]>
Commit | Line | Data |
---|---|---|
1 | ## Copyright (C) 1996-2017 The Squid Software Foundation and contributors | |
2 | ## | |
3 | ## Squid software is distributed under GPLv2+ license and includes | |
4 | ## contributions from numerous individuals and organizations. | |
5 | ## Please see the COPYING and CONTRIBUTORS files for details. | |
6 | ## | |
7 | ||
8 | COMMENT_START | |
9 | WELCOME TO @SQUID@ | |
10 | ---------------------------- | |
11 | ||
12 | This is the documentation for the Squid configuration file. | |
13 | This documentation can also be found online at: | |
14 | http://www.squid-cache.org/Doc/config/ | |
15 | ||
16 | You may wish to look at the Squid home page and wiki for the | |
17 | FAQ and other documentation: | |
18 | http://www.squid-cache.org/ | |
19 | http://wiki.squid-cache.org/SquidFaq | |
20 | http://wiki.squid-cache.org/ConfigExamples | |
21 | ||
22 | This documentation shows what the defaults for various directives | |
23 | happen to be. If you don't need to change the default, you should | |
24 | leave the line out of your squid.conf in most cases. | |
25 | ||
26 | In some cases "none" refers to no default setting at all, | |
27 | while in other cases it refers to the value of the option | |
28 | - the comments for that keyword indicate if this is the case. | |
29 | ||
30 | COMMENT_END | |
31 | ||
32 | COMMENT_START | |
33 | Configuration options can be included using the "include" directive. | |
34 | Include takes a list of files to include. Quoting and wildcards are | |
35 | supported. | |
36 | ||
37 | For example, | |
38 | ||
39 | include /path/to/included/file/squid.acl.config | |
40 | ||
41 | Includes can be nested up to a hard-coded depth of 16 levels. | |
42 | This arbitrary restriction is to prevent recursive include references | |
43 | from causing Squid entering an infinite loop whilst trying to load | |
44 | configuration files. | |
45 | ||
46 | Values with byte units | |
47 | ||
48 | Squid accepts size units on some size related directives. All | |
49 | such directives are documented with a default value displaying | |
50 | a unit. | |
51 | ||
52 | Units accepted by Squid are: | |
53 | bytes - byte | |
54 | KB - Kilobyte (1024 bytes) | |
55 | MB - Megabyte | |
56 | GB - Gigabyte | |
57 | ||
58 | Values with spaces, quotes, and other special characters | |
59 | ||
60 | Squid supports directive parameters with spaces, quotes, and other | |
61 | special characters. Surround such parameters with "double quotes". Use | |
62 | the configuration_includes_quoted_values directive to enable or | |
63 | disable that support. | |
64 | ||
65 | Squid supports reading configuration option parameters from external | |
66 | files using the syntax: | |
67 | parameters("/path/filename") | |
68 | For example: | |
69 | acl whitelist dstdomain parameters("/etc/squid/whitelist.txt") | |
70 | ||
71 | Conditional configuration | |
72 | ||
73 | If-statements can be used to make configuration directives | |
74 | depend on conditions: | |
75 | ||
76 | if <CONDITION> | |
77 | ... regular configuration directives ... | |
78 | [else | |
79 | ... regular configuration directives ...] | |
80 | endif | |
81 | ||
82 | The else part is optional. The keywords "if", "else", and "endif" | |
83 | must be typed on their own lines, as if they were regular | |
84 | configuration directives. | |
85 | ||
86 | NOTE: An else-if condition is not supported. | |
87 | ||
88 | These individual conditions types are supported: | |
89 | ||
90 | true | |
91 | Always evaluates to true. | |
92 | false | |
93 | Always evaluates to false. | |
94 | <integer> = <integer> | |
95 | Equality comparison of two integer numbers. | |
96 | ||
97 | ||
98 | SMP-Related Macros | |
99 | ||
100 | The following SMP-related preprocessor macros can be used. | |
101 | ||
102 | ${process_name} expands to the current Squid process "name" | |
103 | (e.g., squid1, squid2, or cache1). | |
104 | ||
105 | ${process_number} expands to the current Squid process | |
106 | identifier, which is an integer number (e.g., 1, 2, 3) unique | |
107 | across all Squid processes of the current service instance. | |
108 | ||
109 | ${service_name} expands into the current Squid service instance | |
110 | name identifier which is provided by -n on the command line. | |
111 | ||
112 | Logformat Macros | |
113 | ||
114 | Logformat macros can be used in many places outside of the logformat | |
115 | directive. In theory, all of the logformat codes can be used as %macros, | |
116 | where they are supported. In practice, a %macro expands as a dash (-) when | |
117 | the transaction does not yet have enough information and a value is needed. | |
118 | ||
119 | There is no definitive list of what tokens are available at the various | |
120 | stages of the transaction. | |
121 | ||
122 | And some information may already be available to Squid but not yet | |
123 | committed where the macro expansion code can access it (report | |
124 | such instances!). The macro will be expanded into a single dash | |
125 | ('-') in such cases. Not all macros have been tested. | |
126 | ||
127 | COMMENT_END | |
128 | ||
129 | # options still not yet ported from 2.7 to 3.x | |
130 | NAME: broken_vary_encoding | |
131 | TYPE: obsolete | |
132 | DOC_START | |
133 | This option is not yet supported by Squid-3. | |
134 | DOC_END | |
135 | ||
136 | NAME: cache_vary | |
137 | TYPE: obsolete | |
138 | DOC_START | |
139 | This option is not yet supported by Squid-3. | |
140 | DOC_END | |
141 | ||
142 | NAME: error_map | |
143 | TYPE: obsolete | |
144 | DOC_START | |
145 | This option is not yet supported by Squid-3. | |
146 | DOC_END | |
147 | ||
148 | NAME: external_refresh_check | |
149 | TYPE: obsolete | |
150 | DOC_START | |
151 | This option is not yet supported by Squid-3. | |
152 | DOC_END | |
153 | ||
154 | NAME: location_rewrite_program location_rewrite_access location_rewrite_children location_rewrite_concurrency | |
155 | TYPE: obsolete | |
156 | DOC_START | |
157 | This option is not yet supported by Squid-3. | |
158 | DOC_END | |
159 | ||
160 | NAME: refresh_stale_hit | |
161 | TYPE: obsolete | |
162 | DOC_START | |
163 | This option is not yet supported by Squid-3. | |
164 | DOC_END | |
165 | ||
166 | # Options removed in 4.x | |
167 | NAME: cache_peer_domain cache_host_domain | |
168 | TYPE: obsolete | |
169 | DOC_START | |
170 | Replace with dstdomain ACLs and cache_peer_access. | |
171 | DOC_END | |
172 | ||
173 | NAME: ie_refresh | |
174 | TYPE: obsolete | |
175 | DOC_START | |
176 | Remove this line. The behaviour enabled by this is no longer needed. | |
177 | DOC_END | |
178 | ||
179 | NAME: sslproxy_cafile | |
180 | TYPE: obsolete | |
181 | DOC_START | |
182 | Remove this line. Use tls_outgoing_options cafile= instead. | |
183 | DOC_END | |
184 | ||
185 | NAME: sslproxy_capath | |
186 | TYPE: obsolete | |
187 | DOC_START | |
188 | Remove this line. Use tls_outgoing_options capath= instead. | |
189 | DOC_END | |
190 | ||
191 | NAME: sslproxy_cipher | |
192 | TYPE: obsolete | |
193 | DOC_START | |
194 | Remove this line. Use tls_outgoing_options cipher= instead. | |
195 | DOC_END | |
196 | ||
197 | NAME: sslproxy_client_certificate | |
198 | TYPE: obsolete | |
199 | DOC_START | |
200 | Remove this line. Use tls_outgoing_options cert= instead. | |
201 | DOC_END | |
202 | ||
203 | NAME: sslproxy_client_key | |
204 | TYPE: obsolete | |
205 | DOC_START | |
206 | Remove this line. Use tls_outgoing_options key= instead. | |
207 | DOC_END | |
208 | ||
209 | NAME: sslproxy_flags | |
210 | TYPE: obsolete | |
211 | DOC_START | |
212 | Remove this line. Use tls_outgoing_options flags= instead. | |
213 | DOC_END | |
214 | ||
215 | NAME: sslproxy_options | |
216 | TYPE: obsolete | |
217 | DOC_START | |
218 | Remove this line. Use tls_outgoing_options options= instead. | |
219 | DOC_END | |
220 | ||
221 | NAME: sslproxy_version | |
222 | TYPE: obsolete | |
223 | DOC_START | |
224 | Remove this line. Use tls_outgoing_options options= instead. | |
225 | DOC_END | |
226 | ||
227 | # Options removed in 3.5 | |
228 | NAME: hierarchy_stoplist | |
229 | TYPE: obsolete | |
230 | DOC_START | |
231 | Remove this line. Use always_direct or cache_peer_access ACLs instead if you need to prevent cache_peer use. | |
232 | DOC_END | |
233 | ||
234 | # Options removed in 3.4 | |
235 | NAME: log_access | |
236 | TYPE: obsolete | |
237 | DOC_START | |
238 | Remove this line. Use acls with access_log directives to control access logging | |
239 | DOC_END | |
240 | ||
241 | NAME: log_icap | |
242 | TYPE: obsolete | |
243 | DOC_START | |
244 | Remove this line. Use acls with icap_log directives to control icap logging | |
245 | DOC_END | |
246 | ||
247 | # Options Removed in 3.3 | |
248 | NAME: ignore_ims_on_miss | |
249 | TYPE: obsolete | |
250 | DOC_START | |
251 | Remove this line. The HTTP/1.1 feature is now configured by 'cache_miss_revalidate'. | |
252 | DOC_END | |
253 | ||
254 | # Options Removed in 3.2 | |
255 | NAME: chunked_request_body_max_size | |
256 | TYPE: obsolete | |
257 | DOC_START | |
258 | Remove this line. Squid is now HTTP/1.1 compliant. | |
259 | DOC_END | |
260 | ||
261 | NAME: dns_v4_fallback | |
262 | TYPE: obsolete | |
263 | DOC_START | |
264 | Remove this line. Squid performs a 'Happy Eyeballs' algorithm, the 'fallback' algorithm is no longer relevant. | |
265 | DOC_END | |
266 | ||
267 | NAME: emulate_httpd_log | |
268 | TYPE: obsolete | |
269 | DOC_START | |
270 | Replace this with an access_log directive using the format 'common' or 'combined'. | |
271 | DOC_END | |
272 | ||
273 | NAME: forward_log | |
274 | TYPE: obsolete | |
275 | DOC_START | |
276 | Use a regular access.log with ACL limiting it to MISS events. | |
277 | DOC_END | |
278 | ||
279 | NAME: ftp_list_width | |
280 | TYPE: obsolete | |
281 | DOC_START | |
282 | Remove this line. Configure FTP page display using the CSS controls in errorpages.css instead. | |
283 | DOC_END | |
284 | ||
285 | NAME: ignore_expect_100 | |
286 | TYPE: obsolete | |
287 | DOC_START | |
288 | Remove this line. The HTTP/1.1 feature is now fully supported by default. | |
289 | DOC_END | |
290 | ||
291 | NAME: log_fqdn | |
292 | TYPE: obsolete | |
293 | DOC_START | |
294 | Remove this option from your config. To log FQDN use %>A in the log format. | |
295 | DOC_END | |
296 | ||
297 | NAME: log_ip_on_direct | |
298 | TYPE: obsolete | |
299 | DOC_START | |
300 | Remove this option from your config. To log server or peer names use %<A in the log format. | |
301 | DOC_END | |
302 | ||
303 | NAME: maximum_single_addr_tries | |
304 | TYPE: obsolete | |
305 | DOC_START | |
306 | Replaced by connect_retries. The behaviour has changed, please read the documentation before altering. | |
307 | DOC_END | |
308 | ||
309 | NAME: referer_log referrer_log | |
310 | TYPE: obsolete | |
311 | DOC_START | |
312 | Replace this with an access_log directive using the format 'referrer'. | |
313 | DOC_END | |
314 | ||
315 | NAME: update_headers | |
316 | TYPE: obsolete | |
317 | DOC_START | |
318 | Remove this line. The feature is supported by default in storage types where update is implemented. | |
319 | DOC_END | |
320 | ||
321 | NAME: url_rewrite_concurrency | |
322 | TYPE: obsolete | |
323 | DOC_START | |
324 | Remove this line. Set the 'concurrency=' option of url_rewrite_children instead. | |
325 | DOC_END | |
326 | ||
327 | NAME: useragent_log | |
328 | TYPE: obsolete | |
329 | DOC_START | |
330 | Replace this with an access_log directive using the format 'useragent'. | |
331 | DOC_END | |
332 | ||
333 | # Options Removed in 3.1 | |
334 | NAME: dns_testnames | |
335 | TYPE: obsolete | |
336 | DOC_START | |
337 | Remove this line. DNS is no longer tested on startup. | |
338 | DOC_END | |
339 | ||
340 | NAME: extension_methods | |
341 | TYPE: obsolete | |
342 | DOC_START | |
343 | Remove this line. All valid methods for HTTP are accepted by default. | |
344 | DOC_END | |
345 | ||
346 | # 2.7 Options Removed/Replaced in 3.2 | |
347 | NAME: zero_buffers | |
348 | TYPE: obsolete | |
349 | DOC_NONE | |
350 | ||
351 | # 2.7 Options Removed/Replaced in 3.1 | |
352 | NAME: incoming_rate | |
353 | TYPE: obsolete | |
354 | DOC_NONE | |
355 | ||
356 | NAME: server_http11 | |
357 | TYPE: obsolete | |
358 | DOC_START | |
359 | Remove this line. HTTP/1.1 is supported by default. | |
360 | DOC_END | |
361 | ||
362 | NAME: upgrade_http0.9 | |
363 | TYPE: obsolete | |
364 | DOC_START | |
365 | Remove this line. ICY/1.0 streaming protocol is supported by default. | |
366 | DOC_END | |
367 | ||
368 | NAME: zph_local zph_mode zph_option zph_parent zph_sibling | |
369 | TYPE: obsolete | |
370 | DOC_START | |
371 | Alter these entries. Use the qos_flows directive instead. | |
372 | DOC_END | |
373 | ||
374 | # Options Removed in 3.0 | |
375 | NAME: header_access | |
376 | TYPE: obsolete | |
377 | DOC_START | |
378 | Since squid-3.0 replace with request_header_access or reply_header_access | |
379 | depending on whether you wish to match client requests or server replies. | |
380 | DOC_END | |
381 | ||
382 | NAME: httpd_accel_no_pmtu_disc | |
383 | TYPE: obsolete | |
384 | DOC_START | |
385 | Since squid-3.0 use the 'disable-pmtu-discovery' flag on http_port instead. | |
386 | DOC_END | |
387 | ||
388 | NAME: wais_relay_host | |
389 | TYPE: obsolete | |
390 | DOC_START | |
391 | Replace this line with 'cache_peer' configuration. | |
392 | DOC_END | |
393 | ||
394 | NAME: wais_relay_port | |
395 | TYPE: obsolete | |
396 | DOC_START | |
397 | Replace this line with 'cache_peer' configuration. | |
398 | DOC_END | |
399 | ||
400 | COMMENT_START | |
401 | OPTIONS FOR SMP | |
402 | ----------------------------------------------------------------------------- | |
403 | COMMENT_END | |
404 | ||
405 | NAME: workers | |
406 | TYPE: int | |
407 | LOC: Config.workers | |
408 | DEFAULT: 1 | |
409 | DEFAULT_DOC: SMP support disabled. | |
410 | DOC_START | |
411 | Number of main Squid processes or "workers" to fork and maintain. | |
412 | 0: "no daemon" mode, like running "squid -N ..." | |
413 | 1: "no SMP" mode, start one main Squid process daemon (default) | |
414 | N: start N main Squid process daemons (i.e., SMP mode) | |
415 | ||
416 | In SMP mode, each worker does nearly all what a single Squid daemon | |
417 | does (e.g., listen on http_port and forward HTTP requests). | |
418 | DOC_END | |
419 | ||
420 | NAME: cpu_affinity_map | |
421 | TYPE: CpuAffinityMap | |
422 | LOC: Config.cpuAffinityMap | |
423 | DEFAULT: none | |
424 | DEFAULT_DOC: Let operating system decide. | |
425 | DOC_START | |
426 | Usage: cpu_affinity_map process_numbers=P1,P2,... cores=C1,C2,... | |
427 | ||
428 | Sets 1:1 mapping between Squid processes and CPU cores. For example, | |
429 | ||
430 | cpu_affinity_map process_numbers=1,2,3,4 cores=1,3,5,7 | |
431 | ||
432 | affects processes 1 through 4 only and places them on the first | |
433 | four even cores, starting with core #1. | |
434 | ||
435 | CPU cores are numbered starting from 1. Requires support for | |
436 | sched_getaffinity(2) and sched_setaffinity(2) system calls. | |
437 | ||
438 | Multiple cpu_affinity_map options are merged. | |
439 | ||
440 | See also: workers | |
441 | DOC_END | |
442 | ||
443 | NAME: shared_memory_locking | |
444 | TYPE: YesNoNone | |
445 | COMMENT: on|off | |
446 | LOC: Config.shmLocking | |
447 | DEFAULT: off | |
448 | DOC_START | |
449 | Whether to ensure that all required shared memory is available by | |
450 | "locking" that shared memory into RAM when Squid starts. The | |
451 | alternative is faster startup time followed by slightly slower | |
452 | performance and, if not enough RAM is actually available during | |
453 | runtime, mysterious crashes. | |
454 | ||
455 | SMP Squid uses many shared memory segments. These segments are | |
456 | brought into Squid memory space using an mmap(2) system call. During | |
457 | Squid startup, the mmap() call often succeeds regardless of whether | |
458 | the system has enough RAM. In general, Squid cannot tell whether the | |
459 | kernel applies this "optimistic" memory allocation policy (but | |
460 | popular modern kernels usually use it). | |
461 | ||
462 | Later, if Squid attempts to actually access the mapped memory | |
463 | regions beyond what the kernel is willing to allocate, the | |
464 | "optimistic" kernel simply kills Squid kid with a SIGBUS signal. | |
465 | Some of the memory limits enforced by the kernel are currently | |
466 | poorly understood: We do not know how to detect and check them. This | |
467 | option ensures that the mapped memory will be available. | |
468 | ||
469 | This option may have a positive performance side-effect: Locking | |
470 | memory at start avoids runtime paging I/O. Paging slows Squid down. | |
471 | ||
472 | Locking memory may require a large enough RLIMIT_MEMLOCK OS limit, | |
473 | CAP_IPC_LOCK capability, or equivalent. | |
474 | DOC_END | |
475 | ||
476 | COMMENT_START | |
477 | OPTIONS FOR AUTHENTICATION | |
478 | ----------------------------------------------------------------------------- | |
479 | COMMENT_END | |
480 | ||
481 | NAME: auth_param | |
482 | TYPE: authparam | |
483 | IFDEF: USE_AUTH | |
484 | LOC: Auth::TheConfig.schemes | |
485 | DEFAULT: none | |
486 | DOC_START | |
487 | This is used to define parameters for the various authentication | |
488 | schemes supported by Squid. | |
489 | ||
490 | format: auth_param scheme parameter [setting] | |
491 | ||
492 | The order in which authentication schemes are presented to the client is | |
493 | dependent on the order the scheme first appears in config file. IE | |
494 | has a bug (it's not RFC 2617 compliant) in that it will use the basic | |
495 | scheme if basic is the first entry presented, even if more secure | |
496 | schemes are presented. For now use the order in the recommended | |
497 | settings section below. If other browsers have difficulties (don't | |
498 | recognize the schemes offered even if you are using basic) either | |
499 | put basic first, or disable the other schemes (by commenting out their | |
500 | program entry). | |
501 | ||
502 | Once an authentication scheme is fully configured, it can only be | |
503 | shutdown by shutting squid down and restarting. Changes can be made on | |
504 | the fly and activated with a reconfigure. I.E. You can change to a | |
505 | different helper, but not unconfigure the helper completely. | |
506 | ||
507 | Please note that while this directive defines how Squid processes | |
508 | authentication it does not automatically activate authentication. | |
509 | To use authentication you must in addition make use of ACLs based | |
510 | on login name in http_access (proxy_auth, proxy_auth_regex or | |
511 | external with %LOGIN used in the format tag). The browser will be | |
512 | challenged for authentication on the first such acl encountered | |
513 | in http_access processing and will also be re-challenged for new | |
514 | login credentials if the request is being denied by a proxy_auth | |
515 | type acl. | |
516 | ||
517 | WARNING: authentication can't be used in a transparently intercepting | |
518 | proxy as the client then thinks it is talking to an origin server and | |
519 | not the proxy. This is a limitation of bending the TCP/IP protocol to | |
520 | transparently intercepting port 80, not a limitation in Squid. | |
521 | Ports flagged 'transparent', 'intercept', or 'tproxy' have | |
522 | authentication disabled. | |
523 | ||
524 | === Parameters common to all schemes. === | |
525 | ||
526 | "program" cmdline | |
527 | Specifies the command for the external authenticator. | |
528 | ||
529 | By default, each authentication scheme is not used unless a | |
530 | program is specified. | |
531 | ||
532 | See http://wiki.squid-cache.org/Features/AddonHelpers for | |
533 | more details on helper operations and creating your own. | |
534 | ||
535 | "key_extras" format | |
536 | Specifies a string to be append to request line format for | |
537 | the authentication helper. "Quoted" format values may contain | |
538 | spaces and logformat %macros. In theory, any logformat %macro | |
539 | can be used. In practice, a %macro expands as a dash (-) if | |
540 | the helper request is sent before the required macro | |
541 | information is available to Squid. | |
542 | ||
543 | By default, Squid uses request formats provided in | |
544 | scheme-specific examples below (search for %credentials). | |
545 | ||
546 | The expanded key_extras value is added to the Squid credentials | |
547 | cache and, hence, will affect authentication. It can be used to | |
548 | autenticate different users with identical user names (e.g., | |
549 | when user authentication depends on http_port). | |
550 | ||
551 | Avoid adding frequently changing information to key_extras. For | |
552 | example, if you add user source IP, and it changes frequently | |
553 | in your environment, then max_user_ip ACL is going to treat | |
554 | every user+IP combination as a unique "user", breaking the ACL | |
555 | and wasting a lot of memory on those user records. It will also | |
556 | force users to authenticate from scratch whenever their IP | |
557 | changes. | |
558 | ||
559 | "realm" string | |
560 | Specifies the protection scope (aka realm name) which is to be | |
561 | reported to the client for the authentication scheme. It is | |
562 | commonly part of the text the user will see when prompted for | |
563 | their username and password. | |
564 | ||
565 | For Basic the default is "Squid proxy-caching web server". | |
566 | For Digest there is no default, this parameter is mandatory. | |
567 | For NTLM and Negotiate this parameter is ignored. | |
568 | ||
569 | "children" numberofchildren [startup=N] [idle=N] [concurrency=N] | |
570 | [queue-size=N] [on-persistent-overload=action] | |
571 | ||
572 | The maximum number of authenticator processes to spawn. If | |
573 | you start too few Squid will have to wait for them to process | |
574 | a backlog of credential verifications, slowing it down. When | |
575 | password verifications are done via a (slow) network you are | |
576 | likely to need lots of authenticator processes. | |
577 | ||
578 | The startup= and idle= options permit some skew in the exact | |
579 | amount run. A minimum of startup=N will begin during startup | |
580 | and reconfigure. Squid will start more in groups of up to | |
581 | idle=N in an attempt to meet traffic needs and to keep idle=N | |
582 | free above those traffic needs up to the maximum. | |
583 | ||
584 | The concurrency= option sets the number of concurrent requests | |
585 | the helper can process. The default of 0 is used for helpers | |
586 | who only supports one request at a time. Setting this to a | |
587 | number greater than 0 changes the protocol used to include a | |
588 | channel ID field first on the request/response line, allowing | |
589 | multiple requests to be sent to the same helper in parallel | |
590 | without waiting for the response. | |
591 | ||
592 | Concurrency must not be set unless it's known the helper | |
593 | supports the input format with channel-ID fields. | |
594 | ||
595 | The queue-size=N option sets the maximum number of queued | |
596 | requests to N. The default maximum is 2*numberofchildren. Squid | |
597 | is allowed to temporarily exceed the configured maximum, marking | |
598 | the affected helper as "overloaded". If the helper overload | |
599 | lasts more than 3 minutes, the action prescribed by the | |
600 | on-persistent-overload option applies. | |
601 | ||
602 | The on-persistent-overload=action option specifies Squid | |
603 | reaction to a new helper request arriving when the helper | |
604 | has been overloaded for more that 3 minutes already. The number | |
605 | of queued requests determines whether the helper is overloaded | |
606 | (see the queue-size option). | |
607 | ||
608 | Two actions are supported: | |
609 | ||
610 | die Squid worker quits. This is the default behavior. | |
611 | ||
612 | ERR Squid treats the helper request as if it was | |
613 | immediately submitted, and the helper immediately | |
614 | replied with an ERR response. This action has no effect | |
615 | on the already queued and in-progress helper requests. | |
616 | ||
617 | NOTE: NTLM and Negotiate schemes do not support concurrency | |
618 | in the Squid code module even though some helpers can. | |
619 | ||
620 | "keep_alive" on|off | |
621 | If you experience problems with PUT/POST requests when using | |
622 | the NTLM or Negotiate schemes then you can try setting this | |
623 | to off. This will cause Squid to forcibly close the connection | |
624 | on the initial request where the browser asks which schemes | |
625 | are supported by the proxy. | |
626 | ||
627 | For Basic and Digest this parameter is ignored. | |
628 | ||
629 | "utf8" on|off | |
630 | HTTP uses iso-latin-1 as character set, while some | |
631 | authentication backends such as LDAP expects UTF-8. If this is | |
632 | set to on Squid will translate the HTTP iso-latin-1 charset to | |
633 | UTF-8 before sending the username and password to the helper. | |
634 | ||
635 | For NTLM and Negotiate this parameter is ignored. | |
636 | ||
637 | IF HAVE_AUTH_MODULE_BASIC | |
638 | === Basic authentication parameters === | |
639 | ||
640 | "credentialsttl" timetolive | |
641 | Specifies how long squid assumes an externally validated | |
642 | username:password pair is valid for - in other words how | |
643 | often the helper program is called for that user. Set this | |
644 | low to force revalidation with short lived passwords. | |
645 | ||
646 | NOTE: setting this high does not impact your susceptibility | |
647 | to replay attacks unless you are using an one-time password | |
648 | system (such as SecureID). If you are using such a system, | |
649 | you will be vulnerable to replay attacks unless you also | |
650 | use the max_user_ip ACL in an http_access rule. | |
651 | ||
652 | "casesensitive" on|off | |
653 | Specifies if usernames are case sensitive. Most user databases | |
654 | are case insensitive allowing the same username to be spelled | |
655 | using both lower and upper case letters, but some are case | |
656 | sensitive. This makes a big difference for user_max_ip ACL | |
657 | processing and similar. | |
658 | ||
659 | ENDIF | |
660 | IF HAVE_AUTH_MODULE_DIGEST | |
661 | === Digest authentication parameters === | |
662 | ||
663 | "nonce_garbage_interval" timeinterval | |
664 | Specifies the interval that nonces that have been issued | |
665 | to client_agent's are checked for validity. | |
666 | ||
667 | "nonce_max_duration" timeinterval | |
668 | Specifies the maximum length of time a given nonce will be | |
669 | valid for. | |
670 | ||
671 | "nonce_max_count" number | |
672 | Specifies the maximum number of times a given nonce can be | |
673 | used. | |
674 | ||
675 | "nonce_strictness" on|off | |
676 | Determines if squid requires strict increment-by-1 behavior | |
677 | for nonce counts, or just incrementing (off - for use when | |
678 | user agents generate nonce counts that occasionally miss 1 | |
679 | (ie, 1,2,4,6)). Default off. | |
680 | ||
681 | "check_nonce_count" on|off | |
682 | This directive if set to off can disable the nonce count check | |
683 | completely to work around buggy digest qop implementations in | |
684 | certain mainstream browser versions. Default on to check the | |
685 | nonce count to protect from authentication replay attacks. | |
686 | ||
687 | "post_workaround" on|off | |
688 | This is a workaround to certain buggy browsers who send an | |
689 | incorrect request digest in POST requests when reusing the | |
690 | same nonce as acquired earlier on a GET request. | |
691 | ||
692 | ENDIF | |
693 | ||
694 | === Example Configuration === | |
695 | ||
696 | This configuration displays the recommended authentication scheme | |
697 | order from most to least secure with recommended minimum configuration | |
698 | settings for each scheme: | |
699 | ||
700 | #auth_param negotiate program <uncomment and complete this line to activate> | |
701 | #auth_param negotiate children 20 startup=0 idle=1 | |
702 | # | |
703 | #auth_param digest program <uncomment and complete this line to activate> | |
704 | #auth_param digest children 20 startup=0 idle=1 | |
705 | #auth_param digest realm Squid proxy-caching web server | |
706 | #auth_param digest nonce_garbage_interval 5 minutes | |
707 | #auth_param digest nonce_max_duration 30 minutes | |
708 | #auth_param digest nonce_max_count 50 | |
709 | # | |
710 | #auth_param ntlm program <uncomment and complete this line to activate> | |
711 | #auth_param ntlm children 20 startup=0 idle=1 | |
712 | # | |
713 | #auth_param basic program <uncomment and complete this line> | |
714 | #auth_param basic children 5 startup=5 idle=1 | |
715 | #auth_param basic credentialsttl 2 hours | |
716 | DOC_END | |
717 | ||
718 | NAME: authenticate_cache_garbage_interval | |
719 | IFDEF: USE_AUTH | |
720 | TYPE: time_t | |
721 | DEFAULT: 1 hour | |
722 | LOC: Auth::TheConfig.garbageCollectInterval | |
723 | DOC_START | |
724 | The time period between garbage collection across the username cache. | |
725 | This is a trade-off between memory utilization (long intervals - say | |
726 | 2 days) and CPU (short intervals - say 1 minute). Only change if you | |
727 | have good reason to. | |
728 | DOC_END | |
729 | ||
730 | NAME: authenticate_ttl | |
731 | IFDEF: USE_AUTH | |
732 | TYPE: time_t | |
733 | DEFAULT: 1 hour | |
734 | LOC: Auth::TheConfig.credentialsTtl | |
735 | DOC_START | |
736 | The time a user & their credentials stay in the logged in | |
737 | user cache since their last request. When the garbage | |
738 | interval passes, all user credentials that have passed their | |
739 | TTL are removed from memory. | |
740 | DOC_END | |
741 | ||
742 | NAME: authenticate_ip_ttl | |
743 | IFDEF: USE_AUTH | |
744 | TYPE: time_t | |
745 | LOC: Auth::TheConfig.ipTtl | |
746 | DEFAULT: 1 second | |
747 | DOC_START | |
748 | If you use proxy authentication and the 'max_user_ip' ACL, | |
749 | this directive controls how long Squid remembers the IP | |
750 | addresses associated with each user. Use a small value | |
751 | (e.g., 60 seconds) if your users might change addresses | |
752 | quickly, as is the case with dialup. You might be safe | |
753 | using a larger value (e.g., 2 hours) in a corporate LAN | |
754 | environment with relatively static address assignments. | |
755 | DOC_END | |
756 | ||
757 | COMMENT_START | |
758 | ACCESS CONTROLS | |
759 | ----------------------------------------------------------------------------- | |
760 | COMMENT_END | |
761 | ||
762 | NAME: external_acl_type | |
763 | TYPE: externalAclHelper | |
764 | LOC: Config.externalAclHelperList | |
765 | DEFAULT: none | |
766 | DOC_START | |
767 | This option defines external acl classes using a helper program | |
768 | to look up the status | |
769 | ||
770 | external_acl_type name [options] FORMAT /path/to/helper [helper arguments] | |
771 | ||
772 | Options: | |
773 | ||
774 | ttl=n TTL in seconds for cached results (defaults to 3600 | |
775 | for 1 hour) | |
776 | ||
777 | negative_ttl=n | |
778 | TTL for cached negative lookups (default same | |
779 | as ttl) | |
780 | ||
781 | grace=n Percentage remaining of TTL where a refresh of a | |
782 | cached entry should be initiated without needing to | |
783 | wait for a new reply. (default is for no grace period) | |
784 | ||
785 | cache=n The maximum number of entries in the result cache. The | |
786 | default limit is 262144 entries. Each cache entry usually | |
787 | consumes at least 256 bytes. Squid currently does not remove | |
788 | expired cache entries until the limit is reached, so a proxy | |
789 | will sooner or later reach the limit. The expanded FORMAT | |
790 | value is used as the cache key, so if the details in FORMAT | |
791 | are highly variable, a larger cache may be needed to produce | |
792 | reduction in helper load. | |
793 | ||
794 | children-max=n | |
795 | Maximum number of acl helper processes spawned to service | |
796 | external acl lookups of this type. (default 5) | |
797 | ||
798 | children-startup=n | |
799 | Minimum number of acl helper processes to spawn during | |
800 | startup and reconfigure to service external acl lookups | |
801 | of this type. (default 0) | |
802 | ||
803 | children-idle=n | |
804 | Number of acl helper processes to keep ahead of traffic | |
805 | loads. Squid will spawn this many at once whenever load | |
806 | rises above the capabilities of existing processes. | |
807 | Up to the value of children-max. (default 1) | |
808 | ||
809 | concurrency=n concurrency level per process. Only used with helpers | |
810 | capable of processing more than one query at a time. | |
811 | ||
812 | queue-size=N The queue-size= option sets the maximum number of queued | |
813 | requests. If the queued requests exceed queue size | |
814 | the acl is ignored. | |
815 | The default value is set to 2*children-max. | |
816 | ||
817 | protocol=2.5 Compatibility mode for Squid-2.5 external acl helpers. | |
818 | ||
819 | ipv4 / ipv6 IP protocol used to communicate with this helper. | |
820 | The default is to auto-detect IPv6 and use it when available. | |
821 | ||
822 | ||
823 | FORMAT is a series of %macro codes. See logformat directive for a full list | |
824 | of the accepted codes. Although note that at the time of any external ACL | |
825 | being tested data may not be available and thus some %macro expand to '-'. | |
826 | ||
827 | In addition to the logformat codes; when processing external ACLs these | |
828 | additional macros are made available: | |
829 | ||
830 | %ACL The name of the ACL being tested. | |
831 | ||
832 | %DATA The ACL arguments specified in the referencing config | |
833 | 'acl ... external' line, separated by spaces (an | |
834 | "argument string"). see acl external. | |
835 | ||
836 | If there are no ACL arguments %DATA expands to '-'. | |
837 | ||
838 | If you do not specify a DATA macro inside FORMAT, | |
839 | Squid automatically appends %DATA to your FORMAT. | |
840 | ||
841 | By default, Squid applies URL-encoding to each ACL | |
842 | argument inside the argument string. If an explicit | |
843 | encoding modifier is used (e.g., %#DATA), then Squid | |
844 | encodes the whole argument string as a single token | |
845 | (e.g., with %#DATA, spaces between arguments become | |
846 | %20). | |
847 | ||
848 | If SSL is enabled, the following formating codes become available: | |
849 | ||
850 | %USER_CERT SSL User certificate in PEM format | |
851 | %USER_CERTCHAIN SSL User certificate chain in PEM format | |
852 | %USER_CERT_xx SSL User certificate subject attribute xx | |
853 | %USER_CA_CERT_xx SSL User certificate issuer attribute xx | |
854 | ||
855 | ||
856 | NOTE: all other format codes accepted by older Squid versions | |
857 | are deprecated. | |
858 | ||
859 | ||
860 | General request syntax: | |
861 | ||
862 | [channel-ID] FORMAT-values | |
863 | ||
864 | ||
865 | FORMAT-values consists of transaction details expanded with | |
866 | whitespace separation per the config file FORMAT specification | |
867 | using the FORMAT macros listed above. | |
868 | ||
869 | Request values sent to the helper are URL escaped to protect | |
870 | each value in requests against whitespaces. | |
871 | ||
872 | If using protocol=2.5 then the request sent to the helper is not | |
873 | URL escaped to protect against whitespace. | |
874 | ||
875 | NOTE: protocol=3.0 is deprecated as no longer necessary. | |
876 | ||
877 | When using the concurrency= option the protocol is changed by | |
878 | introducing a query channel tag in front of the request/response. | |
879 | The query channel tag is a number between 0 and concurrency-1. | |
880 | This value must be echoed back unchanged to Squid as the first part | |
881 | of the response relating to its request. | |
882 | ||
883 | ||
884 | The helper receives lines expanded per the above format specification | |
885 | and for each input line returns 1 line starting with OK/ERR/BH result | |
886 | code and optionally followed by additional keywords with more details. | |
887 | ||
888 | ||
889 | General result syntax: | |
890 | ||
891 | [channel-ID] result keyword=value ... | |
892 | ||
893 | Result consists of one of the codes: | |
894 | ||
895 | OK | |
896 | the ACL test produced a match. | |
897 | ||
898 | ERR | |
899 | the ACL test does not produce a match. | |
900 | ||
901 | BH | |
902 | An internal error occurred in the helper, preventing | |
903 | a result being identified. | |
904 | ||
905 | The meaning of 'a match' is determined by your squid.conf | |
906 | access control configuration. See the Squid wiki for details. | |
907 | ||
908 | Defined keywords: | |
909 | ||
910 | user= The users name (login) | |
911 | ||
912 | password= The users password (for login= cache_peer option) | |
913 | ||
914 | message= Message describing the reason for this response. | |
915 | Available as %o in error pages. | |
916 | Useful on (ERR and BH results). | |
917 | ||
918 | tag= Apply a tag to a request. Only sets a tag once, | |
919 | does not alter existing tags. | |
920 | ||
921 | log= String to be logged in access.log. Available as | |
922 | %ea in logformat specifications. | |
923 | ||
924 | clt_conn_tag= Associates a TAG with the client TCP connection. | |
925 | Please see url_rewrite_program related documentation | |
926 | for this kv-pair. | |
927 | ||
928 | Any keywords may be sent on any response whether OK, ERR or BH. | |
929 | ||
930 | All response keyword values need to be a single token with URL | |
931 | escaping, or enclosed in double quotes (") and escaped using \ on | |
932 | any double quotes or \ characters within the value. The wrapping | |
933 | double quotes are removed before the value is interpreted by Squid. | |
934 | \r and \n are also replace by CR and LF. | |
935 | ||
936 | Some example key values: | |
937 | ||
938 | user=John%20Smith | |
939 | user="John Smith" | |
940 | user="J. \"Bob\" Smith" | |
941 | DOC_END | |
942 | ||
943 | NAME: acl | |
944 | TYPE: acl | |
945 | LOC: Config.aclList | |
946 | IF USE_OPENSSL | |
947 | DEFAULT: ssl::certHasExpired ssl_error X509_V_ERR_CERT_HAS_EXPIRED | |
948 | DEFAULT: ssl::certNotYetValid ssl_error X509_V_ERR_CERT_NOT_YET_VALID | |
949 | DEFAULT: ssl::certDomainMismatch ssl_error SQUID_X509_V_ERR_DOMAIN_MISMATCH | |
950 | DEFAULT: ssl::certUntrusted ssl_error X509_V_ERR_INVALID_CA X509_V_ERR_SELF_SIGNED_CERT_IN_CHAIN X509_V_ERR_UNABLE_TO_VERIFY_LEAF_SIGNATURE X509_V_ERR_UNABLE_TO_GET_ISSUER_CERT X509_V_ERR_UNABLE_TO_GET_ISSUER_CERT_LOCALLY X509_V_ERR_CERT_UNTRUSTED | |
951 | DEFAULT: ssl::certSelfSigned ssl_error X509_V_ERR_DEPTH_ZERO_SELF_SIGNED_CERT | |
952 | ENDIF | |
953 | DEFAULT: all src all | |
954 | DEFAULT: manager url_regex -i ^cache_object:// +i ^https?://[^/]+/squid-internal-mgr/ | |
955 | DEFAULT: localhost src 127.0.0.1/32 ::1 | |
956 | DEFAULT: to_localhost dst 127.0.0.0/8 0.0.0.0/32 ::1 | |
957 | DEFAULT: CONNECT method CONNECT | |
958 | DEFAULT_DOC: ACLs all, manager, localhost, to_localhost, and CONNECT are predefined. | |
959 | DOC_START | |
960 | Defining an Access List | |
961 | ||
962 | Every access list definition must begin with an aclname and acltype, | |
963 | followed by either type-specific arguments or a quoted filename that | |
964 | they are read from. | |
965 | ||
966 | acl aclname acltype argument ... | |
967 | acl aclname acltype "file" ... | |
968 | ||
969 | When using "file", the file should contain one item per line. | |
970 | ||
971 | ||
972 | ACL Options | |
973 | ||
974 | Some acl types supports options which changes their default behaviour: | |
975 | ||
976 | -i,+i By default, regular expressions are CASE-SENSITIVE. To make them | |
977 | case-insensitive, use the -i option. To return case-sensitive | |
978 | use the +i option between patterns, or make a new ACL line | |
979 | without -i. | |
980 | ||
981 | -n Disable lookups and address type conversions. If lookup or | |
982 | conversion is required because the parameter type (IP or | |
983 | domain name) does not match the message address type (domain | |
984 | name or IP), then the ACL would immediately declare a mismatch | |
985 | without any warnings or lookups. | |
986 | ||
987 | -m[=delimiters] | |
988 | Perform a list membership test, interpreting values as | |
989 | comma-separated token lists and matching against individual | |
990 | tokens instead of whole values. | |
991 | The optional "delimiters" parameter specifies one or more | |
992 | alternative non-alphanumeric delimiter characters. | |
993 | non-alphanumeric delimiter characters. | |
994 | ||
995 | -- Used to stop processing all options, in the case the first acl | |
996 | value has '-' character as first character (for example the '-' | |
997 | is a valid domain name) | |
998 | ||
999 | Some acl types require suspending the current request in order | |
1000 | to access some external data source. | |
1001 | Those which do are marked with the tag [slow], those which | |
1002 | don't are marked as [fast]. | |
1003 | See http://wiki.squid-cache.org/SquidFaq/SquidAcl | |
1004 | for further information | |
1005 | ||
1006 | ***** ACL TYPES AVAILABLE ***** | |
1007 | ||
1008 | acl aclname src ip-address/mask ... # clients IP address [fast] | |
1009 | acl aclname src addr1-addr2/mask ... # range of addresses [fast] | |
1010 | acl aclname dst [-n] ip-address/mask ... # URL host's IP address [slow] | |
1011 | acl aclname localip ip-address/mask ... # IP address the client connected to [fast] | |
1012 | ||
1013 | acl aclname arp mac-address ... (xx:xx:xx:xx:xx:xx notation) | |
1014 | # [fast] | |
1015 | # The 'arp' ACL code is not portable to all operating systems. | |
1016 | # It works on Linux, Solaris, Windows, FreeBSD, and some other | |
1017 | # BSD variants. | |
1018 | # | |
1019 | # NOTE: Squid can only determine the MAC/EUI address for IPv4 | |
1020 | # clients that are on the same subnet. If the client is on a | |
1021 | # different subnet, then Squid cannot find out its address. | |
1022 | # | |
1023 | # NOTE 2: IPv6 protocol does not contain ARP. MAC/EUI is either | |
1024 | # encoded directly in the IPv6 address or not available. | |
1025 | ||
1026 | acl aclname srcdomain .foo.com ... | |
1027 | # reverse lookup, from client IP [slow] | |
1028 | acl aclname dstdomain [-n] .foo.com ... | |
1029 | # Destination server from URL [fast] | |
1030 | acl aclname srcdom_regex [-i] \.foo\.com ... | |
1031 | # regex matching client name [slow] | |
1032 | acl aclname dstdom_regex [-n] [-i] \.foo\.com ... | |
1033 | # regex matching server [fast] | |
1034 | # | |
1035 | # For dstdomain and dstdom_regex a reverse lookup is tried if a IP | |
1036 | # based URL is used and no match is found. The name "none" is used | |
1037 | # if the reverse lookup fails. | |
1038 | ||
1039 | acl aclname src_as number ... | |
1040 | acl aclname dst_as number ... | |
1041 | # [fast] | |
1042 | # Except for access control, AS numbers can be used for | |
1043 | # routing of requests to specific caches. Here's an | |
1044 | # example for routing all requests for AS#1241 and only | |
1045 | # those to mycache.mydomain.net: | |
1046 | # acl asexample dst_as 1241 | |
1047 | # cache_peer_access mycache.mydomain.net allow asexample | |
1048 | # cache_peer_access mycache_mydomain.net deny all | |
1049 | ||
1050 | acl aclname peername myPeer ... | |
1051 | # [fast] | |
1052 | # match against a named cache_peer entry | |
1053 | # set unique name= on cache_peer lines for reliable use. | |
1054 | ||
1055 | acl aclname time [day-abbrevs] [h1:m1-h2:m2] | |
1056 | # [fast] | |
1057 | # day-abbrevs: | |
1058 | # S - Sunday | |
1059 | # M - Monday | |
1060 | # T - Tuesday | |
1061 | # W - Wednesday | |
1062 | # H - Thursday | |
1063 | # F - Friday | |
1064 | # A - Saturday | |
1065 | # h1:m1 must be less than h2:m2 | |
1066 | ||
1067 | acl aclname url_regex [-i] ^http:// ... | |
1068 | # regex matching on whole URL [fast] | |
1069 | acl aclname urllogin [-i] [^a-zA-Z0-9] ... | |
1070 | # regex matching on URL login field | |
1071 | acl aclname urlpath_regex [-i] \.gif$ ... | |
1072 | # regex matching on URL path [fast] | |
1073 | ||
1074 | acl aclname port 80 70 21 0-1024... # destination TCP port [fast] | |
1075 | # ranges are alloed | |
1076 | acl aclname localport 3128 ... # TCP port the client connected to [fast] | |
1077 | # NP: for interception mode this is usually '80' | |
1078 | ||
1079 | acl aclname myportname 3128 ... # *_port name [fast] | |
1080 | ||
1081 | acl aclname proto HTTP FTP ... # request protocol [fast] | |
1082 | ||
1083 | acl aclname method GET POST ... # HTTP request method [fast] | |
1084 | ||
1085 | acl aclname http_status 200 301 500- 400-403 ... | |
1086 | # status code in reply [fast] | |
1087 | ||
1088 | acl aclname browser [-i] regexp ... | |
1089 | # pattern match on User-Agent header (see also req_header below) [fast] | |
1090 | ||
1091 | acl aclname referer_regex [-i] regexp ... | |
1092 | # pattern match on Referer header [fast] | |
1093 | # Referer is highly unreliable, so use with care | |
1094 | ||
1095 | acl aclname ident username ... | |
1096 | acl aclname ident_regex [-i] pattern ... | |
1097 | # string match on ident output [slow] | |
1098 | # use REQUIRED to accept any non-null ident. | |
1099 | ||
1100 | acl aclname proxy_auth [-i] username ... | |
1101 | acl aclname proxy_auth_regex [-i] pattern ... | |
1102 | # perform http authentication challenge to the client and match against | |
1103 | # supplied credentials [slow] | |
1104 | # | |
1105 | # takes a list of allowed usernames. | |
1106 | # use REQUIRED to accept any valid username. | |
1107 | # | |
1108 | # Will use proxy authentication in forward-proxy scenarios, and plain | |
1109 | # http authenticaiton in reverse-proxy scenarios | |
1110 | # | |
1111 | # NOTE: when a Proxy-Authentication header is sent but it is not | |
1112 | # needed during ACL checking the username is NOT logged | |
1113 | # in access.log. | |
1114 | # | |
1115 | # NOTE: proxy_auth requires a EXTERNAL authentication program | |
1116 | # to check username/password combinations (see | |
1117 | # auth_param directive). | |
1118 | # | |
1119 | # NOTE: proxy_auth can't be used in a transparent/intercepting proxy | |
1120 | # as the browser needs to be configured for using a proxy in order | |
1121 | # to respond to proxy authentication. | |
1122 | ||
1123 | acl aclname snmp_community string ... | |
1124 | # A community string to limit access to your SNMP Agent [fast] | |
1125 | # Example: | |
1126 | # | |
1127 | # acl snmppublic snmp_community public | |
1128 | ||
1129 | acl aclname maxconn number | |
1130 | # This will be matched when the client's IP address has | |
1131 | # more than <number> TCP connections established. [fast] | |
1132 | # NOTE: This only measures direct TCP links so X-Forwarded-For | |
1133 | # indirect clients are not counted. | |
1134 | ||
1135 | acl aclname max_user_ip [-s] number | |
1136 | # This will be matched when the user attempts to log in from more | |
1137 | # than <number> different ip addresses. The authenticate_ip_ttl | |
1138 | # parameter controls the timeout on the ip entries. [fast] | |
1139 | # If -s is specified the limit is strict, denying browsing | |
1140 | # from any further IP addresses until the ttl has expired. Without | |
1141 | # -s Squid will just annoy the user by "randomly" denying requests. | |
1142 | # (the counter is reset each time the limit is reached and a | |
1143 | # request is denied) | |
1144 | # NOTE: in acceleration mode or where there is mesh of child proxies, | |
1145 | # clients may appear to come from multiple addresses if they are | |
1146 | # going through proxy farms, so a limit of 1 may cause user problems. | |
1147 | ||
1148 | acl aclname random probability | |
1149 | # Pseudo-randomly match requests. Based on the probability given. | |
1150 | # Probability may be written as a decimal (0.333), fraction (1/3) | |
1151 | # or ratio of matches:non-matches (3:5). | |
1152 | ||
1153 | acl aclname req_mime_type [-i] mime-type ... | |
1154 | # regex match against the mime type of the request generated | |
1155 | # by the client. Can be used to detect file upload or some | |
1156 | # types HTTP tunneling requests [fast] | |
1157 | # NOTE: This does NOT match the reply. You cannot use this | |
1158 | # to match the returned file type. | |
1159 | ||
1160 | acl aclname req_header header-name [-i] any\.regex\.here | |
1161 | # regex match against any of the known request headers. May be | |
1162 | # thought of as a superset of "browser", "referer" and "mime-type" | |
1163 | # ACL [fast] | |
1164 | ||
1165 | acl aclname rep_mime_type [-i] mime-type ... | |
1166 | # regex match against the mime type of the reply received by | |
1167 | # squid. Can be used to detect file download or some | |
1168 | # types HTTP tunneling requests. [fast] | |
1169 | # NOTE: This has no effect in http_access rules. It only has | |
1170 | # effect in rules that affect the reply data stream such as | |
1171 | # http_reply_access. | |
1172 | ||
1173 | acl aclname rep_header header-name [-i] any\.regex\.here | |
1174 | # regex match against any of the known reply headers. May be | |
1175 | # thought of as a superset of "browser", "referer" and "mime-type" | |
1176 | # ACLs [fast] | |
1177 | ||
1178 | acl aclname external class_name [arguments...] | |
1179 | # external ACL lookup via a helper class defined by the | |
1180 | # external_acl_type directive [slow] | |
1181 | ||
1182 | acl aclname user_cert attribute values... | |
1183 | # match against attributes in a user SSL certificate | |
1184 | # attribute is one of DN/C/O/CN/L/ST or a numerical OID [fast] | |
1185 | ||
1186 | acl aclname ca_cert attribute values... | |
1187 | # match against attributes a users issuing CA SSL certificate | |
1188 | # attribute is one of DN/C/O/CN/L/ST or a numerical OID [fast] | |
1189 | ||
1190 | acl aclname ext_user username ... | |
1191 | acl aclname ext_user_regex [-i] pattern ... | |
1192 | # string match on username returned by external acl helper [slow] | |
1193 | # use REQUIRED to accept any non-null user name. | |
1194 | ||
1195 | acl aclname tag tagvalue ... | |
1196 | # string match on tag returned by external acl helper [fast] | |
1197 | # DEPRECATED. Only the first tag will match with this ACL. | |
1198 | # Use the 'note' ACL instead for handling multiple tag values. | |
1199 | ||
1200 | acl aclname hier_code codename ... | |
1201 | # string match against squid hierarchy code(s); [fast] | |
1202 | # e.g., DIRECT, PARENT_HIT, NONE, etc. | |
1203 | # | |
1204 | # NOTE: This has no effect in http_access rules. It only has | |
1205 | # effect in rules that affect the reply data stream such as | |
1206 | # http_reply_access. | |
1207 | ||
1208 | acl aclname note [-m[=delimiters]] name [value ...] | |
1209 | # match transaction annotation [fast] | |
1210 | # Without values, matches any annotation with a given name. | |
1211 | # With value(s), matches any annotation with a given name that | |
1212 | # also has one of the given values. | |
1213 | # If the -m flag is used, then the value of the named | |
1214 | # annotation is interpreted as a list of tokens, and the ACL | |
1215 | # matches individual name=token pairs rather than whole | |
1216 | # name=value pairs. See "ACL Options" above for more info. | |
1217 | # Annotation sources include note and adaptation_meta directives | |
1218 | # as well as helper and eCAP responses. | |
1219 | ||
1220 | acl aclname annotate_transaction [-m[=delimiters]] key=value ... | |
1221 | acl aclname annotate_transaction [-m[=delimiters]] key+=value ... | |
1222 | # Always matches. [fast] | |
1223 | # Used for its side effect: This ACL immediately adds a | |
1224 | # key=value annotation to the current master transaction. | |
1225 | # The added annotation can then be tested using note ACL and | |
1226 | # logged (or sent to helpers) using %note format code. | |
1227 | # | |
1228 | # Annotations can be specified using replacement and addition | |
1229 | # formats. The key=value form replaces old same-key annotation | |
1230 | # value(s). The key+=value form appends a new value to the old | |
1231 | # same-key annotation. Both forms create a new key=value | |
1232 | # annotation if no same-key annotation exists already. If | |
1233 | # -m flag is used, then the value is interpreted as a list | |
1234 | # and the annotation will contain key=token pair(s) instead of the | |
1235 | # whole key=value pair. | |
1236 | # | |
1237 | # This ACL is especially useful for recording complex multi-step | |
1238 | # ACL-driven decisions. For example, the following configuration | |
1239 | # avoids logging transactions accepted after aclX matched: | |
1240 | # | |
1241 | # # First, mark transactions accepted after aclX matched | |
1242 | # acl markSpecial annotate_transaction special=true | |
1243 | # http_access allow acl001 | |
1244 | # ... | |
1245 | # http_access deny acl100 | |
1246 | # http_access allow aclX markSpecial | |
1247 | # | |
1248 | # # Second, do not log marked transactions: | |
1249 | # acl markedSpecial note special true | |
1250 | # access_log ... deny markedSpecial | |
1251 | # | |
1252 | # # Note that the following would not have worked because aclX | |
1253 | # # alone does not determine whether the transaction was allowed: | |
1254 | # access_log ... deny aclX # Wrong! | |
1255 | # | |
1256 | # Warning: This ACL annotates the transaction even when negated | |
1257 | # and even if subsequent ACLs fail to match. For example, the | |
1258 | # following three rules will have exactly the same effect as far | |
1259 | # as annotations set by the "mark" ACL are concerned: | |
1260 | # | |
1261 | # some_directive acl1 ... mark # rule matches if mark is reached | |
1262 | # some_directive acl1 ... !mark # rule never matches | |
1263 | # some_directive acl1 ... mark !all # rule never matches | |
1264 | ||
1265 | acl aclname annotate_client [-m[=delimiters]] key=value ... | |
1266 | acl aclname annotate_client [-m[=delimiters]] key+=value ... | |
1267 | # | |
1268 | # Always matches. [fast] | |
1269 | # Used for its side effect: This ACL immediately adds a | |
1270 | # key=value annotation to the current client-to-Squid | |
1271 | # connection. Connection annotations are propagated to the current | |
1272 | # and all future master transactions on the annotated connection. | |
1273 | # See the annotate_transaction ACL for details. | |
1274 | # | |
1275 | # For example, the following configuration avoids rewriting URLs | |
1276 | # of transactions bumped by SslBump: | |
1277 | # | |
1278 | # # First, mark bumped connections: | |
1279 | # acl markBumped annotate_client bumped=true | |
1280 | # ssl_bump peek acl1 | |
1281 | # ssl_bump stare acl2 | |
1282 | # ssl_bump bump acl3 markBumped | |
1283 | # ssl_bump splice all | |
1284 | # | |
1285 | # # Second, do not send marked transactions to the redirector: | |
1286 | # acl markedBumped note bumped true | |
1287 | # url_rewrite_access deny markedBumped | |
1288 | # | |
1289 | # # Note that the following would not have worked because acl3 alone | |
1290 | # # does not determine whether the connection is going to be bumped: | |
1291 | # url_rewrite_access deny acl3 # Wrong! | |
1292 | ||
1293 | acl aclname adaptation_service service ... | |
1294 | # Matches the name of any icap_service, ecap_service, | |
1295 | # adaptation_service_set, or adaptation_service_chain that Squid | |
1296 | # has used (or attempted to use) for the master transaction. | |
1297 | # This ACL must be defined after the corresponding adaptation | |
1298 | # service is named in squid.conf. This ACL is usable with | |
1299 | # adaptation_meta because it starts matching immediately after | |
1300 | # the service has been selected for adaptation. | |
1301 | ||
1302 | acl aclname transaction_initiator initiator ... | |
1303 | # Matches transaction's initiator [fast] | |
1304 | # | |
1305 | # Supported initiators are: | |
1306 | # esi: matches transactions fetching ESI resources | |
1307 | # certificate-fetching: matches transactions fetching | |
1308 | # a missing intermediate TLS certificate | |
1309 | # cache-digest: matches transactions fetching Cache Digests | |
1310 | # from a cache_peer | |
1311 | # htcp: matches HTCP requests from peers | |
1312 | # icp: matches ICP requests to peers | |
1313 | # icmp: matches ICMP RTT database (NetDB) requests to peers | |
1314 | # asn: matches asns db requests | |
1315 | # internal: matches any of the above | |
1316 | # client: matches transactions containing an HTTP or FTP | |
1317 | # client request received at a Squid *_port | |
1318 | # all: matches any transaction, including internal transactions | |
1319 | # without a configurable initiator and hopefully rare | |
1320 | # transactions without a known-to-Squid initiator | |
1321 | # | |
1322 | # Multiple initiators are ORed. | |
1323 | ||
1324 | acl aclname has component | |
1325 | # matches a transaction "component" [fast] | |
1326 | # | |
1327 | # Supported transaction components are: | |
1328 | # request: transaction has a request header (at least) | |
1329 | # response: transaction has a response header (at least) | |
1330 | # ALE: transaction has an internally-generated Access Log Entry | |
1331 | # structure; bugs notwithstanding, all transaction have it | |
1332 | # | |
1333 | # For example, the following configuration helps when dealing with HTTP | |
1334 | # clients that close connections without sending a request header: | |
1335 | # | |
1336 | # acl hasRequest has request | |
1337 | # acl logMe note important_transaction | |
1338 | # # avoid "logMe ACL is used in context without an HTTP request" warnings | |
1339 | # access_log ... logformat=detailed hasRequest logMe | |
1340 | # # log request-less transactions, instead of ignoring them | |
1341 | # access_log ... logformat=brief !hasRequest | |
1342 | # | |
1343 | # Multiple components are not supported for one "acl" rule, but | |
1344 | # can be specified (and are ORed) using multiple same-name rules: | |
1345 | # | |
1346 | # # OK, this strange logging daemon needs request or response, | |
1347 | # # but can work without either a request or a response: | |
1348 | # acl hasWhatMyLoggingDaemonNeeds has request | |
1349 | # acl hasWhatMyLoggingDaemonNeeds has response | |
1350 | ||
1351 | IF USE_OPENSSL | |
1352 | acl aclname ssl_error errorname | |
1353 | # match against SSL certificate validation error [fast] | |
1354 | # | |
1355 | # For valid error names see in @DEFAULT_ERROR_DIR@/templates/error-details.txt | |
1356 | # template file. | |
1357 | # | |
1358 | # The following can be used as shortcuts for certificate properties: | |
1359 | # [ssl::]certHasExpired: the "not after" field is in the past | |
1360 | # [ssl::]certNotYetValid: the "not before" field is in the future | |
1361 | # [ssl::]certUntrusted: The certificate issuer is not to be trusted. | |
1362 | # [ssl::]certSelfSigned: The certificate is self signed. | |
1363 | # [ssl::]certDomainMismatch: The certificate CN domain does not | |
1364 | # match the name the name of the host we are connecting to. | |
1365 | # | |
1366 | # The ssl::certHasExpired, ssl::certNotYetValid, ssl::certDomainMismatch, | |
1367 | # ssl::certUntrusted, and ssl::certSelfSigned can also be used as | |
1368 | # predefined ACLs, just like the 'all' ACL. | |
1369 | # | |
1370 | # NOTE: The ssl_error ACL is only supported with sslproxy_cert_error, | |
1371 | # sslproxy_cert_sign, and sslproxy_cert_adapt options. | |
1372 | ||
1373 | acl aclname server_cert_fingerprint [-sha1] fingerprint | |
1374 | # match against server SSL certificate fingerprint [fast] | |
1375 | # | |
1376 | # The fingerprint is the digest of the DER encoded version | |
1377 | # of the whole certificate. The user should use the form: XX:XX:... | |
1378 | # Optional argument specifies the digest algorithm to use. | |
1379 | # The SHA1 digest algorithm is the default and is currently | |
1380 | # the only algorithm supported (-sha1). | |
1381 | ||
1382 | acl aclname at_step step | |
1383 | # match against the current step during ssl_bump evaluation [fast] | |
1384 | # Never matches and should not be used outside the ssl_bump context. | |
1385 | # | |
1386 | # At each SslBump step, Squid evaluates ssl_bump directives to find | |
1387 | # the next bumping action (e.g., peek or splice). Valid SslBump step | |
1388 | # values and the corresponding ssl_bump evaluation moments are: | |
1389 | # SslBump1: After getting TCP-level and HTTP CONNECT info. | |
1390 | # SslBump2: After getting SSL Client Hello info. | |
1391 | # SslBump3: After getting SSL Server Hello info. | |
1392 | ||
1393 | acl aclname ssl::server_name [option] .foo.com ... | |
1394 | # matches server name obtained from various sources [fast] | |
1395 | # | |
1396 | # The ACL computes server name(s) using such information sources as | |
1397 | # CONNECT request URI, TLS client SNI, and TLS server certificate | |
1398 | # subject (CN and SubjectAltName). The computed server name(s) usually | |
1399 | # change with each SslBump step, as more info becomes available: | |
1400 | # * SNI is used as the server name instead of the request URI, | |
1401 | # * subject name(s) from the server certificate (CN and | |
1402 | # SubjectAltName) are used as the server names instead of SNI. | |
1403 | # | |
1404 | # When the ACL computes multiple server names, matching any single | |
1405 | # computed name is sufficient for the ACL to match. | |
1406 | # | |
1407 | # The "none" name can be used to match transactions where the ACL | |
1408 | # could not compute the server name using any information source | |
1409 | # that was both available and allowed to be used by the ACL options at | |
1410 | # the ACL evaluation time. | |
1411 | # | |
1412 | # Unlike dstdomain, this ACL does not perform DNS lookups. | |
1413 | # | |
1414 | # An ACL option below may be used to restrict what information | |
1415 | # sources are used to extract the server names from: | |
1416 | # | |
1417 | # --client-requested | |
1418 | # The server name is SNI regardless of what the server says. | |
1419 | # --server-provided | |
1420 | # The server name(s) are the certificate subject name(s), regardless | |
1421 | # of what the client has requested. If the server certificate is | |
1422 | # unavailable, then the name is "none". | |
1423 | # --consensus | |
1424 | # The server name is either SNI (if SNI matches at least one of the | |
1425 | # certificate subject names) or "none" (otherwise). When the server | |
1426 | # certificate is unavailable, the consensus server name is SNI. | |
1427 | # | |
1428 | # Combining multiple options in one ACL is a fatal configuration | |
1429 | # error. | |
1430 | # | |
1431 | # For all options: If no SNI is available, then the CONNECT request | |
1432 | # target (a.k.a. URI) is used instead of SNI (for an intercepted | |
1433 | # connection, this target is the destination IP address). | |
1434 | ||
1435 | acl aclname ssl::server_name_regex [-i] \.foo\.com ... | |
1436 | # regex matches server name obtained from various sources [fast] | |
1437 | ||
1438 | acl aclname connections_encrypted | |
1439 | # matches transactions with all HTTP messages received over TLS | |
1440 | # transport connections. [fast] | |
1441 | # | |
1442 | # The master transaction deals with HTTP messages received from | |
1443 | # various sources. All sources used by the master transaction in the | |
1444 | # past are considered by the ACL. The following rules define whether | |
1445 | # a given message source taints the entire master transaction, | |
1446 | # resulting in ACL mismatches: | |
1447 | # | |
1448 | # * The HTTP client transport connection is not TLS. | |
1449 | # * An adaptation service connection-encryption flag is off. | |
1450 | # * The peer or origin server transport connection is not TLS. | |
1451 | # | |
1452 | # Caching currently does not affect these rules. This cache ignorance | |
1453 | # implies that only the current HTTP client transport and REQMOD | |
1454 | # services status determine whether this ACL matches a from-cache | |
1455 | # transaction. The source of the cached response does not have any | |
1456 | # effect on future transaction that use the cached response without | |
1457 | # revalidation. This may change. | |
1458 | # | |
1459 | # DNS, ICP, and HTCP exchanges during the master transaction do not | |
1460 | # affect these rules. | |
1461 | ENDIF | |
1462 | acl aclname any-of acl1 acl2 ... | |
1463 | # match any one of the acls [fast or slow] | |
1464 | # The first matching ACL stops further ACL evaluation. | |
1465 | # | |
1466 | # ACLs from multiple any-of lines with the same name are ORed. | |
1467 | # For example, A = (a1 or a2) or (a3 or a4) can be written as | |
1468 | # acl A any-of a1 a2 | |
1469 | # acl A any-of a3 a4 | |
1470 | # | |
1471 | # This group ACL is fast if all evaluated ACLs in the group are fast | |
1472 | # and slow otherwise. | |
1473 | ||
1474 | acl aclname all-of acl1 acl2 ... | |
1475 | # match all of the acls [fast or slow] | |
1476 | # The first mismatching ACL stops further ACL evaluation. | |
1477 | # | |
1478 | # ACLs from multiple all-of lines with the same name are ORed. | |
1479 | # For example, B = (b1 and b2) or (b3 and b4) can be written as | |
1480 | # acl B all-of b1 b2 | |
1481 | # acl B all-of b3 b4 | |
1482 | # | |
1483 | # This group ACL is fast if all evaluated ACLs in the group are fast | |
1484 | # and slow otherwise. | |
1485 | ||
1486 | Examples: | |
1487 | acl macaddress arp 09:00:2b:23:45:67 | |
1488 | acl myexample dst_as 1241 | |
1489 | acl password proxy_auth REQUIRED | |
1490 | acl fileupload req_mime_type -i ^multipart/form-data$ | |
1491 | acl javascript rep_mime_type -i ^application/x-javascript$ | |
1492 | ||
1493 | NOCOMMENT_START | |
1494 | # | |
1495 | # Recommended minimum configuration: | |
1496 | # | |
1497 | ||
1498 | # Example rule allowing access from your local networks. | |
1499 | # Adapt to list your (internal) IP networks from where browsing | |
1500 | # should be allowed | |
1501 | acl localnet src 0.0.0.1-0.255.255.255 # RFC 1122 "this" network (LAN) | |
1502 | acl localnet src 10.0.0.0/8 # RFC 1918 local private network (LAN) | |
1503 | acl localnet src 100.64.0.0/10 # RFC 6598 shared address space (CGN) | |
1504 | acl localhet src 169.254.0.0/16 # RFC 3927 link-local (directly plugged) machines | |
1505 | acl localnet src 172.16.0.0/12 # RFC 1918 local private network (LAN) | |
1506 | acl localnet src 192.168.0.0/16 # RFC 1918 local private network (LAN) | |
1507 | acl localnet src fc00::/7 # RFC 4193 local private network range | |
1508 | acl localnet src fe80::/10 # RFC 4291 link-local (directly plugged) machines | |
1509 | ||
1510 | acl SSL_ports port 443 | |
1511 | acl Safe_ports port 80 # http | |
1512 | acl Safe_ports port 21 # ftp | |
1513 | acl Safe_ports port 443 # https | |
1514 | acl Safe_ports port 70 # gopher | |
1515 | acl Safe_ports port 210 # wais | |
1516 | acl Safe_ports port 1025-65535 # unregistered ports | |
1517 | acl Safe_ports port 280 # http-mgmt | |
1518 | acl Safe_ports port 488 # gss-http | |
1519 | acl Safe_ports port 591 # filemaker | |
1520 | acl Safe_ports port 777 # multiling http | |
1521 | NOCOMMENT_END | |
1522 | DOC_END | |
1523 | ||
1524 | NAME: proxy_protocol_access | |
1525 | TYPE: acl_access | |
1526 | LOC: Config.accessList.proxyProtocol | |
1527 | DEFAULT: none | |
1528 | DEFAULT_DOC: all TCP connections to ports with require-proxy-header will be denied | |
1529 | DOC_START | |
1530 | Determine which client proxies can be trusted to provide correct | |
1531 | information regarding real client IP address using PROXY protocol. | |
1532 | ||
1533 | Requests may pass through a chain of several other proxies | |
1534 | before reaching us. The original source details may by sent in: | |
1535 | * HTTP message Forwarded header, or | |
1536 | * HTTP message X-Forwarded-For header, or | |
1537 | * PROXY protocol connection header. | |
1538 | ||
1539 | This directive is solely for validating new PROXY protocol | |
1540 | connections received from a port flagged with require-proxy-header. | |
1541 | It is checked only once after TCP connection setup. | |
1542 | ||
1543 | A deny match results in TCP connection closure. | |
1544 | ||
1545 | An allow match is required for Squid to permit the corresponding | |
1546 | TCP connection, before Squid even looks for HTTP request headers. | |
1547 | If there is an allow match, Squid starts using PROXY header information | |
1548 | to determine the source address of the connection for all future ACL | |
1549 | checks, logging, etc. | |
1550 | ||
1551 | SECURITY CONSIDERATIONS: | |
1552 | ||
1553 | Any host from which we accept client IP details can place | |
1554 | incorrect information in the relevant header, and Squid | |
1555 | will use the incorrect information as if it were the | |
1556 | source address of the request. This may enable remote | |
1557 | hosts to bypass any access control restrictions that are | |
1558 | based on the client's source addresses. | |
1559 | ||
1560 | This clause only supports fast acl types. | |
1561 | See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details. | |
1562 | DOC_END | |
1563 | ||
1564 | NAME: follow_x_forwarded_for | |
1565 | TYPE: acl_access | |
1566 | IFDEF: FOLLOW_X_FORWARDED_FOR | |
1567 | LOC: Config.accessList.followXFF | |
1568 | DEFAULT_IF_NONE: deny all | |
1569 | DEFAULT_DOC: X-Forwarded-For header will be ignored. | |
1570 | DOC_START | |
1571 | Determine which client proxies can be trusted to provide correct | |
1572 | information regarding real client IP address. | |
1573 | ||
1574 | Requests may pass through a chain of several other proxies | |
1575 | before reaching us. The original source details may by sent in: | |
1576 | * HTTP message Forwarded header, or | |
1577 | * HTTP message X-Forwarded-For header, or | |
1578 | * PROXY protocol connection header. | |
1579 | ||
1580 | PROXY protocol connections are controlled by the proxy_protocol_access | |
1581 | directive which is checked before this. | |
1582 | ||
1583 | If a request reaches us from a source that is allowed by this | |
1584 | directive, then we trust the information it provides regarding | |
1585 | the IP of the client it received from (if any). | |
1586 | ||
1587 | For the purpose of ACLs used in this directive the src ACL type always | |
1588 | matches the address we are testing and srcdomain matches its rDNS. | |
1589 | ||
1590 | On each HTTP request Squid checks for X-Forwarded-For header fields. | |
1591 | If found the header values are iterated in reverse order and an allow | |
1592 | match is required for Squid to continue on to the next value. | |
1593 | The verification ends when a value receives a deny match, cannot be | |
1594 | tested, or there are no more values to test. | |
1595 | NOTE: Squid does not yet follow the Forwarded HTTP header. | |
1596 | ||
1597 | The end result of this process is an IP address that we will | |
1598 | refer to as the indirect client address. This address may | |
1599 | be treated as the client address for access control, ICAP, delay | |
1600 | pools and logging, depending on the acl_uses_indirect_client, | |
1601 | icap_uses_indirect_client, delay_pool_uses_indirect_client, | |
1602 | log_uses_indirect_client and tproxy_uses_indirect_client options. | |
1603 | ||
1604 | This clause only supports fast acl types. | |
1605 | See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details. | |
1606 | ||
1607 | SECURITY CONSIDERATIONS: | |
1608 | ||
1609 | Any host from which we accept client IP details can place | |
1610 | incorrect information in the relevant header, and Squid | |
1611 | will use the incorrect information as if it were the | |
1612 | source address of the request. This may enable remote | |
1613 | hosts to bypass any access control restrictions that are | |
1614 | based on the client's source addresses. | |
1615 | ||
1616 | For example: | |
1617 | ||
1618 | acl localhost src 127.0.0.1 | |
1619 | acl my_other_proxy srcdomain .proxy.example.com | |
1620 | follow_x_forwarded_for allow localhost | |
1621 | follow_x_forwarded_for allow my_other_proxy | |
1622 | DOC_END | |
1623 | ||
1624 | NAME: acl_uses_indirect_client | |
1625 | COMMENT: on|off | |
1626 | TYPE: onoff | |
1627 | IFDEF: FOLLOW_X_FORWARDED_FOR | |
1628 | DEFAULT: on | |
1629 | LOC: Config.onoff.acl_uses_indirect_client | |
1630 | DOC_START | |
1631 | Controls whether the indirect client address | |
1632 | (see follow_x_forwarded_for) is used instead of the | |
1633 | direct client address in acl matching. | |
1634 | ||
1635 | NOTE: maxconn ACL considers direct TCP links and indirect | |
1636 | clients will always have zero. So no match. | |
1637 | DOC_END | |
1638 | ||
1639 | NAME: delay_pool_uses_indirect_client | |
1640 | COMMENT: on|off | |
1641 | TYPE: onoff | |
1642 | IFDEF: FOLLOW_X_FORWARDED_FOR&&USE_DELAY_POOLS | |
1643 | DEFAULT: on | |
1644 | LOC: Config.onoff.delay_pool_uses_indirect_client | |
1645 | DOC_START | |
1646 | Controls whether the indirect client address | |
1647 | (see follow_x_forwarded_for) is used instead of the | |
1648 | direct client address in delay pools. | |
1649 | DOC_END | |
1650 | ||
1651 | NAME: log_uses_indirect_client | |
1652 | COMMENT: on|off | |
1653 | TYPE: onoff | |
1654 | IFDEF: FOLLOW_X_FORWARDED_FOR | |
1655 | DEFAULT: on | |
1656 | LOC: Config.onoff.log_uses_indirect_client | |
1657 | DOC_START | |
1658 | Controls whether the indirect client address | |
1659 | (see follow_x_forwarded_for) is used instead of the | |
1660 | direct client address in the access log. | |
1661 | DOC_END | |
1662 | ||
1663 | NAME: tproxy_uses_indirect_client | |
1664 | COMMENT: on|off | |
1665 | TYPE: onoff | |
1666 | IFDEF: FOLLOW_X_FORWARDED_FOR&&LINUX_NETFILTER | |
1667 | DEFAULT: off | |
1668 | LOC: Config.onoff.tproxy_uses_indirect_client | |
1669 | DOC_START | |
1670 | Controls whether the indirect client address | |
1671 | (see follow_x_forwarded_for) is used instead of the | |
1672 | direct client address when spoofing the outgoing client. | |
1673 | ||
1674 | This has no effect on requests arriving in non-tproxy | |
1675 | mode ports. | |
1676 | ||
1677 | SECURITY WARNING: Usage of this option is dangerous | |
1678 | and should not be used trivially. Correct configuration | |
1679 | of follow_x_forwarded_for with a limited set of trusted | |
1680 | sources is required to prevent abuse of your proxy. | |
1681 | DOC_END | |
1682 | ||
1683 | NAME: spoof_client_ip | |
1684 | TYPE: acl_access | |
1685 | LOC: Config.accessList.spoof_client_ip | |
1686 | DEFAULT: none | |
1687 | DEFAULT_DOC: Allow spoofing on all TPROXY traffic. | |
1688 | DOC_START | |
1689 | Control client IP address spoofing of TPROXY traffic based on | |
1690 | defined access lists. | |
1691 | ||
1692 | spoof_client_ip allow|deny [!]aclname ... | |
1693 | ||
1694 | If there are no "spoof_client_ip" lines present, the default | |
1695 | is to "allow" spoofing of any suitable request. | |
1696 | ||
1697 | Note that the cache_peer "no-tproxy" option overrides this ACL. | |
1698 | ||
1699 | This clause supports fast acl types. | |
1700 | See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details. | |
1701 | DOC_END | |
1702 | ||
1703 | NAME: http_access | |
1704 | TYPE: acl_access | |
1705 | LOC: Config.accessList.http | |
1706 | DEFAULT_IF_NONE: deny all | |
1707 | DEFAULT_DOC: Deny, unless rules exist in squid.conf. | |
1708 | DOC_START | |
1709 | Allowing or Denying access based on defined access lists | |
1710 | ||
1711 | To allow or deny a message received on an HTTP, HTTPS, or FTP port: | |
1712 | http_access allow|deny [!]aclname ... | |
1713 | ||
1714 | NOTE on default values: | |
1715 | ||
1716 | If there are no "access" lines present, the default is to deny | |
1717 | the request. | |
1718 | ||
1719 | If none of the "access" lines cause a match, the default is the | |
1720 | opposite of the last line in the list. If the last line was | |
1721 | deny, the default is allow. Conversely, if the last line | |
1722 | is allow, the default will be deny. For these reasons, it is a | |
1723 | good idea to have an "deny all" entry at the end of your access | |
1724 | lists to avoid potential confusion. | |
1725 | ||
1726 | This clause supports both fast and slow acl types. | |
1727 | See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details. | |
1728 | ||
1729 | NOCOMMENT_START | |
1730 | ||
1731 | # | |
1732 | # Recommended minimum Access Permission configuration: | |
1733 | # | |
1734 | # Deny requests to certain unsafe ports | |
1735 | http_access deny !Safe_ports | |
1736 | ||
1737 | # Deny CONNECT to other than secure SSL ports | |
1738 | http_access deny CONNECT !SSL_ports | |
1739 | ||
1740 | # Only allow cachemgr access from localhost | |
1741 | http_access allow localhost manager | |
1742 | http_access deny manager | |
1743 | ||
1744 | # We strongly recommend the following be uncommented to protect innocent | |
1745 | # web applications running on the proxy server who think the only | |
1746 | # one who can access services on "localhost" is a local user | |
1747 | #http_access deny to_localhost | |
1748 | ||
1749 | # | |
1750 | # INSERT YOUR OWN RULE(S) HERE TO ALLOW ACCESS FROM YOUR CLIENTS | |
1751 | # | |
1752 | ||
1753 | # Example rule allowing access from your local networks. | |
1754 | # Adapt localnet in the ACL section to list your (internal) IP networks | |
1755 | # from where browsing should be allowed | |
1756 | http_access allow localnet | |
1757 | http_access allow localhost | |
1758 | ||
1759 | # And finally deny all other access to this proxy | |
1760 | http_access deny all | |
1761 | NOCOMMENT_END | |
1762 | DOC_END | |
1763 | ||
1764 | NAME: adapted_http_access http_access2 | |
1765 | TYPE: acl_access | |
1766 | LOC: Config.accessList.adapted_http | |
1767 | DEFAULT: none | |
1768 | DEFAULT_DOC: Allow, unless rules exist in squid.conf. | |
1769 | DOC_START | |
1770 | Allowing or Denying access based on defined access lists | |
1771 | ||
1772 | Essentially identical to http_access, but runs after redirectors | |
1773 | and ICAP/eCAP adaptation. Allowing access control based on their | |
1774 | output. | |
1775 | ||
1776 | If not set then only http_access is used. | |
1777 | DOC_END | |
1778 | ||
1779 | NAME: http_reply_access | |
1780 | TYPE: acl_access | |
1781 | LOC: Config.accessList.reply | |
1782 | DEFAULT: none | |
1783 | DEFAULT_DOC: Allow, unless rules exist in squid.conf. | |
1784 | DOC_START | |
1785 | Allow replies to client requests. This is complementary to http_access. | |
1786 | ||
1787 | http_reply_access allow|deny [!] aclname ... | |
1788 | ||
1789 | NOTE: if there are no access lines present, the default is to allow | |
1790 | all replies. | |
1791 | ||
1792 | If none of the access lines cause a match the opposite of the | |
1793 | last line will apply. Thus it is good practice to end the rules | |
1794 | with an "allow all" or "deny all" entry. | |
1795 | ||
1796 | This clause supports both fast and slow acl types. | |
1797 | See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details. | |
1798 | DOC_END | |
1799 | ||
1800 | NAME: icp_access | |
1801 | TYPE: acl_access | |
1802 | LOC: Config.accessList.icp | |
1803 | DEFAULT: none | |
1804 | DEFAULT_DOC: Deny, unless rules exist in squid.conf. | |
1805 | DOC_START | |
1806 | Allowing or Denying access to the ICP port based on defined | |
1807 | access lists | |
1808 | ||
1809 | icp_access allow|deny [!]aclname ... | |
1810 | ||
1811 | NOTE: The default if no icp_access lines are present is to | |
1812 | deny all traffic. This default may cause problems with peers | |
1813 | using ICP. | |
1814 | ||
1815 | This clause only supports fast acl types. | |
1816 | See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details. | |
1817 | ||
1818 | # Allow ICP queries from local networks only | |
1819 | #icp_access allow localnet | |
1820 | #icp_access deny all | |
1821 | DOC_END | |
1822 | ||
1823 | NAME: htcp_access | |
1824 | IFDEF: USE_HTCP | |
1825 | TYPE: acl_access | |
1826 | LOC: Config.accessList.htcp | |
1827 | DEFAULT: none | |
1828 | DEFAULT_DOC: Deny, unless rules exist in squid.conf. | |
1829 | DOC_START | |
1830 | Allowing or Denying access to the HTCP port based on defined | |
1831 | access lists | |
1832 | ||
1833 | htcp_access allow|deny [!]aclname ... | |
1834 | ||
1835 | See also htcp_clr_access for details on access control for | |
1836 | cache purge (CLR) HTCP messages. | |
1837 | ||
1838 | NOTE: The default if no htcp_access lines are present is to | |
1839 | deny all traffic. This default may cause problems with peers | |
1840 | using the htcp option. | |
1841 | ||
1842 | This clause only supports fast acl types. | |
1843 | See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details. | |
1844 | ||
1845 | # Allow HTCP queries from local networks only | |
1846 | #htcp_access allow localnet | |
1847 | #htcp_access deny all | |
1848 | DOC_END | |
1849 | ||
1850 | NAME: htcp_clr_access | |
1851 | IFDEF: USE_HTCP | |
1852 | TYPE: acl_access | |
1853 | LOC: Config.accessList.htcp_clr | |
1854 | DEFAULT: none | |
1855 | DEFAULT_DOC: Deny, unless rules exist in squid.conf. | |
1856 | DOC_START | |
1857 | Allowing or Denying access to purge content using HTCP based | |
1858 | on defined access lists. | |
1859 | See htcp_access for details on general HTCP access control. | |
1860 | ||
1861 | htcp_clr_access allow|deny [!]aclname ... | |
1862 | ||
1863 | This clause only supports fast acl types. | |
1864 | See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details. | |
1865 | ||
1866 | # Allow HTCP CLR requests from trusted peers | |
1867 | acl htcp_clr_peer src 192.0.2.2 2001:DB8::2 | |
1868 | htcp_clr_access allow htcp_clr_peer | |
1869 | htcp_clr_access deny all | |
1870 | DOC_END | |
1871 | ||
1872 | NAME: miss_access | |
1873 | TYPE: acl_access | |
1874 | LOC: Config.accessList.miss | |
1875 | DEFAULT: none | |
1876 | DEFAULT_DOC: Allow, unless rules exist in squid.conf. | |
1877 | DOC_START | |
1878 | Determines whether network access is permitted when satisfying a request. | |
1879 | ||
1880 | For example; | |
1881 | to force your neighbors to use you as a sibling instead of | |
1882 | a parent. | |
1883 | ||
1884 | acl localclients src 192.0.2.0/24 2001:DB8::a:0/64 | |
1885 | miss_access deny !localclients | |
1886 | miss_access allow all | |
1887 | ||
1888 | This means only your local clients are allowed to fetch relayed/MISS | |
1889 | replies from the network and all other clients can only fetch cached | |
1890 | objects (HITs). | |
1891 | ||
1892 | The default for this setting allows all clients who passed the | |
1893 | http_access rules to relay via this proxy. | |
1894 | ||
1895 | This clause only supports fast acl types. | |
1896 | See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details. | |
1897 | DOC_END | |
1898 | ||
1899 | NAME: ident_lookup_access | |
1900 | TYPE: acl_access | |
1901 | IFDEF: USE_IDENT | |
1902 | DEFAULT: none | |
1903 | DEFAULT_DOC: Unless rules exist in squid.conf, IDENT is not fetched. | |
1904 | LOC: Ident::TheConfig.identLookup | |
1905 | DOC_START | |
1906 | A list of ACL elements which, if matched, cause an ident | |
1907 | (RFC 931) lookup to be performed for this request. For | |
1908 | example, you might choose to always perform ident lookups | |
1909 | for your main multi-user Unix boxes, but not for your Macs | |
1910 | and PCs. By default, ident lookups are not performed for | |
1911 | any requests. | |
1912 | ||
1913 | To enable ident lookups for specific client addresses, you | |
1914 | can follow this example: | |
1915 | ||
1916 | acl ident_aware_hosts src 198.168.1.0/24 | |
1917 | ident_lookup_access allow ident_aware_hosts | |
1918 | ident_lookup_access deny all | |
1919 | ||
1920 | Only src type ACL checks are fully supported. A srcdomain | |
1921 | ACL might work at times, but it will not always provide | |
1922 | the correct result. | |
1923 | ||
1924 | This clause only supports fast acl types. | |
1925 | See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details. | |
1926 | DOC_END | |
1927 | ||
1928 | NAME: reply_body_max_size | |
1929 | COMMENT: size [acl acl...] | |
1930 | TYPE: acl_b_size_t | |
1931 | DEFAULT: none | |
1932 | DEFAULT_DOC: No limit is applied. | |
1933 | LOC: Config.ReplyBodySize | |
1934 | DOC_START | |
1935 | This option specifies the maximum size of a reply body. It can be | |
1936 | used to prevent users from downloading very large files, such as | |
1937 | MP3's and movies. When the reply headers are received, the | |
1938 | reply_body_max_size lines are processed, and the first line where | |
1939 | all (if any) listed ACLs are true is used as the maximum body size | |
1940 | for this reply. | |
1941 | ||
1942 | This size is checked twice. First when we get the reply headers, | |
1943 | we check the content-length value. If the content length value exists | |
1944 | and is larger than the allowed size, the request is denied and the | |
1945 | user receives an error message that says "the request or reply | |
1946 | is too large." If there is no content-length, and the reply | |
1947 | size exceeds this limit, the client's connection is just closed | |
1948 | and they will receive a partial reply. | |
1949 | ||
1950 | WARNING: downstream caches probably can not detect a partial reply | |
1951 | if there is no content-length header, so they will cache | |
1952 | partial responses and give them out as hits. You should NOT | |
1953 | use this option if you have downstream caches. | |
1954 | ||
1955 | WARNING: A maximum size smaller than the size of squid's error messages | |
1956 | will cause an infinite loop and crash squid. Ensure that the smallest | |
1957 | non-zero value you use is greater that the maximum header size plus | |
1958 | the size of your largest error page. | |
1959 | ||
1960 | If you set this parameter none (the default), there will be | |
1961 | no limit imposed. | |
1962 | ||
1963 | Configuration Format is: | |
1964 | reply_body_max_size SIZE UNITS [acl ...] | |
1965 | ie. | |
1966 | reply_body_max_size 10 MB | |
1967 | ||
1968 | DOC_END | |
1969 | ||
1970 | NAME: on_unsupported_protocol | |
1971 | TYPE: on_unsupported_protocol | |
1972 | LOC: Config.accessList.on_unsupported_protocol | |
1973 | DEFAULT: none | |
1974 | DEFAULT_DOC: Respond with an error message to unidentifiable traffic | |
1975 | DOC_START | |
1976 | Determines Squid behavior when encountering strange requests at the | |
1977 | beginning of an accepted TCP connection or the beginning of a bumped | |
1978 | CONNECT tunnel. Controlling Squid reaction to unexpected traffic is | |
1979 | especially useful in interception environments where Squid is likely | |
1980 | to see connections for unsupported protocols that Squid should either | |
1981 | terminate or tunnel at TCP level. | |
1982 | ||
1983 | on_unsupported_protocol <action> [!]acl ... | |
1984 | ||
1985 | The first matching action wins. Only fast ACLs are supported. | |
1986 | ||
1987 | Supported actions are: | |
1988 | ||
1989 | tunnel: Establish a TCP connection with the intended server and | |
1990 | blindly shovel TCP packets between the client and server. | |
1991 | ||
1992 | respond: Respond with an error message, using the transfer protocol | |
1993 | for the Squid port that received the request (e.g., HTTP | |
1994 | for connections intercepted at the http_port). This is the | |
1995 | default. | |
1996 | ||
1997 | Squid expects the following traffic patterns: | |
1998 | ||
1999 | http_port: a plain HTTP request | |
2000 | https_port: SSL/TLS handshake followed by an [encrypted] HTTP request | |
2001 | ftp_port: a plain FTP command (no on_unsupported_protocol support yet!) | |
2002 | CONNECT tunnel on http_port: same as https_port | |
2003 | CONNECT tunnel on https_port: same as https_port | |
2004 | ||
2005 | Currently, this directive has effect on intercepted connections and | |
2006 | bumped tunnels only. Other cases are not supported because Squid | |
2007 | cannot know the intended destination of other traffic. | |
2008 | ||
2009 | For example: | |
2010 | # define what Squid errors indicate receiving non-HTTP traffic: | |
2011 | acl foreignProtocol squid_error ERR_PROTOCOL_UNKNOWN ERR_TOO_BIG | |
2012 | # define what Squid errors indicate receiving nothing: | |
2013 | acl serverTalksFirstProtocol squid_error ERR_REQUEST_START_TIMEOUT | |
2014 | # tunnel everything that does not look like HTTP: | |
2015 | on_unsupported_protocol tunnel foreignProtocol | |
2016 | # tunnel if we think the client waits for the server to talk first: | |
2017 | on_unsupported_protocol tunnel serverTalksFirstProtocol | |
2018 | # in all other error cases, just send an HTTP "error page" response: | |
2019 | on_unsupported_protocol respond all | |
2020 | ||
2021 | See also: squid_error ACL | |
2022 | DOC_END | |
2023 | ||
2024 | NAME: auth_schemes | |
2025 | TYPE: AuthSchemes | |
2026 | IFDEF: USE_AUTH | |
2027 | LOC: Auth::TheConfig.schemeAccess | |
2028 | DEFAULT: none | |
2029 | DEFAULT_DOC: use all auth_param schemes in their configuration order | |
2030 | DOC_START | |
2031 | Use this directive to customize authentication schemes presence and | |
2032 | order in Squid's Unauthorized and Authentication Required responses. | |
2033 | ||
2034 | auth_schemes scheme1,scheme2,... [!]aclname ... | |
2035 | ||
2036 | where schemeN is the name of one of the authentication schemes | |
2037 | configured using auth_param directives. At least one scheme name is | |
2038 | required. Multiple scheme names are separated by commas. Either | |
2039 | avoid whitespace or quote the entire schemes list. | |
2040 | ||
2041 | A special "ALL" scheme name expands to all auth_param-configured | |
2042 | schemes in their configuration order. This directive cannot be used | |
2043 | to configure Squid to offer no authentication schemes at all. | |
2044 | ||
2045 | The first matching auth_schemes rule determines the schemes order | |
2046 | for the current Authentication Required transaction. Note that the | |
2047 | future response is not yet available during auth_schemes evaluation. | |
2048 | ||
2049 | If this directive is not used or none of its rules match, then Squid | |
2050 | responds with all configured authentication schemes in the order of | |
2051 | auth_param directives in the configuration file. | |
2052 | ||
2053 | This directive does not determine when authentication is used or | |
2054 | how each authentication scheme authenticates clients. | |
2055 | ||
2056 | The following example sends basic and negotiate authentication | |
2057 | schemes, in that order, when requesting authentication of HTTP | |
2058 | requests matching the isIE ACL (not shown) while sending all | |
2059 | auth_param schemes in their configuration order to other clients: | |
2060 | ||
2061 | auth_schemes basic,negotiate isIE | |
2062 | auth_schemes ALL all # explicit default | |
2063 | ||
2064 | This directive supports fast ACLs only. | |
2065 | ||
2066 | See also: auth_param. | |
2067 | DOC_END | |
2068 | ||
2069 | COMMENT_START | |
2070 | NETWORK OPTIONS | |
2071 | ----------------------------------------------------------------------------- | |
2072 | COMMENT_END | |
2073 | ||
2074 | NAME: http_port ascii_port | |
2075 | TYPE: PortCfg | |
2076 | DEFAULT: none | |
2077 | LOC: HttpPortList | |
2078 | DOC_START | |
2079 | Usage: port [mode] [options] | |
2080 | hostname:port [mode] [options] | |
2081 | 1.2.3.4:port [mode] [options] | |
2082 | ||
2083 | The socket addresses where Squid will listen for HTTP client | |
2084 | requests. You may specify multiple socket addresses. | |
2085 | There are three forms: port alone, hostname with port, and | |
2086 | IP address with port. If you specify a hostname or IP | |
2087 | address, Squid binds the socket to that specific | |
2088 | address. Most likely, you do not need to bind to a specific | |
2089 | address, so you can use the port number alone. | |
2090 | ||
2091 | If you are running Squid in accelerator mode, you | |
2092 | probably want to listen on port 80 also, or instead. | |
2093 | ||
2094 | The -a command line option may be used to specify additional | |
2095 | port(s) where Squid listens for proxy request. Such ports will | |
2096 | be plain proxy ports with no options. | |
2097 | ||
2098 | You may specify multiple socket addresses on multiple lines. | |
2099 | ||
2100 | Modes: | |
2101 | ||
2102 | intercept Support for IP-Layer NAT interception delivering | |
2103 | traffic to this Squid port. | |
2104 | NP: disables authentication on the port. | |
2105 | ||
2106 | tproxy Support Linux TPROXY (or BSD divert-to) with spoofing | |
2107 | of outgoing connections using the client IP address. | |
2108 | NP: disables authentication on the port. | |
2109 | ||
2110 | accel Accelerator / reverse proxy mode | |
2111 | ||
2112 | ssl-bump For each CONNECT request allowed by ssl_bump ACLs, | |
2113 | establish secure connection with the client and with | |
2114 | the server, decrypt HTTPS messages as they pass through | |
2115 | Squid, and treat them as unencrypted HTTP messages, | |
2116 | becoming the man-in-the-middle. | |
2117 | ||
2118 | The ssl_bump option is required to fully enable | |
2119 | bumping of CONNECT requests. | |
2120 | ||
2121 | Omitting the mode flag causes default forward proxy mode to be used. | |
2122 | ||
2123 | ||
2124 | Accelerator Mode Options: | |
2125 | ||
2126 | defaultsite=domainname | |
2127 | What to use for the Host: header if it is not present | |
2128 | in a request. Determines what site (not origin server) | |
2129 | accelerators should consider the default. | |
2130 | ||
2131 | no-vhost Disable using HTTP/1.1 Host header for virtual domain support. | |
2132 | ||
2133 | protocol= Protocol to reconstruct accelerated and intercepted | |
2134 | requests with. Defaults to HTTP/1.1 for http_port and | |
2135 | HTTPS/1.1 for https_port. | |
2136 | When an unsupported value is configured Squid will | |
2137 | produce a FATAL error. | |
2138 | Values: HTTP or HTTP/1.1, HTTPS or HTTPS/1.1 | |
2139 | ||
2140 | vport Virtual host port support. Using the http_port number | |
2141 | instead of the port passed on Host: headers. | |
2142 | ||
2143 | vport=NN Virtual host port support. Using the specified port | |
2144 | number instead of the port passed on Host: headers. | |
2145 | ||
2146 | act-as-origin | |
2147 | Act as if this Squid is the origin server. | |
2148 | This currently means generate new Date: and Expires: | |
2149 | headers on HIT instead of adding Age:. | |
2150 | ||
2151 | ignore-cc Ignore request Cache-Control headers. | |
2152 | ||
2153 | WARNING: This option violates HTTP specifications if | |
2154 | used in non-accelerator setups. | |
2155 | ||
2156 | allow-direct Allow direct forwarding in accelerator mode. Normally | |
2157 | accelerated requests are denied direct forwarding as if | |
2158 | never_direct was used. | |
2159 | ||
2160 | WARNING: this option opens accelerator mode to security | |
2161 | vulnerabilities usually only affecting in interception | |
2162 | mode. Make sure to protect forwarding with suitable | |
2163 | http_access rules when using this. | |
2164 | ||
2165 | ||
2166 | SSL Bump Mode Options: | |
2167 | In addition to these options ssl-bump requires TLS/SSL options. | |
2168 | ||
2169 | generate-host-certificates[=<on|off>] | |
2170 | Dynamically create SSL server certificates for the | |
2171 | destination hosts of bumped CONNECT requests.When | |
2172 | enabled, the cert and key options are used to sign | |
2173 | generated certificates. Otherwise generated | |
2174 | certificate will be selfsigned. | |
2175 | If there is a CA certificate lifetime of the generated | |
2176 | certificate equals lifetime of the CA certificate. If | |
2177 | generated certificate is selfsigned lifetime is three | |
2178 | years. | |
2179 | This option is enabled by default when ssl-bump is used. | |
2180 | See the ssl-bump option above for more information. | |
2181 | ||
2182 | dynamic_cert_mem_cache_size=SIZE | |
2183 | Approximate total RAM size spent on cached generated | |
2184 | certificates. If set to zero, caching is disabled. The | |
2185 | default value is 4MB. | |
2186 | ||
2187 | TLS / SSL Options: | |
2188 | ||
2189 | cert= Path to SSL certificate (PEM format). | |
2190 | ||
2191 | key= Path to SSL private key file (PEM format) | |
2192 | if not specified, the certificate file is | |
2193 | assumed to be a combined certificate and | |
2194 | key file. | |
2195 | ||
2196 | cipher= Colon separated list of supported ciphers. | |
2197 | NOTE: some ciphers such as EDH ciphers depend on | |
2198 | additional settings. If those settings are | |
2199 | omitted the ciphers may be silently ignored | |
2200 | by the OpenSSL library. | |
2201 | ||
2202 | options= Various SSL implementation options. The most important | |
2203 | being: | |
2204 | ||
2205 | NO_SSLv3 Disallow the use of SSLv3 | |
2206 | ||
2207 | NO_TLSv1 Disallow the use of TLSv1.0 | |
2208 | ||
2209 | NO_TLSv1_1 Disallow the use of TLSv1.1 | |
2210 | ||
2211 | NO_TLSv1_2 Disallow the use of TLSv1.2 | |
2212 | ||
2213 | SINGLE_DH_USE | |
2214 | Always create a new key when using | |
2215 | temporary/ephemeral DH key exchanges | |
2216 | ||
2217 | SINGLE_ECDH_USE | |
2218 | Enable ephemeral ECDH key exchange. | |
2219 | The adopted curve should be specified | |
2220 | using the tls-dh option. | |
2221 | ||
2222 | NO_TICKET | |
2223 | Disable use of RFC5077 session tickets. | |
2224 | Some servers may have problems | |
2225 | understanding the TLS extension due | |
2226 | to ambiguous specification in RFC4507. | |
2227 | ||
2228 | ALL Enable various bug workarounds | |
2229 | suggested as "harmless" by OpenSSL | |
2230 | Be warned that this reduces SSL/TLS | |
2231 | strength to some attacks. | |
2232 | ||
2233 | See the OpenSSL SSL_CTX_set_options documentation for a | |
2234 | more complete list. | |
2235 | ||
2236 | clientca= File containing the list of CAs to use when | |
2237 | requesting a client certificate. | |
2238 | ||
2239 | tls-cafile= PEM file containing CA certificates to use when verifying | |
2240 | client certificates. If not configured clientca will be | |
2241 | used. May be repeated to load multiple files. | |
2242 | ||
2243 | capath= Directory containing additional CA certificates | |
2244 | and CRL lists to use when verifying client certificates. | |
2245 | Requires OpenSSL or LibreSSL. | |
2246 | ||
2247 | crlfile= File of additional CRL lists to use when verifying | |
2248 | the client certificate, in addition to CRLs stored in | |
2249 | the capath. Implies VERIFY_CRL flag below. | |
2250 | ||
2251 | tls-dh=[curve:]file | |
2252 | File containing DH parameters for temporary/ephemeral DH key | |
2253 | exchanges, optionally prefixed by a curve for ephemeral ECDH | |
2254 | key exchanges. | |
2255 | See OpenSSL documentation for details on how to create the | |
2256 | DH parameter file. Supported curves for ECDH can be listed | |
2257 | using the "openssl ecparam -list_curves" command. | |
2258 | WARNING: EDH and EECDH ciphers will be silently disabled if | |
2259 | this option is not set. | |
2260 | ||
2261 | sslflags= Various flags modifying the use of SSL: | |
2262 | DELAYED_AUTH | |
2263 | Don't request client certificates | |
2264 | immediately, but wait until acl processing | |
2265 | requires a certificate (not yet implemented). | |
2266 | NO_SESSION_REUSE | |
2267 | Don't allow for session reuse. Each connection | |
2268 | will result in a new SSL session. | |
2269 | VERIFY_CRL | |
2270 | Verify CRL lists when accepting client | |
2271 | certificates. | |
2272 | VERIFY_CRL_ALL | |
2273 | Verify CRL lists for all certificates in the | |
2274 | client certificate chain. | |
2275 | ||
2276 | tls-default-ca[=off] | |
2277 | Whether to use the system Trusted CAs. Default is OFF. | |
2278 | ||
2279 | tls-no-npn Do not use the TLS NPN extension to advertise HTTP/1.1. | |
2280 | ||
2281 | sslcontext= SSL session ID context identifier. | |
2282 | ||
2283 | Other Options: | |
2284 | ||
2285 | connection-auth[=on|off] | |
2286 | use connection-auth=off to tell Squid to prevent | |
2287 | forwarding Microsoft connection oriented authentication | |
2288 | (NTLM, Negotiate and Kerberos) | |
2289 | ||
2290 | disable-pmtu-discovery= | |
2291 | Control Path-MTU discovery usage: | |
2292 | off lets OS decide on what to do (default). | |
2293 | transparent disable PMTU discovery when transparent | |
2294 | support is enabled. | |
2295 | always disable always PMTU discovery. | |
2296 | ||
2297 | In many setups of transparently intercepting proxies | |
2298 | Path-MTU discovery can not work on traffic towards the | |
2299 | clients. This is the case when the intercepting device | |
2300 | does not fully track connections and fails to forward | |
2301 | ICMP must fragment messages to the cache server. If you | |
2302 | have such setup and experience that certain clients | |
2303 | sporadically hang or never complete requests set | |
2304 | disable-pmtu-discovery option to 'transparent'. | |
2305 | ||
2306 | name= Specifies a internal name for the port. Defaults to | |
2307 | the port specification (port or addr:port) | |
2308 | ||
2309 | tcpkeepalive[=idle,interval,timeout] | |
2310 | Enable TCP keepalive probes of idle connections. | |
2311 | In seconds; idle is the initial time before TCP starts | |
2312 | probing the connection, interval how often to probe, and | |
2313 | timeout the time before giving up. | |
2314 | ||
2315 | require-proxy-header | |
2316 | Require PROXY protocol version 1 or 2 connections. | |
2317 | The proxy_protocol_access is required to whitelist | |
2318 | downstream proxies which can be trusted. | |
2319 | ||
2320 | If you run Squid on a dual-homed machine with an internal | |
2321 | and an external interface we recommend you to specify the | |
2322 | internal address:port in http_port. This way Squid will only be | |
2323 | visible on the internal address. | |
2324 | ||
2325 | NOCOMMENT_START | |
2326 | ||
2327 | # Squid normally listens to port 3128 | |
2328 | http_port @DEFAULT_HTTP_PORT@ | |
2329 | NOCOMMENT_END | |
2330 | DOC_END | |
2331 | ||
2332 | NAME: https_port | |
2333 | IFDEF: USE_GNUTLS||USE_OPENSSL | |
2334 | TYPE: PortCfg | |
2335 | DEFAULT: none | |
2336 | LOC: HttpPortList | |
2337 | DOC_START | |
2338 | Usage: [ip:]port [mode] cert=certificate.pem [options] | |
2339 | ||
2340 | The socket address where Squid will listen for client requests made | |
2341 | over TLS or SSL connections. Commonly referred to as HTTPS. | |
2342 | ||
2343 | This is most useful for situations where you are running squid in | |
2344 | accelerator mode and you want to do the TLS work at the accelerator level. | |
2345 | ||
2346 | You may specify multiple socket addresses on multiple lines, | |
2347 | each with their own certificate and/or options. | |
2348 | ||
2349 | The TLS cert= option is mandatory on HTTPS ports. | |
2350 | ||
2351 | See http_port for a list of modes and options. | |
2352 | DOC_END | |
2353 | ||
2354 | NAME: ftp_port | |
2355 | TYPE: PortCfg | |
2356 | DEFAULT: none | |
2357 | LOC: FtpPortList | |
2358 | DOC_START | |
2359 | Enables Native FTP proxy by specifying the socket address where Squid | |
2360 | listens for FTP client requests. See http_port directive for various | |
2361 | ways to specify the listening address and mode. | |
2362 | ||
2363 | Usage: ftp_port address [mode] [options] | |
2364 | ||
2365 | WARNING: This is a new, experimental, complex feature that has seen | |
2366 | limited production exposure. Some Squid modules (e.g., caching) do not | |
2367 | currently work with native FTP proxying, and many features have not | |
2368 | even been tested for compatibility. Test well before deploying! | |
2369 | ||
2370 | Native FTP proxying differs substantially from proxying HTTP requests | |
2371 | with ftp:// URIs because Squid works as an FTP server and receives | |
2372 | actual FTP commands (rather than HTTP requests with FTP URLs). | |
2373 | ||
2374 | Native FTP commands accepted at ftp_port are internally converted or | |
2375 | wrapped into HTTP-like messages. The same happens to Native FTP | |
2376 | responses received from FTP origin servers. Those HTTP-like messages | |
2377 | are shoveled through regular access control and adaptation layers | |
2378 | between the FTP client and the FTP origin server. This allows Squid to | |
2379 | examine, adapt, block, and log FTP exchanges. Squid reuses most HTTP | |
2380 | mechanisms when shoveling wrapped FTP messages. For example, | |
2381 | http_access and adaptation_access directives are used. | |
2382 | ||
2383 | Modes: | |
2384 | ||
2385 | intercept Same as http_port intercept. The FTP origin address is | |
2386 | determined based on the intended destination of the | |
2387 | intercepted connection. | |
2388 | ||
2389 | tproxy Support Linux TPROXY for spoofing outgoing | |
2390 | connections using the client IP address. | |
2391 | NP: disables authentication and maybe IPv6 on the port. | |
2392 | ||
2393 | By default (i.e., without an explicit mode option), Squid extracts the | |
2394 | FTP origin address from the login@origin parameter of the FTP USER | |
2395 | command. Many popular FTP clients support such native FTP proxying. | |
2396 | ||
2397 | Options: | |
2398 | ||
2399 | name=token Specifies an internal name for the port. Defaults to | |
2400 | the port address. Usable with myportname ACL. | |
2401 | ||
2402 | ftp-track-dirs | |
2403 | Enables tracking of FTP directories by injecting extra | |
2404 | PWD commands and adjusting Request-URI (in wrapping | |
2405 | HTTP requests) to reflect the current FTP server | |
2406 | directory. Tracking is disabled by default. | |
2407 | ||
2408 | protocol=FTP Protocol to reconstruct accelerated and intercepted | |
2409 | requests with. Defaults to FTP. No other accepted | |
2410 | values have been tested with. An unsupported value | |
2411 | results in a FATAL error. Accepted values are FTP, | |
2412 | HTTP (or HTTP/1.1), and HTTPS (or HTTPS/1.1). | |
2413 | ||
2414 | Other http_port modes and options that are not specific to HTTP and | |
2415 | HTTPS may also work. | |
2416 | DOC_END | |
2417 | ||
2418 | NAME: tcp_outgoing_tos tcp_outgoing_ds tcp_outgoing_dscp | |
2419 | TYPE: acl_tos | |
2420 | DEFAULT: none | |
2421 | LOC: Ip::Qos::TheConfig.tosToServer | |
2422 | DOC_START | |
2423 | Allows you to select a TOS/Diffserv value for packets outgoing | |
2424 | on the server side, based on an ACL. | |
2425 | ||
2426 | tcp_outgoing_tos ds-field [!]aclname ... | |
2427 | ||
2428 | Example where normal_service_net uses the TOS value 0x00 | |
2429 | and good_service_net uses 0x20 | |
2430 | ||
2431 | acl normal_service_net src 10.0.0.0/24 | |
2432 | acl good_service_net src 10.0.1.0/24 | |
2433 | tcp_outgoing_tos 0x00 normal_service_net | |
2434 | tcp_outgoing_tos 0x20 good_service_net | |
2435 | ||
2436 | TOS/DSCP values really only have local significance - so you should | |
2437 | know what you're specifying. For more information, see RFC2474, | |
2438 | RFC2475, and RFC3260. | |
2439 | ||
2440 | The TOS/DSCP byte must be exactly that - a octet value 0 - 255, or | |
2441 | "default" to use whatever default your host has. | |
2442 | Note that only multiples of 4 are usable as the two rightmost bits have | |
2443 | been redefined for use by ECN (RFC 3168 section 23.1). | |
2444 | The squid parser will enforce this by masking away the ECN bits. | |
2445 | ||
2446 | Processing proceeds in the order specified, and stops at first fully | |
2447 | matching line. | |
2448 | ||
2449 | Only fast ACLs are supported. | |
2450 | DOC_END | |
2451 | ||
2452 | NAME: clientside_tos | |
2453 | TYPE: acl_tos | |
2454 | DEFAULT: none | |
2455 | LOC: Ip::Qos::TheConfig.tosToClient | |
2456 | DOC_START | |
2457 | Allows you to select a TOS/DSCP value for packets being transmitted | |
2458 | on the client-side, based on an ACL. | |
2459 | ||
2460 | clientside_tos ds-field [!]aclname ... | |
2461 | ||
2462 | Example where normal_service_net uses the TOS value 0x00 | |
2463 | and good_service_net uses 0x20 | |
2464 | ||
2465 | acl normal_service_net src 10.0.0.0/24 | |
2466 | acl good_service_net src 10.0.1.0/24 | |
2467 | clientside_tos 0x00 normal_service_net | |
2468 | clientside_tos 0x20 good_service_net | |
2469 | ||
2470 | Note: This feature is incompatible with qos_flows. Any TOS values set here | |
2471 | will be overwritten by TOS values in qos_flows. | |
2472 | ||
2473 | The TOS/DSCP byte must be exactly that - a octet value 0 - 255, or | |
2474 | "default" to use whatever default your host has. | |
2475 | Note that only multiples of 4 are usable as the two rightmost bits have | |
2476 | been redefined for use by ECN (RFC 3168 section 23.1). | |
2477 | The squid parser will enforce this by masking away the ECN bits. | |
2478 | ||
2479 | DOC_END | |
2480 | ||
2481 | NAME: tcp_outgoing_mark | |
2482 | TYPE: acl_nfmark | |
2483 | IFDEF: SO_MARK&&USE_LIBCAP | |
2484 | DEFAULT: none | |
2485 | LOC: Ip::Qos::TheConfig.nfmarkToServer | |
2486 | DOC_START | |
2487 | Allows you to apply a Netfilter mark value to outgoing packets | |
2488 | on the server side, based on an ACL. | |
2489 | ||
2490 | tcp_outgoing_mark mark-value [!]aclname ... | |
2491 | ||
2492 | Example where normal_service_net uses the mark value 0x00 | |
2493 | and good_service_net uses 0x20 | |
2494 | ||
2495 | acl normal_service_net src 10.0.0.0/24 | |
2496 | acl good_service_net src 10.0.1.0/24 | |
2497 | tcp_outgoing_mark 0x00 normal_service_net | |
2498 | tcp_outgoing_mark 0x20 good_service_net | |
2499 | ||
2500 | Only fast ACLs are supported. | |
2501 | DOC_END | |
2502 | ||
2503 | NAME: clientside_mark | |
2504 | TYPE: acl_nfmark | |
2505 | IFDEF: SO_MARK&&USE_LIBCAP | |
2506 | DEFAULT: none | |
2507 | LOC: Ip::Qos::TheConfig.nfmarkToClient | |
2508 | DOC_START | |
2509 | Allows you to apply a Netfilter mark value to packets being transmitted | |
2510 | on the client-side, based on an ACL. | |
2511 | ||
2512 | clientside_mark mark-value [!]aclname ... | |
2513 | ||
2514 | Example where normal_service_net uses the mark value 0x00 | |
2515 | and good_service_net uses 0x20 | |
2516 | ||
2517 | acl normal_service_net src 10.0.0.0/24 | |
2518 | acl good_service_net src 10.0.1.0/24 | |
2519 | clientside_mark 0x00 normal_service_net | |
2520 | clientside_mark 0x20 good_service_net | |
2521 | ||
2522 | Note: This feature is incompatible with qos_flows. Any mark values set here | |
2523 | will be overwritten by mark values in qos_flows. | |
2524 | DOC_END | |
2525 | ||
2526 | NAME: qos_flows | |
2527 | TYPE: QosConfig | |
2528 | IFDEF: USE_QOS_TOS | |
2529 | DEFAULT: none | |
2530 | LOC: Ip::Qos::TheConfig | |
2531 | DOC_START | |
2532 | Allows you to select a TOS/DSCP value to mark outgoing | |
2533 | connections to the client, based on where the reply was sourced. | |
2534 | For platforms using netfilter, allows you to set a netfilter mark | |
2535 | value instead of, or in addition to, a TOS value. | |
2536 | ||
2537 | By default this functionality is disabled. To enable it with the default | |
2538 | settings simply use "qos_flows mark" or "qos_flows tos". Default | |
2539 | settings will result in the netfilter mark or TOS value being copied | |
2540 | from the upstream connection to the client. Note that it is the connection | |
2541 | CONNMARK value not the packet MARK value that is copied. | |
2542 | ||
2543 | It is not currently possible to copy the mark or TOS value from the | |
2544 | client to the upstream connection request. | |
2545 | ||
2546 | TOS values really only have local significance - so you should | |
2547 | know what you're specifying. For more information, see RFC2474, | |
2548 | RFC2475, and RFC3260. | |
2549 | ||
2550 | The TOS/DSCP byte must be exactly that - a octet value 0 - 255. | |
2551 | Note that only multiples of 4 are usable as the two rightmost bits have | |
2552 | been redefined for use by ECN (RFC 3168 section 23.1). | |
2553 | The squid parser will enforce this by masking away the ECN bits. | |
2554 | ||
2555 | Mark values can be any unsigned 32-bit integer value. | |
2556 | ||
2557 | This setting is configured by setting the following values: | |
2558 | ||
2559 | tos|mark Whether to set TOS or netfilter mark values | |
2560 | ||
2561 | local-hit=0xFF Value to mark local cache hits. | |
2562 | ||
2563 | sibling-hit=0xFF Value to mark hits from sibling peers. | |
2564 | ||
2565 | parent-hit=0xFF Value to mark hits from parent peers. | |
2566 | ||
2567 | miss=0xFF[/mask] Value to mark cache misses. Takes precedence | |
2568 | over the preserve-miss feature (see below), unless | |
2569 | mask is specified, in which case only the bits | |
2570 | specified in the mask are written. | |
2571 | ||
2572 | The TOS variant of the following features are only possible on Linux | |
2573 | and require your kernel to be patched with the TOS preserving ZPH | |
2574 | patch, available from http://zph.bratcheda.org | |
2575 | No patch is needed to preserve the netfilter mark, which will work | |
2576 | with all variants of netfilter. | |
2577 | ||
2578 | disable-preserve-miss | |
2579 | This option disables the preservation of the TOS or netfilter | |
2580 | mark. By default, the existing TOS or netfilter mark value of | |
2581 | the response coming from the remote server will be retained | |
2582 | and masked with miss-mark. | |
2583 | NOTE: in the case of a netfilter mark, the mark must be set on | |
2584 | the connection (using the CONNMARK target) not on the packet | |
2585 | (MARK target). | |
2586 | ||
2587 | miss-mask=0xFF | |
2588 | Allows you to mask certain bits in the TOS or mark value | |
2589 | received from the remote server, before copying the value to | |
2590 | the TOS sent towards clients. | |
2591 | Default for tos: 0xFF (TOS from server is not changed). | |
2592 | Default for mark: 0xFFFFFFFF (mark from server is not changed). | |
2593 | ||
2594 | All of these features require the --enable-zph-qos compilation flag | |
2595 | (enabled by default). Netfilter marking also requires the | |
2596 | libnetfilter_conntrack libraries (--with-netfilter-conntrack) and | |
2597 | libcap 2.09+ (--with-libcap). | |
2598 | ||
2599 | DOC_END | |
2600 | ||
2601 | NAME: tcp_outgoing_address | |
2602 | TYPE: acl_address | |
2603 | DEFAULT: none | |
2604 | DEFAULT_DOC: Address selection is performed by the operating system. | |
2605 | LOC: Config.accessList.outgoing_address | |
2606 | DOC_START | |
2607 | Allows you to map requests to different outgoing IP addresses | |
2608 | based on the username or source address of the user making | |
2609 | the request. | |
2610 | ||
2611 | tcp_outgoing_address ipaddr [[!]aclname] ... | |
2612 | ||
2613 | For example; | |
2614 | Forwarding clients with dedicated IPs for certain subnets. | |
2615 | ||
2616 | acl normal_service_net src 10.0.0.0/24 | |
2617 | acl good_service_net src 10.0.2.0/24 | |
2618 | ||
2619 | tcp_outgoing_address 2001:db8::c001 good_service_net | |
2620 | tcp_outgoing_address 10.1.0.2 good_service_net | |
2621 | ||
2622 | tcp_outgoing_address 2001:db8::beef normal_service_net | |
2623 | tcp_outgoing_address 10.1.0.1 normal_service_net | |
2624 | ||
2625 | tcp_outgoing_address 2001:db8::1 | |
2626 | tcp_outgoing_address 10.1.0.3 | |
2627 | ||
2628 | Processing proceeds in the order specified, and stops at first fully | |
2629 | matching line. | |
2630 | ||
2631 | Squid will add an implicit IP version test to each line. | |
2632 | Requests going to IPv4 websites will use the outgoing 10.1.0.* addresses. | |
2633 | Requests going to IPv6 websites will use the outgoing 2001:db8:* addresses. | |
2634 | ||
2635 | ||
2636 | NOTE: The use of this directive using client dependent ACLs is | |
2637 | incompatible with the use of server side persistent connections. To | |
2638 | ensure correct results it is best to set server_persistent_connections | |
2639 | to off when using this directive in such configurations. | |
2640 | ||
2641 | NOTE: The use of this directive to set a local IP on outgoing TCP links | |
2642 | is incompatible with using TPROXY to set client IP out outbound TCP links. | |
2643 | When needing to contact peers use the no-tproxy cache_peer option and the | |
2644 | client_dst_passthru directive re-enable normal forwarding such as this. | |
2645 | ||
2646 | DOC_END | |
2647 | ||
2648 | NAME: host_verify_strict | |
2649 | TYPE: onoff | |
2650 | DEFAULT: off | |
2651 | LOC: Config.onoff.hostStrictVerify | |
2652 | DOC_START | |
2653 | Regardless of this option setting, when dealing with intercepted | |
2654 | traffic, Squid always verifies that the destination IP address matches | |
2655 | the Host header domain or IP (called 'authority form URL'). | |
2656 | ||
2657 | This enforcement is performed to satisfy a MUST-level requirement in | |
2658 | RFC 2616 section 14.23: "The Host field value MUST represent the naming | |
2659 | authority of the origin server or gateway given by the original URL". | |
2660 | ||
2661 | When set to ON: | |
2662 | Squid always responds with an HTTP 409 (Conflict) error | |
2663 | page and logs a security warning if there is no match. | |
2664 | ||
2665 | Squid verifies that the destination IP address matches | |
2666 | the Host header for forward-proxy and reverse-proxy traffic | |
2667 | as well. For those traffic types, Squid also enables the | |
2668 | following checks, comparing the corresponding Host header | |
2669 | and Request-URI components: | |
2670 | ||
2671 | * The host names (domain or IP) must be identical, | |
2672 | but valueless or missing Host header disables all checks. | |
2673 | For the two host names to match, both must be either IP | |
2674 | or FQDN. | |
2675 | ||
2676 | * Port numbers must be identical, but if a port is missing | |
2677 | the scheme-default port is assumed. | |
2678 | ||
2679 | ||
2680 | When set to OFF (the default): | |
2681 | Squid allows suspicious requests to continue but logs a | |
2682 | security warning and blocks caching of the response. | |
2683 | ||
2684 | * Forward-proxy traffic is not checked at all. | |
2685 | ||
2686 | * Reverse-proxy traffic is not checked at all. | |
2687 | ||
2688 | * Intercepted traffic which passes verification is handled | |
2689 | according to client_dst_passthru. | |
2690 | ||
2691 | * Intercepted requests which fail verification are sent | |
2692 | to the client original destination instead of DIRECT. | |
2693 | This overrides 'client_dst_passthru off'. | |
2694 | ||
2695 | For now suspicious intercepted CONNECT requests are always | |
2696 | responded to with an HTTP 409 (Conflict) error page. | |
2697 | ||
2698 | ||
2699 | SECURITY NOTE: | |
2700 | ||
2701 | As described in CVE-2009-0801 when the Host: header alone is used | |
2702 | to determine the destination of a request it becomes trivial for | |
2703 | malicious scripts on remote websites to bypass browser same-origin | |
2704 | security policy and sandboxing protections. | |
2705 | ||
2706 | The cause of this is that such applets are allowed to perform their | |
2707 | own HTTP stack, in which case the same-origin policy of the browser | |
2708 | sandbox only verifies that the applet tries to contact the same IP | |
2709 | as from where it was loaded at the IP level. The Host: header may | |
2710 | be different from the connected IP and approved origin. | |
2711 | ||
2712 | DOC_END | |
2713 | ||
2714 | NAME: client_dst_passthru | |
2715 | TYPE: onoff | |
2716 | DEFAULT: on | |
2717 | LOC: Config.onoff.client_dst_passthru | |
2718 | DOC_START | |
2719 | With NAT or TPROXY intercepted traffic Squid may pass the request | |
2720 | directly to the original client destination IP or seek a faster | |
2721 | source using the HTTP Host header. | |
2722 | ||
2723 | Using Host to locate alternative servers can provide faster | |
2724 | connectivity with a range of failure recovery options. | |
2725 | But can also lead to connectivity trouble when the client and | |
2726 | server are attempting stateful interactions unaware of the proxy. | |
2727 | ||
2728 | This option (on by default) prevents alternative DNS entries being | |
2729 | located to send intercepted traffic DIRECT to an origin server. | |
2730 | The clients original destination IP and port will be used instead. | |
2731 | ||
2732 | Regardless of this option setting, when dealing with intercepted | |
2733 | traffic Squid will verify the Host: header and any traffic which | |
2734 | fails Host verification will be treated as if this option were ON. | |
2735 | ||
2736 | see host_verify_strict for details on the verification process. | |
2737 | DOC_END | |
2738 | ||
2739 | COMMENT_START | |
2740 | TLS OPTIONS | |
2741 | ----------------------------------------------------------------------------- | |
2742 | COMMENT_END | |
2743 | ||
2744 | NAME: tls_outgoing_options | |
2745 | IFDEF: USE_GNUTLS||USE_OPENSSL | |
2746 | TYPE: securePeerOptions | |
2747 | DEFAULT: min-version=1.0 | |
2748 | LOC: Security::ProxyOutgoingConfig | |
2749 | DOC_START | |
2750 | disable Do not support https:// URLs. | |
2751 | ||
2752 | cert=/path/to/client/certificate | |
2753 | A client TLS certificate to use when connecting. | |
2754 | ||
2755 | key=/path/to/client/private_key | |
2756 | The private TLS key corresponding to the cert= above. | |
2757 | If key= is not specified cert= is assumed to reference | |
2758 | a PEM file containing both the certificate and the key. | |
2759 | ||
2760 | cipher=... The list of valid TLS ciphers to use. | |
2761 | ||
2762 | min-version=1.N | |
2763 | The minimum TLS protocol version to permit. | |
2764 | To control SSLv3 use the options= parameter. | |
2765 | Supported Values: 1.0 (default), 1.1, 1.2 | |
2766 | ||
2767 | options=... Specify various TLS/SSL implementation options. | |
2768 | ||
2769 | OpenSSL options most important are: | |
2770 | ||
2771 | NO_SSLv3 Disallow the use of SSLv3 | |
2772 | ||
2773 | SINGLE_DH_USE | |
2774 | Always create a new key when using | |
2775 | temporary/ephemeral DH key exchanges | |
2776 | ||
2777 | NO_TICKET | |
2778 | Disable use of RFC5077 session tickets. | |
2779 | Some servers may have problems | |
2780 | understanding the TLS extension due | |
2781 | to ambiguous specification in RFC4507. | |
2782 | ||
2783 | ALL Enable various bug workarounds | |
2784 | suggested as "harmless" by OpenSSL | |
2785 | Be warned that this reduces SSL/TLS | |
2786 | strength to some attacks. | |
2787 | ||
2788 | See the OpenSSL SSL_CTX_set_options documentation | |
2789 | for a more complete list. | |
2790 | ||
2791 | GnuTLS options most important are: | |
2792 | ||
2793 | %NO_TICKETS | |
2794 | Disable use of RFC5077 session tickets. | |
2795 | Some servers may have problems | |
2796 | understanding the TLS extension due | |
2797 | to ambiguous specification in RFC4507. | |
2798 | ||
2799 | See the GnuTLS Priority Strings documentation | |
2800 | for a more complete list. | |
2801 | http://www.gnutls.org/manual/gnutls.html#Priority-Strings | |
2802 | ||
2803 | ||
2804 | cafile= PEM file containing CA certificates to use when verifying | |
2805 | the peer certificate. May be repeated to load multiple files. | |
2806 | ||
2807 | capath= A directory containing additional CA certificates to | |
2808 | use when verifying the peer certificate. | |
2809 | Requires OpenSSL or LibreSSL. | |
2810 | ||
2811 | crlfile=... A certificate revocation list file to use when | |
2812 | verifying the peer certificate. | |
2813 | ||
2814 | flags=... Specify various flags modifying the TLS implementation: | |
2815 | ||
2816 | DONT_VERIFY_PEER | |
2817 | Accept certificates even if they fail to | |
2818 | verify. | |
2819 | DONT_VERIFY_DOMAIN | |
2820 | Don't verify the peer certificate | |
2821 | matches the server name | |
2822 | ||
2823 | default-ca[=off] | |
2824 | Whether to use the system Trusted CAs. Default is ON. | |
2825 | ||
2826 | domain= The peer name as advertised in its certificate. | |
2827 | Used for verifying the correctness of the received peer | |
2828 | certificate. If not specified the peer hostname will be | |
2829 | used. | |
2830 | DOC_END | |
2831 | ||
2832 | COMMENT_START | |
2833 | SSL OPTIONS | |
2834 | ----------------------------------------------------------------------------- | |
2835 | COMMENT_END | |
2836 | ||
2837 | NAME: ssl_unclean_shutdown | |
2838 | IFDEF: USE_OPENSSL | |
2839 | TYPE: onoff | |
2840 | DEFAULT: off | |
2841 | LOC: Config.SSL.unclean_shutdown | |
2842 | DOC_START | |
2843 | Some browsers (especially MSIE) bugs out on SSL shutdown | |
2844 | messages. | |
2845 | DOC_END | |
2846 | ||
2847 | NAME: ssl_engine | |
2848 | IFDEF: USE_OPENSSL | |
2849 | TYPE: string | |
2850 | LOC: Config.SSL.ssl_engine | |
2851 | DEFAULT: none | |
2852 | DOC_START | |
2853 | The OpenSSL engine to use. You will need to set this if you | |
2854 | would like to use hardware SSL acceleration for example. | |
2855 | DOC_END | |
2856 | ||
2857 | NAME: sslproxy_session_ttl | |
2858 | IFDEF: USE_OPENSSL | |
2859 | DEFAULT: 300 | |
2860 | LOC: Config.SSL.session_ttl | |
2861 | TYPE: int | |
2862 | DOC_START | |
2863 | Sets the timeout value for SSL sessions | |
2864 | DOC_END | |
2865 | ||
2866 | NAME: sslproxy_session_cache_size | |
2867 | IFDEF: USE_OPENSSL | |
2868 | DEFAULT: 2 MB | |
2869 | LOC: Config.SSL.sessionCacheSize | |
2870 | TYPE: b_size_t | |
2871 | DOC_START | |
2872 | Sets the cache size to use for ssl session | |
2873 | DOC_END | |
2874 | ||
2875 | NAME: sslproxy_foreign_intermediate_certs | |
2876 | IFDEF: USE_OPENSSL | |
2877 | DEFAULT: none | |
2878 | LOC: Config.ssl_client.foreignIntermediateCertsPath | |
2879 | TYPE: string | |
2880 | DOC_START | |
2881 | Many origin servers fail to send their full server certificate | |
2882 | chain for verification, assuming the client already has or can | |
2883 | easily locate any missing intermediate certificates. | |
2884 | ||
2885 | Squid uses the certificates from the specified file to fill in | |
2886 | these missing chains when trying to validate origin server | |
2887 | certificate chains. | |
2888 | ||
2889 | The file is expected to contain zero or more PEM-encoded | |
2890 | intermediate certificates. These certificates are not treated | |
2891 | as trusted root certificates, and any self-signed certificate in | |
2892 | this file will be ignored. | |
2893 | DOC_END | |
2894 | ||
2895 | NAME: sslproxy_cert_sign_hash | |
2896 | IFDEF: USE_OPENSSL | |
2897 | DEFAULT: none | |
2898 | LOC: Config.SSL.certSignHash | |
2899 | TYPE: string | |
2900 | DOC_START | |
2901 | Sets the hashing algorithm to use when signing generated certificates. | |
2902 | Valid algorithm names depend on the OpenSSL library used. The following | |
2903 | names are usually available: sha1, sha256, sha512, and md5. Please see | |
2904 | your OpenSSL library manual for the available hashes. By default, Squids | |
2905 | that support this option use sha256 hashes. | |
2906 | ||
2907 | Squid does not forcefully purge cached certificates that were generated | |
2908 | with an algorithm other than the currently configured one. They remain | |
2909 | in the cache, subject to the regular cache eviction policy, and become | |
2910 | useful if the algorithm changes again. | |
2911 | DOC_END | |
2912 | ||
2913 | NAME: ssl_bump | |
2914 | IFDEF: USE_OPENSSL | |
2915 | TYPE: sslproxy_ssl_bump | |
2916 | LOC: Config.accessList.ssl_bump | |
2917 | DEFAULT_DOC: Become a TCP tunnel without decrypting proxied traffic. | |
2918 | DEFAULT: none | |
2919 | DOC_START | |
2920 | This option is consulted when a CONNECT request is received on | |
2921 | an http_port (or a new connection is intercepted at an | |
2922 | https_port), provided that port was configured with an ssl-bump | |
2923 | flag. The subsequent data on the connection is either treated as | |
2924 | HTTPS and decrypted OR tunneled at TCP level without decryption, | |
2925 | depending on the first matching bumping "action". | |
2926 | ||
2927 | ssl_bump <action> [!]acl ... | |
2928 | ||
2929 | The following bumping actions are currently supported: | |
2930 | ||
2931 | splice | |
2932 | Become a TCP tunnel without decrypting proxied traffic. | |
2933 | This is the default action. | |
2934 | ||
2935 | bump | |
2936 | When used on step SslBump1, establishes a secure connection | |
2937 | with the client first, then connect to the server. | |
2938 | When used on step SslBump2 or SslBump3, establishes a secure | |
2939 | connection with the server and, using a mimicked server | |
2940 | certificate, with the client. | |
2941 | ||
2942 | peek | |
2943 | Receive client (step SslBump1) or server (step SslBump2) | |
2944 | certificate while preserving the possibility of splicing the | |
2945 | connection. Peeking at the server certificate (during step 2) | |
2946 | usually precludes bumping of the connection at step 3. | |
2947 | ||
2948 | stare | |
2949 | Receive client (step SslBump1) or server (step SslBump2) | |
2950 | certificate while preserving the possibility of bumping the | |
2951 | connection. Staring at the server certificate (during step 2) | |
2952 | usually precludes splicing of the connection at step 3. | |
2953 | ||
2954 | terminate | |
2955 | Close client and server connections. | |
2956 | ||
2957 | Backward compatibility actions available at step SslBump1: | |
2958 | ||
2959 | client-first | |
2960 | Bump the connection. Establish a secure connection with the | |
2961 | client first, then connect to the server. This old mode does | |
2962 | not allow Squid to mimic server SSL certificate and does not | |
2963 | work with intercepted SSL connections. | |
2964 | ||
2965 | server-first | |
2966 | Bump the connection. Establish a secure connection with the | |
2967 | server first, then establish a secure connection with the | |
2968 | client, using a mimicked server certificate. Works with both | |
2969 | CONNECT requests and intercepted SSL connections, but does | |
2970 | not allow to make decisions based on SSL handshake info. | |
2971 | ||
2972 | peek-and-splice | |
2973 | Decide whether to bump or splice the connection based on | |
2974 | client-to-squid and server-to-squid SSL hello messages. | |
2975 | XXX: Remove. | |
2976 | ||
2977 | none | |
2978 | Same as the "splice" action. | |
2979 | ||
2980 | All ssl_bump rules are evaluated at each of the supported bumping | |
2981 | steps. Rules with actions that are impossible at the current step are | |
2982 | ignored. The first matching ssl_bump action wins and is applied at the | |
2983 | end of the current step. If no rules match, the splice action is used. | |
2984 | See the at_step ACL for a list of the supported SslBump steps. | |
2985 | ||
2986 | This clause supports both fast and slow acl types. | |
2987 | See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details. | |
2988 | ||
2989 | See also: http_port ssl-bump, https_port ssl-bump, and acl at_step. | |
2990 | ||
2991 | ||
2992 | # Example: Bump all TLS connections except those originating from | |
2993 | # localhost or those going to example.com. | |
2994 | ||
2995 | acl broken_sites ssl::server_name .example.com | |
2996 | ssl_bump splice localhost | |
2997 | ssl_bump splice broken_sites | |
2998 | ssl_bump bump all | |
2999 | DOC_END | |
3000 | ||
3001 | NAME: sslproxy_cert_error | |
3002 | IFDEF: USE_OPENSSL | |
3003 | DEFAULT: none | |
3004 | DEFAULT_DOC: Server certificate errors terminate the transaction. | |
3005 | LOC: Config.ssl_client.cert_error | |
3006 | TYPE: acl_access | |
3007 | DOC_START | |
3008 | Use this ACL to bypass server certificate validation errors. | |
3009 | ||
3010 | For example, the following lines will bypass all validation errors | |
3011 | when talking to servers for example.com. All other | |
3012 | validation errors will result in ERR_SECURE_CONNECT_FAIL error. | |
3013 | ||
3014 | acl BrokenButTrustedServers dstdomain example.com | |
3015 | sslproxy_cert_error allow BrokenButTrustedServers | |
3016 | sslproxy_cert_error deny all | |
3017 | ||
3018 | This clause only supports fast acl types. | |
3019 | See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details. | |
3020 | Using slow acl types may result in server crashes | |
3021 | ||
3022 | Without this option, all server certificate validation errors | |
3023 | terminate the transaction to protect Squid and the client. | |
3024 | ||
3025 | SQUID_X509_V_ERR_INFINITE_VALIDATION error cannot be bypassed | |
3026 | but should not happen unless your OpenSSL library is buggy. | |
3027 | ||
3028 | SECURITY WARNING: | |
3029 | Bypassing validation errors is dangerous because an | |
3030 | error usually implies that the server cannot be trusted | |
3031 | and the connection may be insecure. | |
3032 | ||
3033 | See also: sslproxy_flags and DONT_VERIFY_PEER. | |
3034 | DOC_END | |
3035 | ||
3036 | NAME: sslproxy_cert_sign | |
3037 | IFDEF: USE_OPENSSL | |
3038 | DEFAULT: none | |
3039 | POSTSCRIPTUM: signUntrusted ssl::certUntrusted | |
3040 | POSTSCRIPTUM: signSelf ssl::certSelfSigned | |
3041 | POSTSCRIPTUM: signTrusted all | |
3042 | TYPE: sslproxy_cert_sign | |
3043 | LOC: Config.ssl_client.cert_sign | |
3044 | DOC_START | |
3045 | ||
3046 | sslproxy_cert_sign <signing algorithm> acl ... | |
3047 | ||
3048 | The following certificate signing algorithms are supported: | |
3049 | ||
3050 | signTrusted | |
3051 | Sign using the configured CA certificate which is usually | |
3052 | placed in and trusted by end-user browsers. This is the | |
3053 | default for trusted origin server certificates. | |
3054 | ||
3055 | signUntrusted | |
3056 | Sign to guarantee an X509_V_ERR_CERT_UNTRUSTED browser error. | |
3057 | This is the default for untrusted origin server certificates | |
3058 | that are not self-signed (see ssl::certUntrusted). | |
3059 | ||
3060 | signSelf | |
3061 | Sign using a self-signed certificate with the right CN to | |
3062 | generate a X509_V_ERR_DEPTH_ZERO_SELF_SIGNED_CERT error in the | |
3063 | browser. This is the default for self-signed origin server | |
3064 | certificates (see ssl::certSelfSigned). | |
3065 | ||
3066 | This clause only supports fast acl types. | |
3067 | ||
3068 | When sslproxy_cert_sign acl(s) match, Squid uses the corresponding | |
3069 | signing algorithm to generate the certificate and ignores all | |
3070 | subsequent sslproxy_cert_sign options (the first match wins). If no | |
3071 | acl(s) match, the default signing algorithm is determined by errors | |
3072 | detected when obtaining and validating the origin server certificate. | |
3073 | ||
3074 | WARNING: SQUID_X509_V_ERR_DOMAIN_MISMATCH and ssl:certDomainMismatch can | |
3075 | be used with sslproxy_cert_adapt, but if and only if Squid is bumping a | |
3076 | CONNECT request that carries a domain name. In all other cases (CONNECT | |
3077 | to an IP address or an intercepted SSL connection), Squid cannot detect | |
3078 | the domain mismatch at certificate generation time when | |
3079 | bump-server-first is used. | |
3080 | DOC_END | |
3081 | ||
3082 | NAME: sslproxy_cert_adapt | |
3083 | IFDEF: USE_OPENSSL | |
3084 | DEFAULT: none | |
3085 | TYPE: sslproxy_cert_adapt | |
3086 | LOC: Config.ssl_client.cert_adapt | |
3087 | DOC_START | |
3088 | ||
3089 | sslproxy_cert_adapt <adaptation algorithm> acl ... | |
3090 | ||
3091 | The following certificate adaptation algorithms are supported: | |
3092 | ||
3093 | setValidAfter | |
3094 | Sets the "Not After" property to the "Not After" property of | |
3095 | the CA certificate used to sign generated certificates. | |
3096 | ||
3097 | setValidBefore | |
3098 | Sets the "Not Before" property to the "Not Before" property of | |
3099 | the CA certificate used to sign generated certificates. | |
3100 | ||
3101 | setCommonName or setCommonName{CN} | |
3102 | Sets Subject.CN property to the host name specified as a | |
3103 | CN parameter or, if no explicit CN parameter was specified, | |
3104 | extracted from the CONNECT request. It is a misconfiguration | |
3105 | to use setCommonName without an explicit parameter for | |
3106 | intercepted or tproxied SSL connections. | |
3107 | ||
3108 | This clause only supports fast acl types. | |
3109 | ||
3110 | Squid first groups sslproxy_cert_adapt options by adaptation algorithm. | |
3111 | Within a group, when sslproxy_cert_adapt acl(s) match, Squid uses the | |
3112 | corresponding adaptation algorithm to generate the certificate and | |
3113 | ignores all subsequent sslproxy_cert_adapt options in that algorithm's | |
3114 | group (i.e., the first match wins within each algorithm group). If no | |
3115 | acl(s) match, the default mimicking action takes place. | |
3116 | ||
3117 | WARNING: SQUID_X509_V_ERR_DOMAIN_MISMATCH and ssl:certDomainMismatch can | |
3118 | be used with sslproxy_cert_adapt, but if and only if Squid is bumping a | |
3119 | CONNECT request that carries a domain name. In all other cases (CONNECT | |
3120 | to an IP address or an intercepted SSL connection), Squid cannot detect | |
3121 | the domain mismatch at certificate generation time when | |
3122 | bump-server-first is used. | |
3123 | DOC_END | |
3124 | ||
3125 | NAME: sslpassword_program | |
3126 | IFDEF: USE_OPENSSL | |
3127 | DEFAULT: none | |
3128 | LOC: Config.Program.ssl_password | |
3129 | TYPE: string | |
3130 | DOC_START | |
3131 | Specify a program used for entering SSL key passphrases | |
3132 | when using encrypted SSL certificate keys. If not specified | |
3133 | keys must either be unencrypted, or Squid started with the -N | |
3134 | option to allow it to query interactively for the passphrase. | |
3135 | ||
3136 | The key file name is given as argument to the program allowing | |
3137 | selection of the right password if you have multiple encrypted | |
3138 | keys. | |
3139 | DOC_END | |
3140 | ||
3141 | COMMENT_START | |
3142 | OPTIONS RELATING TO EXTERNAL SSL_CRTD | |
3143 | ----------------------------------------------------------------------------- | |
3144 | COMMENT_END | |
3145 | ||
3146 | NAME: sslcrtd_program | |
3147 | TYPE: eol | |
3148 | IFDEF: USE_SSL_CRTD | |
3149 | DEFAULT: @DEFAULT_SSL_CRTD@ -s @DEFAULT_SSL_DB_DIR@ -M 4MB | |
3150 | LOC: Ssl::TheConfig.ssl_crtd | |
3151 | DOC_START | |
3152 | Specify the location and options of the executable for certificate | |
3153 | generator. | |
3154 | @DEFAULT_SSL_CRTD@ program requires -s and -M parameters | |
3155 | For more information use: | |
3156 | @DEFAULT_SSL_CRTD@ -h | |
3157 | DOC_END | |
3158 | ||
3159 | NAME: sslcrtd_children | |
3160 | TYPE: HelperChildConfig | |
3161 | IFDEF: USE_SSL_CRTD | |
3162 | DEFAULT: 32 startup=5 idle=1 | |
3163 | LOC: Ssl::TheConfig.ssl_crtdChildren | |
3164 | DOC_START | |
3165 | The maximum number of processes spawn to service ssl server. | |
3166 | The maximum this may be safely set to is 32. | |
3167 | ||
3168 | The startup= and idle= options allow some measure of skew in your | |
3169 | tuning. | |
3170 | ||
3171 | startup=N | |
3172 | ||
3173 | Sets the minimum number of processes to spawn when Squid | |
3174 | starts or reconfigures. When set to zero the first request will | |
3175 | cause spawning of the first child process to handle it. | |
3176 | ||
3177 | Starting too few children temporary slows Squid under load while it | |
3178 | tries to spawn enough additional processes to cope with traffic. | |
3179 | ||
3180 | idle=N | |
3181 | ||
3182 | Sets a minimum of how many processes Squid is to try and keep available | |
3183 | at all times. When traffic begins to rise above what the existing | |
3184 | processes can handle this many more will be spawned up to the maximum | |
3185 | configured. A minimum setting of 1 is required. | |
3186 | ||
3187 | queue-size=N | |
3188 | ||
3189 | Sets the maximum number of queued requests. | |
3190 | If the queued requests exceed queue size for more than 3 minutes | |
3191 | squid aborts its operation. | |
3192 | The default value is set to 2*numberofchildren. | |
3193 | ||
3194 | You must have at least one ssl_crtd process. | |
3195 | DOC_END | |
3196 | ||
3197 | NAME: sslcrtvalidator_program | |
3198 | TYPE: eol | |
3199 | IFDEF: USE_OPENSSL | |
3200 | DEFAULT: none | |
3201 | LOC: Ssl::TheConfig.ssl_crt_validator | |
3202 | DOC_START | |
3203 | Specify the location and options of the executable for ssl_crt_validator | |
3204 | process. | |
3205 | ||
3206 | Usage: sslcrtvalidator_program [ttl=n] [cache=n] path ... | |
3207 | ||
3208 | Options: | |
3209 | ttl=n TTL in seconds for cached results. The default is 60 secs | |
3210 | cache=n limit the result cache size. The default value is 2048 | |
3211 | DOC_END | |
3212 | ||
3213 | NAME: sslcrtvalidator_children | |
3214 | TYPE: HelperChildConfig | |
3215 | IFDEF: USE_OPENSSL | |
3216 | DEFAULT: 32 startup=5 idle=1 concurrency=1 | |
3217 | LOC: Ssl::TheConfig.ssl_crt_validator_Children | |
3218 | DOC_START | |
3219 | The maximum number of processes spawn to service SSL server. | |
3220 | The maximum this may be safely set to is 32. | |
3221 | ||
3222 | The startup= and idle= options allow some measure of skew in your | |
3223 | tuning. | |
3224 | ||
3225 | startup=N | |
3226 | ||
3227 | Sets the minimum number of processes to spawn when Squid | |
3228 | starts or reconfigures. When set to zero the first request will | |
3229 | cause spawning of the first child process to handle it. | |
3230 | ||
3231 | Starting too few children temporary slows Squid under load while it | |
3232 | tries to spawn enough additional processes to cope with traffic. | |
3233 | ||
3234 | idle=N | |
3235 | ||
3236 | Sets a minimum of how many processes Squid is to try and keep available | |
3237 | at all times. When traffic begins to rise above what the existing | |
3238 | processes can handle this many more will be spawned up to the maximum | |
3239 | configured. A minimum setting of 1 is required. | |
3240 | ||
3241 | concurrency= | |
3242 | ||
3243 | The number of requests each certificate validator helper can handle in | |
3244 | parallel. A value of 0 indicates the certficate validator does not | |
3245 | support concurrency. Defaults to 1. | |
3246 | ||
3247 | When this directive is set to a value >= 1 then the protocol | |
3248 | used to communicate with the helper is modified to include | |
3249 | a request ID in front of the request/response. The request | |
3250 | ID from the request must be echoed back with the response | |
3251 | to that request. | |
3252 | ||
3253 | queue-size=N | |
3254 | ||
3255 | Sets the maximum number of queued requests. | |
3256 | If the queued requests exceed queue size for more than 3 minutes | |
3257 | squid aborts its operation. | |
3258 | The default value is set to 2*numberofchildren. | |
3259 | ||
3260 | You must have at least one ssl_crt_validator process. | |
3261 | DOC_END | |
3262 | ||
3263 | COMMENT_START | |
3264 | OPTIONS WHICH AFFECT THE NEIGHBOR SELECTION ALGORITHM | |
3265 | ----------------------------------------------------------------------------- | |
3266 | COMMENT_END | |
3267 | ||
3268 | NAME: cache_peer | |
3269 | TYPE: peer | |
3270 | DEFAULT: none | |
3271 | LOC: Config.peers | |
3272 | DOC_START | |
3273 | To specify other caches in a hierarchy, use the format: | |
3274 | ||
3275 | cache_peer hostname type http-port icp-port [options] | |
3276 | ||
3277 | For example, | |
3278 | ||
3279 | # proxy icp | |
3280 | # hostname type port port options | |
3281 | # -------------------- -------- ----- ----- ----------- | |
3282 | cache_peer parent.foo.net parent 3128 3130 default | |
3283 | cache_peer sib1.foo.net sibling 3128 3130 proxy-only | |
3284 | cache_peer sib2.foo.net sibling 3128 3130 proxy-only | |
3285 | cache_peer example.com parent 80 0 default | |
3286 | cache_peer cdn.example.com sibling 3128 0 | |
3287 | ||
3288 | type: either 'parent', 'sibling', or 'multicast'. | |
3289 | ||
3290 | proxy-port: The port number where the peer accept HTTP requests. | |
3291 | For other Squid proxies this is usually 3128 | |
3292 | For web servers this is usually 80 | |
3293 | ||
3294 | icp-port: Used for querying neighbor caches about objects. | |
3295 | Set to 0 if the peer does not support ICP or HTCP. | |
3296 | See ICP and HTCP options below for additional details. | |
3297 | ||
3298 | ||
3299 | ==== ICP OPTIONS ==== | |
3300 | ||
3301 | You MUST also set icp_port and icp_access explicitly when using these options. | |
3302 | The defaults will prevent peer traffic using ICP. | |
3303 | ||
3304 | ||
3305 | no-query Disable ICP queries to this neighbor. | |
3306 | ||
3307 | multicast-responder | |
3308 | Indicates the named peer is a member of a multicast group. | |
3309 | ICP queries will not be sent directly to the peer, but ICP | |
3310 | replies will be accepted from it. | |
3311 | ||
3312 | closest-only Indicates that, for ICP_OP_MISS replies, we'll only forward | |
3313 | CLOSEST_PARENT_MISSes and never FIRST_PARENT_MISSes. | |
3314 | ||
3315 | background-ping | |
3316 | To only send ICP queries to this neighbor infrequently. | |
3317 | This is used to keep the neighbor round trip time updated | |
3318 | and is usually used in conjunction with weighted-round-robin. | |
3319 | ||
3320 | ||
3321 | ==== HTCP OPTIONS ==== | |
3322 | ||
3323 | You MUST also set htcp_port and htcp_access explicitly when using these options. | |
3324 | The defaults will prevent peer traffic using HTCP. | |
3325 | ||
3326 | ||
3327 | htcp Send HTCP, instead of ICP, queries to the neighbor. | |
3328 | You probably also want to set the "icp-port" to 4827 | |
3329 | instead of 3130. This directive accepts a comma separated | |
3330 | list of options described below. | |
3331 | ||
3332 | htcp=oldsquid Send HTCP to old Squid versions (2.5 or earlier). | |
3333 | ||
3334 | htcp=no-clr Send HTCP to the neighbor but without | |
3335 | sending any CLR requests. This cannot be used with | |
3336 | only-clr. | |
3337 | ||
3338 | htcp=only-clr Send HTCP to the neighbor but ONLY CLR requests. | |
3339 | This cannot be used with no-clr. | |
3340 | ||
3341 | htcp=no-purge-clr | |
3342 | Send HTCP to the neighbor including CLRs but only when | |
3343 | they do not result from PURGE requests. | |
3344 | ||
3345 | htcp=forward-clr | |
3346 | Forward any HTCP CLR requests this proxy receives to the peer. | |
3347 | ||
3348 | ||
3349 | ==== PEER SELECTION METHODS ==== | |
3350 | ||
3351 | The default peer selection method is ICP, with the first responding peer | |
3352 | being used as source. These options can be used for better load balancing. | |
3353 | ||
3354 | ||
3355 | default This is a parent cache which can be used as a "last-resort" | |
3356 | if a peer cannot be located by any of the peer-selection methods. | |
3357 | If specified more than once, only the first is used. | |
3358 | ||
3359 | round-robin Load-Balance parents which should be used in a round-robin | |
3360 | fashion in the absence of any ICP queries. | |
3361 | weight=N can be used to add bias. | |
3362 | ||
3363 | weighted-round-robin | |
3364 | Load-Balance parents which should be used in a round-robin | |
3365 | fashion with the frequency of each parent being based on the | |
3366 | round trip time. Closer parents are used more often. | |
3367 | Usually used for background-ping parents. | |
3368 | weight=N can be used to add bias. | |
3369 | ||
3370 | carp Load-Balance parents which should be used as a CARP array. | |
3371 | The requests will be distributed among the parents based on the | |
3372 | CARP load balancing hash function based on their weight. | |
3373 | ||
3374 | userhash Load-balance parents based on the client proxy_auth or ident username. | |
3375 | ||
3376 | sourcehash Load-balance parents based on the client source IP. | |
3377 | ||
3378 | multicast-siblings | |
3379 | To be used only for cache peers of type "multicast". | |
3380 | ALL members of this multicast group have "sibling" | |
3381 | relationship with it, not "parent". This is to a multicast | |
3382 | group when the requested object would be fetched only from | |
3383 | a "parent" cache, anyway. It's useful, e.g., when | |
3384 | configuring a pool of redundant Squid proxies, being | |
3385 | members of the same multicast group. | |
3386 | ||
3387 | ||
3388 | ==== PEER SELECTION OPTIONS ==== | |
3389 | ||
3390 | weight=N use to affect the selection of a peer during any weighted | |
3391 | peer-selection mechanisms. | |
3392 | The weight must be an integer; default is 1, | |
3393 | larger weights are favored more. | |
3394 | This option does not affect parent selection if a peering | |
3395 | protocol is not in use. | |
3396 | ||
3397 | basetime=N Specify a base amount to be subtracted from round trip | |
3398 | times of parents. | |
3399 | It is subtracted before division by weight in calculating | |
3400 | which parent to fectch from. If the rtt is less than the | |
3401 | base time the rtt is set to a minimal value. | |
3402 | ||
3403 | ttl=N Specify a TTL to use when sending multicast ICP queries | |
3404 | to this address. | |
3405 | Only useful when sending to a multicast group. | |
3406 | Because we don't accept ICP replies from random | |
3407 | hosts, you must configure other group members as | |
3408 | peers with the 'multicast-responder' option. | |
3409 | ||
3410 | no-delay To prevent access to this neighbor from influencing the | |
3411 | delay pools. | |
3412 | ||
3413 | digest-url=URL Tell Squid to fetch the cache digest (if digests are | |
3414 | enabled) for this host from the specified URL rather | |
3415 | than the Squid default location. | |
3416 | ||
3417 | ||
3418 | ==== CARP OPTIONS ==== | |
3419 | ||
3420 | carp-key=key-specification | |
3421 | use a different key than the full URL to hash against the peer. | |
3422 | the key-specification is a comma-separated list of the keywords | |
3423 | scheme, host, port, path, params | |
3424 | Order is not important. | |
3425 | ||
3426 | ==== ACCELERATOR / REVERSE-PROXY OPTIONS ==== | |
3427 | ||
3428 | originserver Causes this parent to be contacted as an origin server. | |
3429 | Meant to be used in accelerator setups when the peer | |
3430 | is a web server. | |
3431 | ||
3432 | forceddomain=name | |
3433 | Set the Host header of requests forwarded to this peer. | |
3434 | Useful in accelerator setups where the server (peer) | |
3435 | expects a certain domain name but clients may request | |
3436 | others. ie example.com or www.example.com | |
3437 | ||
3438 | no-digest Disable request of cache digests. | |
3439 | ||
3440 | no-netdb-exchange | |
3441 | Disables requesting ICMP RTT database (NetDB). | |
3442 | ||
3443 | ||
3444 | ==== AUTHENTICATION OPTIONS ==== | |
3445 | ||
3446 | login=user:password | |
3447 | If this is a personal/workgroup proxy and your parent | |
3448 | requires proxy authentication. | |
3449 | ||
3450 | Note: The string can include URL escapes (i.e. %20 for | |
3451 | spaces). This also means % must be written as %%. | |
3452 | ||
3453 | login=PASSTHRU | |
3454 | Send login details received from client to this peer. | |
3455 | Both Proxy- and WWW-Authorization headers are passed | |
3456 | without alteration to the peer. | |
3457 | Authentication is not required by Squid for this to work. | |
3458 | ||
3459 | Note: This will pass any form of authentication but | |
3460 | only Basic auth will work through a proxy unless the | |
3461 | connection-auth options are also used. | |
3462 | ||
3463 | login=PASS Send login details received from client to this peer. | |
3464 | Authentication is not required by this option. | |
3465 | ||
3466 | If there are no client-provided authentication headers | |
3467 | to pass on, but username and password are available | |
3468 | from an external ACL user= and password= result tags | |
3469 | they may be sent instead. | |
3470 | ||
3471 | Note: To combine this with proxy_auth both proxies must | |
3472 | share the same user database as HTTP only allows for | |
3473 | a single login (one for proxy, one for origin server). | |
3474 | Also be warned this will expose your users proxy | |
3475 | password to the peer. USE WITH CAUTION | |
3476 | ||
3477 | login=*:password | |
3478 | Send the username to the upstream cache, but with a | |
3479 | fixed password. This is meant to be used when the peer | |
3480 | is in another administrative domain, but it is still | |
3481 | needed to identify each user. | |
3482 | The star can optionally be followed by some extra | |
3483 | information which is added to the username. This can | |
3484 | be used to identify this proxy to the peer, similar to | |
3485 | the login=username:password option above. | |
3486 | ||
3487 | login=NEGOTIATE | |
3488 | If this is a personal/workgroup proxy and your parent | |
3489 | requires a secure proxy authentication. | |
3490 | The first principal from the default keytab or defined by | |
3491 | the environment variable KRB5_KTNAME will be used. | |
3492 | ||
3493 | WARNING: The connection may transmit requests from multiple | |
3494 | clients. Negotiate often assumes end-to-end authentication | |
3495 | and a single-client. Which is not strictly true here. | |
3496 | ||
3497 | login=NEGOTIATE:principal_name | |
3498 | If this is a personal/workgroup proxy and your parent | |
3499 | requires a secure proxy authentication. | |
3500 | The principal principal_name from the default keytab or | |
3501 | defined by the environment variable KRB5_KTNAME will be | |
3502 | used. | |
3503 | ||
3504 | WARNING: The connection may transmit requests from multiple | |
3505 | clients. Negotiate often assumes end-to-end authentication | |
3506 | and a single-client. Which is not strictly true here. | |
3507 | ||
3508 | connection-auth=on|off | |
3509 | Tell Squid that this peer does or not support Microsoft | |
3510 | connection oriented authentication, and any such | |
3511 | challenges received from there should be ignored. | |
3512 | Default is auto to automatically determine the status | |
3513 | of the peer. | |
3514 | ||
3515 | auth-no-keytab | |
3516 | Do not use a keytab to authenticate to a peer when | |
3517 | login=NEGOTIATE is specified. Let the GSSAPI | |
3518 | implementation determine which already existing | |
3519 | credentials cache to use instead. | |
3520 | ||
3521 | ||
3522 | ==== SSL / HTTPS / TLS OPTIONS ==== | |
3523 | ||
3524 | tls Encrypt connections to this peer with TLS. | |
3525 | ||
3526 | sslcert=/path/to/ssl/certificate | |
3527 | A client SSL certificate to use when connecting to | |
3528 | this peer. | |
3529 | ||
3530 | sslkey=/path/to/ssl/key | |
3531 | The private SSL key corresponding to sslcert above. | |
3532 | If 'sslkey' is not specified 'sslcert' is assumed to | |
3533 | reference a combined file containing both the | |
3534 | certificate and the key. | |
3535 | ||
3536 | sslcipher=... The list of valid SSL ciphers to use when connecting | |
3537 | to this peer. | |
3538 | ||
3539 | tls-min-version=1.N | |
3540 | The minimum TLS protocol version to permit. To control | |
3541 | SSLv3 use the tls-options= parameter. | |
3542 | Supported Values: 1.0 (default), 1.1, 1.2 | |
3543 | ||
3544 | tls-options=... Specify various TLS implementation options. | |
3545 | ||
3546 | OpenSSL options most important are: | |
3547 | ||
3548 | NO_SSLv3 Disallow the use of SSLv3 | |
3549 | ||
3550 | SINGLE_DH_USE | |
3551 | Always create a new key when using | |
3552 | temporary/ephemeral DH key exchanges | |
3553 | ||
3554 | NO_TICKET | |
3555 | Disable use of RFC5077 session tickets. | |
3556 | Some servers may have problems | |
3557 | understanding the TLS extension due | |
3558 | to ambiguous specification in RFC4507. | |
3559 | ||
3560 | ALL Enable various bug workarounds | |
3561 | suggested as "harmless" by OpenSSL | |
3562 | Be warned that this reduces SSL/TLS | |
3563 | strength to some attacks. | |
3564 | ||
3565 | See the OpenSSL SSL_CTX_set_options documentation for a | |
3566 | more complete list. | |
3567 | ||
3568 | GnuTLS options most important are: | |
3569 | ||
3570 | %NO_TICKETS | |
3571 | Disable use of RFC5077 session tickets. | |
3572 | Some servers may have problems | |
3573 | understanding the TLS extension due | |
3574 | to ambiguous specification in RFC4507. | |
3575 | ||
3576 | See the GnuTLS Priority Strings documentation | |
3577 | for a more complete list. | |
3578 | http://www.gnutls.org/manual/gnutls.html#Priority-Strings | |
3579 | ||
3580 | tls-cafile= PEM file containing CA certificates to use when verifying | |
3581 | the peer certificate. May be repeated to load multiple files. | |
3582 | ||
3583 | sslcapath=... A directory containing additional CA certificates to | |
3584 | use when verifying the peer certificate. | |
3585 | Requires OpenSSL or LibreSSL. | |
3586 | ||
3587 | sslcrlfile=... A certificate revocation list file to use when | |
3588 | verifying the peer certificate. | |
3589 | ||
3590 | sslflags=... Specify various flags modifying the SSL implementation: | |
3591 | ||
3592 | DONT_VERIFY_PEER | |
3593 | Accept certificates even if they fail to | |
3594 | verify. | |
3595 | ||
3596 | DONT_VERIFY_DOMAIN | |
3597 | Don't verify the peer certificate | |
3598 | matches the server name | |
3599 | ||
3600 | ssldomain= The peer name as advertised in it's certificate. | |
3601 | Used for verifying the correctness of the received peer | |
3602 | certificate. If not specified the peer hostname will be | |
3603 | used. | |
3604 | ||
3605 | front-end-https[=off|on|auto] | |
3606 | Enable the "Front-End-Https: On" header needed when | |
3607 | using Squid as a SSL frontend in front of Microsoft OWA. | |
3608 | See MS KB document Q307347 for details on this header. | |
3609 | If set to auto the header will only be added if the | |
3610 | request is forwarded as a https:// URL. | |
3611 | ||
3612 | tls-default-ca[=off] | |
3613 | Whether to use the system Trusted CAs. Default is ON. | |
3614 | ||
3615 | tls-no-npn Do not use the TLS NPN extension to advertise HTTP/1.1. | |
3616 | ||
3617 | ==== GENERAL OPTIONS ==== | |
3618 | ||
3619 | connect-timeout=N | |
3620 | A peer-specific connect timeout. | |
3621 | Also see the peer_connect_timeout directive. | |
3622 | ||
3623 | connect-fail-limit=N | |
3624 | How many times connecting to a peer must fail before | |
3625 | it is marked as down. Standby connection failures | |
3626 | count towards this limit. Default is 10. | |
3627 | ||
3628 | allow-miss Disable Squid's use of only-if-cached when forwarding | |
3629 | requests to siblings. This is primarily useful when | |
3630 | icp_hit_stale is used by the sibling. Excessive use | |
3631 | of this option may result in forwarding loops. One way | |
3632 | to prevent peering loops when using this option, is to | |
3633 | deny cache peer usage on requests from a peer: | |
3634 | acl fromPeer ... | |
3635 | cache_peer_access peerName deny fromPeer | |
3636 | ||
3637 | max-conn=N Limit the number of concurrent connections the Squid | |
3638 | may open to this peer, including already opened idle | |
3639 | and standby connections. There is no peer-specific | |
3640 | connection limit by default. | |
3641 | ||
3642 | A peer exceeding the limit is not used for new | |
3643 | requests unless a standby connection is available. | |
3644 | ||
3645 | max-conn currently works poorly with idle persistent | |
3646 | connections: When a peer reaches its max-conn limit, | |
3647 | and there are idle persistent connections to the peer, | |
3648 | the peer may not be selected because the limiting code | |
3649 | does not know whether Squid can reuse those idle | |
3650 | connections. | |
3651 | ||
3652 | standby=N Maintain a pool of N "hot standby" connections to an | |
3653 | UP peer, available for requests when no idle | |
3654 | persistent connection is available (or safe) to use. | |
3655 | By default and with zero N, no such pool is maintained. | |
3656 | N must not exceed the max-conn limit (if any). | |
3657 | ||
3658 | At start or after reconfiguration, Squid opens new TCP | |
3659 | standby connections until there are N connections | |
3660 | available and then replenishes the standby pool as | |
3661 | opened connections are used up for requests. A used | |
3662 | connection never goes back to the standby pool, but | |
3663 | may go to the regular idle persistent connection pool | |
3664 | shared by all peers and origin servers. | |
3665 | ||
3666 | Squid never opens multiple new standby connections | |
3667 | concurrently. This one-at-a-time approach minimizes | |
3668 | flooding-like effect on peers. Furthermore, just a few | |
3669 | standby connections should be sufficient in most cases | |
3670 | to supply most new requests with a ready-to-use | |
3671 | connection. | |
3672 | ||
3673 | Standby connections obey server_idle_pconn_timeout. | |
3674 | For the feature to work as intended, the peer must be | |
3675 | configured to accept and keep them open longer than | |
3676 | the idle timeout at the connecting Squid, to minimize | |
3677 | race conditions typical to idle used persistent | |
3678 | connections. Default request_timeout and | |
3679 | server_idle_pconn_timeout values ensure such a | |
3680 | configuration. | |
3681 | ||
3682 | name=xxx Unique name for the peer. | |
3683 | Required if you have multiple peers on the same host | |
3684 | but different ports. | |
3685 | This name can be used in cache_peer_access and similar | |
3686 | directives to identify the peer. | |
3687 | Can be used by outgoing access controls through the | |
3688 | peername ACL type. | |
3689 | ||
3690 | no-tproxy Do not use the client-spoof TPROXY support when forwarding | |
3691 | requests to this peer. Use normal address selection instead. | |
3692 | This overrides the spoof_client_ip ACL. | |
3693 | ||
3694 | proxy-only objects fetched from the peer will not be stored locally. | |
3695 | ||
3696 | DOC_END | |
3697 | ||
3698 | NAME: cache_peer_access | |
3699 | TYPE: peer_access | |
3700 | DEFAULT: none | |
3701 | DEFAULT_DOC: No peer usage restrictions. | |
3702 | LOC: none | |
3703 | DOC_START | |
3704 | Restricts usage of cache_peer proxies. | |
3705 | ||
3706 | Usage: | |
3707 | cache_peer_access peer-name allow|deny [!]aclname ... | |
3708 | ||
3709 | For the required peer-name parameter, use either the value of the | |
3710 | cache_peer name=value parameter or, if name=value is missing, the | |
3711 | cache_peer hostname parameter. | |
3712 | ||
3713 | This directive narrows down the selection of peering candidates, but | |
3714 | does not determine the order in which the selected candidates are | |
3715 | contacted. That order is determined by the peer selection algorithms | |
3716 | (see PEER SELECTION sections in the cache_peer documentation). | |
3717 | ||
3718 | If a deny rule matches, the corresponding peer will not be contacted | |
3719 | for the current transaction -- Squid will not send ICP queries and | |
3720 | will not forward HTTP requests to that peer. An allow match leaves | |
3721 | the corresponding peer in the selection. The first match for a given | |
3722 | peer wins for that peer. | |
3723 | ||
3724 | The relative order of cache_peer_access directives for the same peer | |
3725 | matters. The relative order of any two cache_peer_access directives | |
3726 | for different peers does not matter. To ease interpretation, it is a | |
3727 | good idea to group cache_peer_access directives for the same peer | |
3728 | together. | |
3729 | ||
3730 | A single cache_peer_access directive may be evaluated multiple times | |
3731 | for a given transaction because individual peer selection algorithms | |
3732 | may check it independently from each other. These redundant checks | |
3733 | may be optimized away in future Squid versions. | |
3734 | ||
3735 | This clause only supports fast acl types. | |
3736 | See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details. | |
3737 | ||
3738 | DOC_END | |
3739 | ||
3740 | NAME: neighbor_type_domain | |
3741 | TYPE: hostdomaintype | |
3742 | DEFAULT: none | |
3743 | DEFAULT_DOC: The peer type from cache_peer directive is used for all requests to that peer. | |
3744 | LOC: none | |
3745 | DOC_START | |
3746 | Modify the cache_peer neighbor type when passing requests | |
3747 | about specific domains to the peer. | |
3748 | ||
3749 | Usage: | |
3750 | neighbor_type_domain neighbor parent|sibling domain domain ... | |
3751 | ||
3752 | For example: | |
3753 | cache_peer foo.example.com parent 3128 3130 | |
3754 | neighbor_type_domain foo.example.com sibling .au .de | |
3755 | ||
3756 | The above configuration treats all requests to foo.example.com as a | |
3757 | parent proxy unless the request is for a .au or .de ccTLD domain name. | |
3758 | DOC_END | |
3759 | ||
3760 | NAME: dead_peer_timeout | |
3761 | COMMENT: (seconds) | |
3762 | DEFAULT: 10 seconds | |
3763 | TYPE: time_t | |
3764 | LOC: Config.Timeout.deadPeer | |
3765 | DOC_START | |
3766 | This controls how long Squid waits to declare a peer cache | |
3767 | as "dead." If there are no ICP replies received in this | |
3768 | amount of time, Squid will declare the peer dead and not | |
3769 | expect to receive any further ICP replies. However, it | |
3770 | continues to send ICP queries, and will mark the peer as | |
3771 | alive upon receipt of the first subsequent ICP reply. | |
3772 | ||
3773 | This timeout also affects when Squid expects to receive ICP | |
3774 | replies from peers. If more than 'dead_peer' seconds have | |
3775 | passed since the last ICP reply was received, Squid will not | |
3776 | expect to receive an ICP reply on the next query. Thus, if | |
3777 | your time between requests is greater than this timeout, you | |
3778 | will see a lot of requests sent DIRECT to origin servers | |
3779 | instead of to your parents. | |
3780 | DOC_END | |
3781 | ||
3782 | NAME: forward_max_tries | |
3783 | DEFAULT: 25 | |
3784 | TYPE: int | |
3785 | LOC: Config.forward_max_tries | |
3786 | DOC_START | |
3787 | Controls how many different forward paths Squid will try | |
3788 | before giving up. See also forward_timeout. | |
3789 | ||
3790 | NOTE: connect_retries (default: none) can make each of these | |
3791 | possible forwarding paths be tried multiple times. | |
3792 | DOC_END | |
3793 | ||
3794 | COMMENT_START | |
3795 | MEMORY CACHE OPTIONS | |
3796 | ----------------------------------------------------------------------------- | |
3797 | COMMENT_END | |
3798 | ||
3799 | NAME: cache_mem | |
3800 | COMMENT: (bytes) | |
3801 | TYPE: b_size_t | |
3802 | DEFAULT: 256 MB | |
3803 | LOC: Config.memMaxSize | |
3804 | DOC_START | |
3805 | NOTE: THIS PARAMETER DOES NOT SPECIFY THE MAXIMUM PROCESS SIZE. | |
3806 | IT ONLY PLACES A LIMIT ON HOW MUCH ADDITIONAL MEMORY SQUID WILL | |
3807 | USE AS A MEMORY CACHE OF OBJECTS. SQUID USES MEMORY FOR OTHER | |
3808 | THINGS AS WELL. SEE THE SQUID FAQ SECTION 8 FOR DETAILS. | |
3809 | ||
3810 | 'cache_mem' specifies the ideal amount of memory to be used | |
3811 | for: | |
3812 | * In-Transit objects | |
3813 | * Hot Objects | |
3814 | * Negative-Cached objects | |
3815 | ||
3816 | Data for these objects are stored in 4 KB blocks. This | |
3817 | parameter specifies the ideal upper limit on the total size of | |
3818 | 4 KB blocks allocated. In-Transit objects take the highest | |
3819 | priority. | |
3820 | ||
3821 | In-transit objects have priority over the others. When | |
3822 | additional space is needed for incoming data, negative-cached | |
3823 | and hot objects will be released. In other words, the | |
3824 | negative-cached and hot objects will fill up any unused space | |
3825 | not needed for in-transit objects. | |
3826 | ||
3827 | If circumstances require, this limit will be exceeded. | |
3828 | Specifically, if your incoming request rate requires more than | |
3829 | 'cache_mem' of memory to hold in-transit objects, Squid will | |
3830 | exceed this limit to satisfy the new requests. When the load | |
3831 | decreases, blocks will be freed until the high-water mark is | |
3832 | reached. Thereafter, blocks will be used to store hot | |
3833 | objects. | |
3834 | ||
3835 | If shared memory caching is enabled, Squid does not use the shared | |
3836 | cache space for in-transit objects, but they still consume as much | |
3837 | local memory as they need. For more details about the shared memory | |
3838 | cache, see memory_cache_shared. | |
3839 | DOC_END | |
3840 | ||
3841 | NAME: maximum_object_size_in_memory | |
3842 | COMMENT: (bytes) | |
3843 | TYPE: b_size_t | |
3844 | DEFAULT: 512 KB | |
3845 | LOC: Config.Store.maxInMemObjSize | |
3846 | DOC_START | |
3847 | Objects greater than this size will not be attempted to kept in | |
3848 | the memory cache. This should be set high enough to keep objects | |
3849 | accessed frequently in memory to improve performance whilst low | |
3850 | enough to keep larger objects from hoarding cache_mem. | |
3851 | DOC_END | |
3852 | ||
3853 | NAME: memory_cache_shared | |
3854 | COMMENT: on|off | |
3855 | TYPE: YesNoNone | |
3856 | LOC: Config.memShared | |
3857 | DEFAULT: none | |
3858 | DEFAULT_DOC: "on" where supported if doing memory caching with multiple SMP workers. | |
3859 | DOC_START | |
3860 | Controls whether the memory cache is shared among SMP workers. | |
3861 | ||
3862 | The shared memory cache is meant to occupy cache_mem bytes and replace | |
3863 | the non-shared memory cache, although some entities may still be | |
3864 | cached locally by workers for now (e.g., internal and in-transit | |
3865 | objects may be served from a local memory cache even if shared memory | |
3866 | caching is enabled). | |
3867 | ||
3868 | By default, the memory cache is shared if and only if all of the | |
3869 | following conditions are satisfied: Squid runs in SMP mode with | |
3870 | multiple workers, cache_mem is positive, and Squid environment | |
3871 | supports required IPC primitives (e.g., POSIX shared memory segments | |
3872 | and GCC-style atomic operations). | |
3873 | ||
3874 | To avoid blocking locks, shared memory uses opportunistic algorithms | |
3875 | that do not guarantee that every cachable entity that could have been | |
3876 | shared among SMP workers will actually be shared. | |
3877 | DOC_END | |
3878 | ||
3879 | NAME: memory_cache_mode | |
3880 | TYPE: memcachemode | |
3881 | LOC: Config | |
3882 | DEFAULT: always | |
3883 | DEFAULT_DOC: Keep the most recently fetched objects in memory | |
3884 | DOC_START | |
3885 | Controls which objects to keep in the memory cache (cache_mem) | |
3886 | ||
3887 | always Keep most recently fetched objects in memory (default) | |
3888 | ||
3889 | disk Only disk cache hits are kept in memory, which means | |
3890 | an object must first be cached on disk and then hit | |
3891 | a second time before cached in memory. | |
3892 | ||
3893 | network Only objects fetched from network is kept in memory | |
3894 | DOC_END | |
3895 | ||
3896 | NAME: memory_replacement_policy | |
3897 | TYPE: removalpolicy | |
3898 | LOC: Config.memPolicy | |
3899 | DEFAULT: lru | |
3900 | DOC_START | |
3901 | The memory replacement policy parameter determines which | |
3902 | objects are purged from memory when memory space is needed. | |
3903 | ||
3904 | See cache_replacement_policy for details on algorithms. | |
3905 | DOC_END | |
3906 | ||
3907 | COMMENT_START | |
3908 | DISK CACHE OPTIONS | |
3909 | ----------------------------------------------------------------------------- | |
3910 | COMMENT_END | |
3911 | ||
3912 | NAME: cache_replacement_policy | |
3913 | TYPE: removalpolicy | |
3914 | LOC: Config.replPolicy | |
3915 | DEFAULT: lru | |
3916 | DOC_START | |
3917 | The cache replacement policy parameter determines which | |
3918 | objects are evicted (replaced) when disk space is needed. | |
3919 | ||
3920 | lru : Squid's original list based LRU policy | |
3921 | heap GDSF : Greedy-Dual Size Frequency | |
3922 | heap LFUDA: Least Frequently Used with Dynamic Aging | |
3923 | heap LRU : LRU policy implemented using a heap | |
3924 | ||
3925 | Applies to any cache_dir lines listed below this directive. | |
3926 | ||
3927 | The LRU policies keeps recently referenced objects. | |
3928 | ||
3929 | The heap GDSF policy optimizes object hit rate by keeping smaller | |
3930 | popular objects in cache so it has a better chance of getting a | |
3931 | hit. It achieves a lower byte hit rate than LFUDA though since | |
3932 | it evicts larger (possibly popular) objects. | |
3933 | ||
3934 | The heap LFUDA policy keeps popular objects in cache regardless of | |
3935 | their size and thus optimizes byte hit rate at the expense of | |
3936 | hit rate since one large, popular object will prevent many | |
3937 | smaller, slightly less popular objects from being cached. | |
3938 | ||
3939 | Both policies utilize a dynamic aging mechanism that prevents | |
3940 | cache pollution that can otherwise occur with frequency-based | |
3941 | replacement policies. | |
3942 | ||
3943 | NOTE: if using the LFUDA replacement policy you should increase | |
3944 | the value of maximum_object_size above its default of 4 MB to | |
3945 | to maximize the potential byte hit rate improvement of LFUDA. | |
3946 | ||
3947 | For more information about the GDSF and LFUDA cache replacement | |
3948 | policies see http://www.hpl.hp.com/techreports/1999/HPL-1999-69.html | |
3949 | and http://fog.hpl.external.hp.com/techreports/98/HPL-98-173.html. | |
3950 | DOC_END | |
3951 | ||
3952 | NAME: minimum_object_size | |
3953 | COMMENT: (bytes) | |
3954 | TYPE: b_int64_t | |
3955 | DEFAULT: 0 KB | |
3956 | DEFAULT_DOC: no limit | |
3957 | LOC: Config.Store.minObjectSize | |
3958 | DOC_START | |
3959 | Objects smaller than this size will NOT be saved on disk. The | |
3960 | value is specified in bytes, and the default is 0 KB, which | |
3961 | means all responses can be stored. | |
3962 | DOC_END | |
3963 | ||
3964 | NAME: maximum_object_size | |
3965 | COMMENT: (bytes) | |
3966 | TYPE: b_int64_t | |
3967 | DEFAULT: 4 MB | |
3968 | LOC: Config.Store.maxObjectSize | |
3969 | DOC_START | |
3970 | Set the default value for max-size parameter on any cache_dir. | |
3971 | The value is specified in bytes, and the default is 4 MB. | |
3972 | ||
3973 | If you wish to get a high BYTES hit ratio, you should probably | |
3974 | increase this (one 32 MB object hit counts for 3200 10KB | |
3975 | hits). | |
3976 | ||
3977 | If you wish to increase hit ratio more than you want to | |
3978 | save bandwidth you should leave this low. | |
3979 | ||
3980 | NOTE: if using the LFUDA replacement policy you should increase | |
3981 | this value to maximize the byte hit rate improvement of LFUDA! | |
3982 | See cache_replacement_policy for a discussion of this policy. | |
3983 | DOC_END | |
3984 | ||
3985 | NAME: cache_dir | |
3986 | TYPE: cachedir | |
3987 | DEFAULT: none | |
3988 | DEFAULT_DOC: No disk cache. Store cache ojects only in memory. | |
3989 | LOC: Config.cacheSwap | |
3990 | DOC_START | |
3991 | Format: | |
3992 | cache_dir Type Directory-Name Fs-specific-data [options] | |
3993 | ||
3994 | You can specify multiple cache_dir lines to spread the | |
3995 | cache among different disk partitions. | |
3996 | ||
3997 | Type specifies the kind of storage system to use. Only "ufs" | |
3998 | is built by default. To enable any of the other storage systems | |
3999 | see the --enable-storeio configure option. | |
4000 | ||
4001 | 'Directory' is a top-level directory where cache swap | |
4002 | files will be stored. If you want to use an entire disk | |
4003 | for caching, this can be the mount-point directory. | |
4004 | The directory must exist and be writable by the Squid | |
4005 | process. Squid will NOT create this directory for you. | |
4006 | ||
4007 | In SMP configurations, cache_dir must not precede the workers option | |
4008 | and should use configuration macros or conditionals to give each | |
4009 | worker interested in disk caching a dedicated cache directory. | |
4010 | ||
4011 | ||
4012 | ==== The ufs store type ==== | |
4013 | ||
4014 | "ufs" is the old well-known Squid storage format that has always | |
4015 | been there. | |
4016 | ||
4017 | Usage: | |
4018 | cache_dir ufs Directory-Name Mbytes L1 L2 [options] | |
4019 | ||
4020 | 'Mbytes' is the amount of disk space (MB) to use under this | |
4021 | directory. The default is 100 MB. Change this to suit your | |
4022 | configuration. Do NOT put the size of your disk drive here. | |
4023 | Instead, if you want Squid to use the entire disk drive, | |
4024 | subtract 20% and use that value. | |
4025 | ||
4026 | 'L1' is the number of first-level subdirectories which | |
4027 | will be created under the 'Directory'. The default is 16. | |
4028 | ||
4029 | 'L2' is the number of second-level subdirectories which | |
4030 | will be created under each first-level directory. The default | |
4031 | is 256. | |
4032 | ||
4033 | ||
4034 | ==== The aufs store type ==== | |
4035 | ||
4036 | "aufs" uses the same storage format as "ufs", utilizing | |
4037 | POSIX-threads to avoid blocking the main Squid process on | |
4038 | disk-I/O. This was formerly known in Squid as async-io. | |
4039 | ||
4040 | Usage: | |
4041 | cache_dir aufs Directory-Name Mbytes L1 L2 [options] | |
4042 | ||
4043 | see argument descriptions under ufs above | |
4044 | ||
4045 | ||
4046 | ==== The diskd store type ==== | |
4047 | ||
4048 | "diskd" uses the same storage format as "ufs", utilizing a | |
4049 | separate process to avoid blocking the main Squid process on | |
4050 | disk-I/O. | |
4051 | ||
4052 | Usage: | |
4053 | cache_dir diskd Directory-Name Mbytes L1 L2 [options] [Q1=n] [Q2=n] | |
4054 | ||
4055 | see argument descriptions under ufs above | |
4056 | ||
4057 | Q1 specifies the number of unacknowledged I/O requests when Squid | |
4058 | stops opening new files. If this many messages are in the queues, | |
4059 | Squid won't open new files. Default is 64 | |
4060 | ||
4061 | Q2 specifies the number of unacknowledged messages when Squid | |
4062 | starts blocking. If this many messages are in the queues, | |
4063 | Squid blocks until it receives some replies. Default is 72 | |
4064 | ||
4065 | When Q1 < Q2 (the default), the cache directory is optimized | |
4066 | for lower response time at the expense of a decrease in hit | |
4067 | ratio. If Q1 > Q2, the cache directory is optimized for | |
4068 | higher hit ratio at the expense of an increase in response | |
4069 | time. | |
4070 | ||
4071 | ||
4072 | ==== The rock store type ==== | |
4073 | ||
4074 | Usage: | |
4075 | cache_dir rock Directory-Name Mbytes [options] | |
4076 | ||
4077 | The Rock Store type is a database-style storage. All cached | |
4078 | entries are stored in a "database" file, using fixed-size slots. | |
4079 | A single entry occupies one or more slots. | |
4080 | ||
4081 | If possible, Squid using Rock Store creates a dedicated kid | |
4082 | process called "disker" to avoid blocking Squid worker(s) on disk | |
4083 | I/O. One disker kid is created for each rock cache_dir. Diskers | |
4084 | are created only when Squid, running in daemon mode, has support | |
4085 | for the IpcIo disk I/O module. | |
4086 | ||
4087 | swap-timeout=msec: Squid will not start writing a miss to or | |
4088 | reading a hit from disk if it estimates that the swap operation | |
4089 | will take more than the specified number of milliseconds. By | |
4090 | default and when set to zero, disables the disk I/O time limit | |
4091 | enforcement. Ignored when using blocking I/O module because | |
4092 | blocking synchronous I/O does not allow Squid to estimate the | |
4093 | expected swap wait time. | |
4094 | ||
4095 | max-swap-rate=swaps/sec: Artificially limits disk access using | |
4096 | the specified I/O rate limit. Swap out requests that | |
4097 | would cause the average I/O rate to exceed the limit are | |
4098 | delayed. Individual swap in requests (i.e., hits or reads) are | |
4099 | not delayed, but they do contribute to measured swap rate and | |
4100 | since they are placed in the same FIFO queue as swap out | |
4101 | requests, they may wait longer if max-swap-rate is smaller. | |
4102 | This is necessary on file systems that buffer "too | |
4103 | many" writes and then start blocking Squid and other processes | |
4104 | while committing those writes to disk. Usually used together | |
4105 | with swap-timeout to avoid excessive delays and queue overflows | |
4106 | when disk demand exceeds available disk "bandwidth". By default | |
4107 | and when set to zero, disables the disk I/O rate limit | |
4108 | enforcement. Currently supported by IpcIo module only. | |
4109 | ||
4110 | slot-size=bytes: The size of a database "record" used for | |
4111 | storing cached responses. A cached response occupies at least | |
4112 | one slot and all database I/O is done using individual slots so | |
4113 | increasing this parameter leads to more disk space waste while | |
4114 | decreasing it leads to more disk I/O overheads. Should be a | |
4115 | multiple of your operating system I/O page size. Defaults to | |
4116 | 16KBytes. A housekeeping header is stored with each slot and | |
4117 | smaller slot-sizes will be rejected. The header is smaller than | |
4118 | 100 bytes. | |
4119 | ||
4120 | ||
4121 | ==== COMMON OPTIONS ==== | |
4122 | ||
4123 | no-store no new objects should be stored to this cache_dir. | |
4124 | ||
4125 | min-size=n the minimum object size in bytes this cache_dir | |
4126 | will accept. It's used to restrict a cache_dir | |
4127 | to only store large objects (e.g. AUFS) while | |
4128 | other stores are optimized for smaller objects | |
4129 | (e.g. Rock). | |
4130 | Defaults to 0. | |
4131 | ||
4132 | max-size=n the maximum object size in bytes this cache_dir | |
4133 | supports. | |
4134 | The value in maximum_object_size directive sets | |
4135 | the default unless more specific details are | |
4136 | available (ie a small store capacity). | |
4137 | ||
4138 | Note: To make optimal use of the max-size limits you should order | |
4139 | the cache_dir lines with the smallest max-size value first. | |
4140 | ||
4141 | NOCOMMENT_START | |
4142 | ||
4143 | # Uncomment and adjust the following to add a disk cache directory. | |
4144 | #cache_dir ufs @DEFAULT_SWAP_DIR@ 100 16 256 | |
4145 | NOCOMMENT_END | |
4146 | DOC_END | |
4147 | ||
4148 | NAME: store_dir_select_algorithm | |
4149 | TYPE: string | |
4150 | LOC: Config.store_dir_select_algorithm | |
4151 | DEFAULT: least-load | |
4152 | DOC_START | |
4153 | How Squid selects which cache_dir to use when the response | |
4154 | object will fit into more than one. | |
4155 | ||
4156 | Regardless of which algorithm is used the cache_dir min-size | |
4157 | and max-size parameters are obeyed. As such they can affect | |
4158 | the selection algorithm by limiting the set of considered | |
4159 | cache_dir. | |
4160 | ||
4161 | Algorithms: | |
4162 | ||
4163 | least-load | |
4164 | ||
4165 | This algorithm is suited to caches with similar cache_dir | |
4166 | sizes and disk speeds. | |
4167 | ||
4168 | The disk with the least I/O pending is selected. | |
4169 | When there are multiple disks with the same I/O load ranking | |
4170 | the cache_dir with most available capacity is selected. | |
4171 | ||
4172 | When a mix of cache_dir sizes are configured the faster disks | |
4173 | have a naturally lower I/O loading and larger disks have more | |
4174 | capacity. So space used to store objects and data throughput | |
4175 | may be very unbalanced towards larger disks. | |
4176 | ||
4177 | ||
4178 | round-robin | |
4179 | ||
4180 | This algorithm is suited to caches with unequal cache_dir | |
4181 | disk sizes. | |
4182 | ||
4183 | Each cache_dir is selected in a rotation. The next suitable | |
4184 | cache_dir is used. | |
4185 | ||
4186 | Available cache_dir capacity is only considered in relation | |
4187 | to whether the object will fit and meets the min-size and | |
4188 | max-size parameters. | |
4189 | ||
4190 | Disk I/O loading is only considered to prevent overload on slow | |
4191 | disks. This algorithm does not spread objects by size, so any | |
4192 | I/O loading per-disk may appear very unbalanced and volatile. | |
4193 | ||
4194 | If several cache_dirs use similar min-size, max-size, or other | |
4195 | limits to to reject certain responses, then do not group such | |
4196 | cache_dir lines together, to avoid round-robin selection bias | |
4197 | towards the first cache_dir after the group. Instead, interleave | |
4198 | cache_dir lines from different groups. For example: | |
4199 | ||
4200 | store_dir_select_algorithm round-robin | |
4201 | cache_dir rock /hdd1 ... min-size=100000 | |
4202 | cache_dir rock /ssd1 ... max-size=99999 | |
4203 | cache_dir rock /hdd2 ... min-size=100000 | |
4204 | cache_dir rock /ssd2 ... max-size=99999 | |
4205 | cache_dir rock /hdd3 ... min-size=100000 | |
4206 | cache_dir rock /ssd3 ... max-size=99999 | |
4207 | DOC_END | |
4208 | ||
4209 | NAME: max_open_disk_fds | |
4210 | TYPE: int | |
4211 | LOC: Config.max_open_disk_fds | |
4212 | DEFAULT: 0 | |
4213 | DEFAULT_DOC: no limit | |
4214 | DOC_START | |
4215 | To avoid having disk as the I/O bottleneck Squid can optionally | |
4216 | bypass the on-disk cache if more than this amount of disk file | |
4217 | descriptors are open. | |
4218 | ||
4219 | A value of 0 indicates no limit. | |
4220 | DOC_END | |
4221 | ||
4222 | NAME: cache_swap_low | |
4223 | COMMENT: (percent, 0-100) | |
4224 | TYPE: int | |
4225 | DEFAULT: 90 | |
4226 | LOC: Config.Swap.lowWaterMark | |
4227 | DOC_START | |
4228 | The low-water mark for AUFS/UFS/diskd cache object eviction by | |
4229 | the cache_replacement_policy algorithm. | |
4230 | ||
4231 | Removal begins when the swap (disk) usage of a cache_dir is | |
4232 | above this low-water mark and attempts to maintain utilization | |
4233 | near the low-water mark. | |
4234 | ||
4235 | As swap utilization increases towards the high-water mark set | |
4236 | by cache_swap_high object eviction becomes more agressive. | |
4237 | ||
4238 | The value difference in percentages between low- and high-water | |
4239 | marks represent an eviction rate of 300 objects per second and | |
4240 | the rate continues to scale in agressiveness by multiples of | |
4241 | this above the high-water mark. | |
4242 | ||
4243 | Defaults are 90% and 95%. If you have a large cache, 5% could be | |
4244 | hundreds of MB. If this is the case you may wish to set these | |
4245 | numbers closer together. | |
4246 | ||
4247 | See also cache_swap_high and cache_replacement_policy | |
4248 | DOC_END | |
4249 | ||
4250 | NAME: cache_swap_high | |
4251 | COMMENT: (percent, 0-100) | |
4252 | TYPE: int | |
4253 | DEFAULT: 95 | |
4254 | LOC: Config.Swap.highWaterMark | |
4255 | DOC_START | |
4256 | The high-water mark for AUFS/UFS/diskd cache object eviction by | |
4257 | the cache_replacement_policy algorithm. | |
4258 | ||
4259 | Removal begins when the swap (disk) usage of a cache_dir is | |
4260 | above the low-water mark set by cache_swap_low and attempts to | |
4261 | maintain utilization near the low-water mark. | |
4262 | ||
4263 | As swap utilization increases towards this high-water mark object | |
4264 | eviction becomes more agressive. | |
4265 | ||
4266 | The value difference in percentages between low- and high-water | |
4267 | marks represent an eviction rate of 300 objects per second and | |
4268 | the rate continues to scale in agressiveness by multiples of | |
4269 | this above the high-water mark. | |
4270 | ||
4271 | Defaults are 90% and 95%. If you have a large cache, 5% could be | |
4272 | hundreds of MB. If this is the case you may wish to set these | |
4273 | numbers closer together. | |
4274 | ||
4275 | See also cache_swap_low and cache_replacement_policy | |
4276 | DOC_END | |
4277 | ||
4278 | COMMENT_START | |
4279 | LOGFILE OPTIONS | |
4280 | ----------------------------------------------------------------------------- | |
4281 | COMMENT_END | |
4282 | ||
4283 | NAME: logformat | |
4284 | TYPE: logformat | |
4285 | LOC: Log::TheConfig | |
4286 | DEFAULT: none | |
4287 | DEFAULT_DOC: The format definitions squid, common, combined, referrer, useragent are built in. | |
4288 | DOC_START | |
4289 | Usage: | |
4290 | ||
4291 | logformat <name> <format specification> | |
4292 | ||
4293 | Defines an access log format. | |
4294 | ||
4295 | The <format specification> is a string with embedded % format codes | |
4296 | ||
4297 | % format codes all follow the same basic structure where all but | |
4298 | the formatcode is optional. Output strings are automatically escaped | |
4299 | as required according to their context and the output format | |
4300 | modifiers are usually not needed, but can be specified if an explicit | |
4301 | output format is desired. | |
4302 | ||
4303 | % ["|[|'|#|/] [-] [[0]width] [{arg}] formatcode [{arg}] | |
4304 | ||
4305 | " output in quoted string format | |
4306 | [ output in squid text log format as used by log_mime_hdrs | |
4307 | # output in URL quoted format | |
4308 | / output in shell \-escaped format | |
4309 | ' output as-is | |
4310 | ||
4311 | - left aligned | |
4312 | ||
4313 | width minimum and/or maximum field width: | |
4314 | [width_min][.width_max] | |
4315 | When minimum starts with 0, the field is zero-padded. | |
4316 | String values exceeding maximum width are truncated. | |
4317 | ||
4318 | {arg} argument such as header name etc. This field may be | |
4319 | placed before or after the token, but not both at once. | |
4320 | ||
4321 | Format codes: | |
4322 | ||
4323 | % a literal % character | |
4324 | sn Unique sequence number per log line entry | |
4325 | err_code The ID of an error response served by Squid or | |
4326 | a similar internal error identifier. | |
4327 | err_detail Additional err_code-dependent error information. | |
4328 | note The annotation specified by the argument. Also | |
4329 | logs the adaptation meta headers set by the | |
4330 | adaptation_meta configuration parameter. | |
4331 | If no argument given all annotations logged. | |
4332 | The argument may include a separator to use with | |
4333 | annotation values: | |
4334 | name[:separator] | |
4335 | By default, multiple note values are separated with "," | |
4336 | and multiple notes are separated with "\r\n". | |
4337 | When logging named notes with %{name}note, the | |
4338 | explicitly configured separator is used between note | |
4339 | values. When logging all notes with %note, the | |
4340 | explicitly configured separator is used between | |
4341 | individual notes. There is currently no way to | |
4342 | specify both value and notes separators when logging | |
4343 | all notes with %note. | |
4344 | ||
4345 | Connection related format codes: | |
4346 | ||
4347 | >a Client source IP address | |
4348 | >A Client FQDN | |
4349 | >p Client source port | |
4350 | >eui Client source EUI (MAC address, EUI-48 or EUI-64 identifier) | |
4351 | >la Local IP address the client connected to | |
4352 | >lp Local port number the client connected to | |
4353 | >qos Client connection TOS/DSCP value set by Squid | |
4354 | >nfmark Client connection netfilter mark set by Squid | |
4355 | ||
4356 | la Local listening IP address the client connection was connected to. | |
4357 | lp Local listening port number the client connection was connected to. | |
4358 | ||
4359 | <a Server IP address of the last server or peer connection | |
4360 | <A Server FQDN or peer name | |
4361 | <p Server port number of the last server or peer connection | |
4362 | <la Local IP address of the last server or peer connection | |
4363 | <lp Local port number of the last server or peer connection | |
4364 | <qos Server connection TOS/DSCP value set by Squid | |
4365 | <nfmark Server connection netfilter mark set by Squid | |
4366 | ||
4367 | Time related format codes: | |
4368 | ||
4369 | ts Seconds since epoch | |
4370 | tu subsecond time (milliseconds) | |
4371 | tl Local time. Optional strftime format argument | |
4372 | default %d/%b/%Y:%H:%M:%S %z | |
4373 | tg GMT time. Optional strftime format argument | |
4374 | default %d/%b/%Y:%H:%M:%S %z | |
4375 | tr Response time (milliseconds) | |
4376 | dt Total time spent making DNS lookups (milliseconds) | |
4377 | tS Approximate master transaction start time in | |
4378 | <full seconds since epoch>.<fractional seconds> format. | |
4379 | Currently, Squid considers the master transaction | |
4380 | started when a complete HTTP request header initiating | |
4381 | the transaction is received from the client. This is | |
4382 | the same value that Squid uses to calculate transaction | |
4383 | response time when logging %tr to access.log. Currently, | |
4384 | Squid uses millisecond resolution for %tS values, | |
4385 | similar to the default access.log "current time" field | |
4386 | (%ts.%03tu). | |
4387 | ||
4388 | Access Control related format codes: | |
4389 | ||
4390 | et Tag returned by external acl | |
4391 | ea Log string returned by external acl | |
4392 | un User name (any available) | |
4393 | ul User name from authentication | |
4394 | ue User name from external acl helper | |
4395 | ui User name from ident | |
4396 | un A user name. Expands to the first available name | |
4397 | from the following list of information sources: | |
4398 | - authenticated user name, like %ul | |
4399 | - user name supplied by an external ACL, like %ue | |
4400 | - SSL client name, like %us | |
4401 | - ident user name, like %ui | |
4402 | credentials Client credentials. The exact meaning depends on | |
4403 | the authentication scheme: For Basic authentication, | |
4404 | it is the password; for Digest, the realm sent by the | |
4405 | client; for NTLM and Negotiate, the client challenge | |
4406 | or client credentials prefixed with "YR " or "KK ". | |
4407 | ||
4408 | HTTP related format codes: | |
4409 | ||
4410 | REQUEST | |
4411 | ||
4412 | [http::]rm Request method (GET/POST etc) | |
4413 | [http::]>rm Request method from client | |
4414 | [http::]<rm Request method sent to server or peer | |
4415 | [http::]ru Request URL from client (historic, filtered for logging) | |
4416 | [http::]>ru Request URL from client | |
4417 | [http::]<ru Request URL sent to server or peer | |
4418 | [http::]>rs Request URL scheme from client | |
4419 | [http::]<rs Request URL scheme sent to server or peer | |
4420 | [http::]>rd Request URL domain from client | |
4421 | [http::]<rd Request URL domain sent to server or peer | |
4422 | [http::]>rP Request URL port from client | |
4423 | [http::]<rP Request URL port sent to server or peer | |
4424 | [http::]rp Request URL path excluding hostname | |
4425 | [http::]>rp Request URL path excluding hostname from client | |
4426 | [http::]<rp Request URL path excluding hostname sent to server or peer | |
4427 | [http::]rv Request protocol version | |
4428 | [http::]>rv Request protocol version from client | |
4429 | [http::]<rv Request protocol version sent to server or peer | |
4430 | ||
4431 | [http::]>h Original received request header. | |
4432 | Usually differs from the request header sent by | |
4433 | Squid, although most fields are often preserved. | |
4434 | Accepts optional header field name/value filter | |
4435 | argument using name[:[separator]element] format. | |
4436 | [http::]>ha Received request header after adaptation and | |
4437 | redirection (pre-cache REQMOD vectoring point). | |
4438 | Usually differs from the request header sent by | |
4439 | Squid, although most fields are often preserved. | |
4440 | Optional header name argument as for >h | |
4441 | ||
4442 | RESPONSE | |
4443 | ||
4444 | [http::]<Hs HTTP status code received from the next hop | |
4445 | [http::]>Hs HTTP status code sent to the client | |
4446 | ||
4447 | [http::]<h Reply header. Optional header name argument | |
4448 | as for >h | |
4449 | ||
4450 | [http::]mt MIME content type | |
4451 | ||
4452 | ||
4453 | SIZE COUNTERS | |
4454 | ||
4455 | [http::]st Total size of request + reply traffic with client | |
4456 | [http::]>st Total size of request received from client. | |
4457 | Excluding chunked encoding bytes. | |
4458 | [http::]<st Total size of reply sent to client (after adaptation) | |
4459 | ||
4460 | [http::]>sh Size of request headers received from client | |
4461 | [http::]<sh Size of reply headers sent to client (after adaptation) | |
4462 | ||
4463 | [http::]<sH Reply high offset sent | |
4464 | [http::]<sS Upstream object size | |
4465 | ||
4466 | [http::]<bs Number of HTTP-equivalent message body bytes | |
4467 | received from the next hop, excluding chunked | |
4468 | transfer encoding and control messages. | |
4469 | Generated FTP/Gopher listings are treated as | |
4470 | received bodies. | |
4471 | ||
4472 | TIMING | |
4473 | ||
4474 | [http::]<pt Peer response time in milliseconds. The timer starts | |
4475 | when the last request byte is sent to the next hop | |
4476 | and stops when the last response byte is received. | |
4477 | [http::]<tt Total time in milliseconds. The timer | |
4478 | starts with the first connect request (or write I/O) | |
4479 | sent to the first selected peer. The timer stops | |
4480 | with the last I/O with the last peer. | |
4481 | ||
4482 | Squid handling related format codes: | |
4483 | ||
4484 | Ss Squid request status (TCP_MISS etc) | |
4485 | Sh Squid hierarchy status (DEFAULT_PARENT etc) | |
4486 | ||
4487 | SSL-related format codes: | |
4488 | ||
4489 | ssl::bump_mode SslBump decision for the transaction: | |
4490 | ||
4491 | For CONNECT requests that initiated bumping of | |
4492 | a connection and for any request received on | |
4493 | an already bumped connection, Squid logs the | |
4494 | corresponding SslBump mode ("splice", "bump", | |
4495 | "peek", "stare", "terminate", "server-first" | |
4496 | or "client-first"). See the ssl_bump option | |
4497 | for more information about these modes. | |
4498 | ||
4499 | A "none" token is logged for requests that | |
4500 | triggered "ssl_bump" ACL evaluation matching | |
4501 | a "none" rule. | |
4502 | ||
4503 | In all other cases, a single dash ("-") is | |
4504 | logged. | |
4505 | ||
4506 | ssl::>sni SSL client SNI sent to Squid. | |
4507 | ||
4508 | ssl::>cert_subject | |
4509 | The Subject field of the received client | |
4510 | SSL certificate or a dash ('-') if Squid has | |
4511 | received an invalid/malformed certificate or | |
4512 | no certificate at all. Consider encoding the | |
4513 | logged value because Subject often has spaces. | |
4514 | ||
4515 | ssl::>cert_issuer | |
4516 | The Issuer field of the received client | |
4517 | SSL certificate or a dash ('-') if Squid has | |
4518 | received an invalid/malformed certificate or | |
4519 | no certificate at all. Consider encoding the | |
4520 | logged value because Issuer often has spaces. | |
4521 | ||
4522 | ssl::<cert_errors | |
4523 | The list of certificate validation errors | |
4524 | detected by Squid (including OpenSSL and | |
4525 | certificate validation helper components). The | |
4526 | errors are listed in the discovery order. By | |
4527 | default, the error codes are separated by ':'. | |
4528 | Accepts an optional separator argument. | |
4529 | ||
4530 | %ssl::>negotiated_version The negotiated TLS version of the | |
4531 | client connection. | |
4532 | ||
4533 | %ssl::<negotiated_version The negotiated TLS version of the | |
4534 | last server or peer connection. | |
4535 | ||
4536 | %ssl::>received_hello_version The TLS version of the Hello | |
4537 | message received from TLS client. | |
4538 | ||
4539 | %ssl::<received_hello_version The TLS version of the Hello | |
4540 | message received from TLS server. | |
4541 | ||
4542 | %ssl::>received_supported_version The maximum TLS version | |
4543 | supported by the TLS client. | |
4544 | ||
4545 | %ssl::<received_supported_version The maximum TLS version | |
4546 | supported by the TLS server. | |
4547 | ||
4548 | %ssl::>negotiated_cipher The negotiated cipher of the | |
4549 | client connection. | |
4550 | ||
4551 | %ssl::<negotiated_cipher The negotiated cipher of the | |
4552 | last server or peer connection. | |
4553 | ||
4554 | If ICAP is enabled, the following code becomes available (as | |
4555 | well as ICAP log codes documented with the icap_log option): | |
4556 | ||
4557 | icap::tt Total ICAP processing time for the HTTP | |
4558 | transaction. The timer ticks when ICAP | |
4559 | ACLs are checked and when ICAP | |
4560 | transaction is in progress. | |
4561 | ||
4562 | If adaptation is enabled the following codes become available: | |
4563 | ||
4564 | adapt::<last_h The header of the last ICAP response or | |
4565 | meta-information from the last eCAP | |
4566 | transaction related to the HTTP transaction. | |
4567 | Like <h, accepts an optional header name | |
4568 | argument. | |
4569 | ||
4570 | adapt::sum_trs Summed adaptation transaction response | |
4571 | times recorded as a comma-separated list in | |
4572 | the order of transaction start time. Each time | |
4573 | value is recorded as an integer number, | |
4574 | representing response time of one or more | |
4575 | adaptation (ICAP or eCAP) transaction in | |
4576 | milliseconds. When a failed transaction is | |
4577 | being retried or repeated, its time is not | |
4578 | logged individually but added to the | |
4579 | replacement (next) transaction. See also: | |
4580 | adapt::all_trs. | |
4581 | ||
4582 | adapt::all_trs All adaptation transaction response times. | |
4583 | Same as adaptation_strs but response times of | |
4584 | individual transactions are never added | |
4585 | together. Instead, all transaction response | |
4586 | times are recorded individually. | |
4587 | ||
4588 | You can prefix adapt::*_trs format codes with adaptation | |
4589 | service name in curly braces to record response time(s) specific | |
4590 | to that service. For example: %{my_service}adapt::sum_trs | |
4591 | ||
4592 | The default formats available (which do not need re-defining) are: | |
4593 | ||
4594 | logformat squid %ts.%03tu %6tr %>a %Ss/%03>Hs %<st %rm %ru %[un %Sh/%<a %mt | |
4595 | logformat common %>a %[ui %[un [%tl] "%rm %ru HTTP/%rv" %>Hs %<st %Ss:%Sh | |
4596 | logformat combined %>a %[ui %[un [%tl] "%rm %ru HTTP/%rv" %>Hs %<st "%{Referer}>h" "%{User-Agent}>h" %Ss:%Sh | |
4597 | logformat referrer %ts.%03tu %>a %{Referer}>h %ru | |
4598 | logformat useragent %>a [%tl] "%{User-Agent}>h" | |
4599 | ||
4600 | NOTE: When the log_mime_hdrs directive is set to ON. | |
4601 | The squid, common and combined formats have a safely encoded copy | |
4602 | of the mime headers appended to each line within a pair of brackets. | |
4603 | ||
4604 | NOTE: The common and combined formats are not quite true to the Apache definition. | |
4605 | The logs from Squid contain an extra status and hierarchy code appended. | |
4606 | ||
4607 | DOC_END | |
4608 | ||
4609 | NAME: access_log cache_access_log | |
4610 | TYPE: access_log | |
4611 | LOC: Config.Log.accesslogs | |
4612 | DEFAULT_IF_NONE: daemon:@DEFAULT_ACCESS_LOG@ squid | |
4613 | DOC_START | |
4614 | Configures whether and how Squid logs HTTP and ICP transactions. | |
4615 | If access logging is enabled, a single line is logged for every | |
4616 | matching HTTP or ICP request. The recommended directive formats are: | |
4617 | ||
4618 | access_log <module>:<place> [option ...] [acl acl ...] | |
4619 | access_log none [acl acl ...] | |
4620 | ||
4621 | The following directive format is accepted but may be deprecated: | |
4622 | access_log <module>:<place> [<logformat name> [acl acl ...]] | |
4623 | ||
4624 | In most cases, the first ACL name must not contain the '=' character | |
4625 | and should not be equal to an existing logformat name. You can always | |
4626 | start with an 'all' ACL to work around those restrictions. | |
4627 | ||
4628 | Will log to the specified module:place using the specified format (which | |
4629 | must be defined in a logformat directive) those entries which match | |
4630 | ALL the acl's specified (which must be defined in acl clauses). | |
4631 | If no acl is specified, all requests will be logged to this destination. | |
4632 | ||
4633 | ===== Available options for the recommended directive format ===== | |
4634 | ||
4635 | logformat=name Names log line format (either built-in or | |
4636 | defined by a logformat directive). Defaults | |
4637 | to 'squid'. | |
4638 | ||
4639 | buffer-size=64KB Defines approximate buffering limit for log | |
4640 | records (see buffered_logs). Squid should not | |
4641 | keep more than the specified size and, hence, | |
4642 | should flush records before the buffer becomes | |
4643 | full to avoid overflows under normal | |
4644 | conditions (the exact flushing algorithm is | |
4645 | module-dependent though). The on-error option | |
4646 | controls overflow handling. | |
4647 | ||
4648 | on-error=die|drop Defines action on unrecoverable errors. The | |
4649 | 'drop' action ignores (i.e., does not log) | |
4650 | affected log records. The default 'die' action | |
4651 | kills the affected worker. The drop action | |
4652 | support has not been tested for modules other | |
4653 | than tcp. | |
4654 | ||
4655 | rotate=N Specifies the number of log file rotations to | |
4656 | make when you run 'squid -k rotate'. The default | |
4657 | is to obey the logfile_rotate directive. Setting | |
4658 | rotate=0 will disable the file name rotation, | |
4659 | but the log files are still closed and re-opened. | |
4660 | This will enable you to rename the logfiles | |
4661 | yourself just before sending the rotate signal. | |
4662 | Only supported by the stdio module. | |
4663 | ||
4664 | ===== Modules Currently available ===== | |
4665 | ||
4666 | none Do not log any requests matching these ACL. | |
4667 | Do not specify Place or logformat name. | |
4668 | ||
4669 | stdio Write each log line to disk immediately at the completion of | |
4670 | each request. | |
4671 | Place: the filename and path to be written. | |
4672 | ||
4673 | daemon Very similar to stdio. But instead of writing to disk the log | |
4674 | line is passed to a daemon helper for asychronous handling instead. | |
4675 | Place: varies depending on the daemon. | |
4676 | ||
4677 | log_file_daemon Place: the file name and path to be written. | |
4678 | ||
4679 | syslog To log each request via syslog facility. | |
4680 | Place: The syslog facility and priority level for these entries. | |
4681 | Place Format: facility.priority | |
4682 | ||
4683 | where facility could be any of: | |
4684 | authpriv, daemon, local0 ... local7 or user. | |
4685 | ||
4686 | And priority could be any of: | |
4687 | err, warning, notice, info, debug. | |
4688 | ||
4689 | udp To send each log line as text data to a UDP receiver. | |
4690 | Place: The destination host name or IP and port. | |
4691 | Place Format: //host:port | |
4692 | ||
4693 | tcp To send each log line as text data to a TCP receiver. | |
4694 | Lines may be accumulated before sending (see buffered_logs). | |
4695 | Place: The destination host name or IP and port. | |
4696 | Place Format: //host:port | |
4697 | ||
4698 | Default: | |
4699 | access_log daemon:@DEFAULT_ACCESS_LOG@ squid | |
4700 | DOC_END | |
4701 | ||
4702 | NAME: icap_log | |
4703 | TYPE: access_log | |
4704 | IFDEF: ICAP_CLIENT | |
4705 | LOC: Config.Log.icaplogs | |
4706 | DEFAULT: none | |
4707 | DOC_START | |
4708 | ICAP log files record ICAP transaction summaries, one line per | |
4709 | transaction. | |
4710 | ||
4711 | The icap_log option format is: | |
4712 | icap_log <filepath> [<logformat name> [acl acl ...]] | |
4713 | icap_log none [acl acl ...]] | |
4714 | ||
4715 | Please see access_log option documentation for details. The two | |
4716 | kinds of logs share the overall configuration approach and many | |
4717 | features. | |
4718 | ||
4719 | ICAP processing of a single HTTP message or transaction may | |
4720 | require multiple ICAP transactions. In such cases, multiple | |
4721 | ICAP transaction log lines will correspond to a single access | |
4722 | log line. | |
4723 | ||
4724 | ICAP log supports many access.log logformat %codes. In ICAP context, | |
4725 | HTTP message-related %codes are applied to the HTTP message embedded | |
4726 | in an ICAP message. Logformat "%http::>..." codes are used for HTTP | |
4727 | messages embedded in ICAP requests while "%http::<..." codes are used | |
4728 | for HTTP messages embedded in ICAP responses. For example: | |
4729 | ||
4730 | http::>h To-be-adapted HTTP message headers sent by Squid to | |
4731 | the ICAP service. For REQMOD transactions, these are | |
4732 | HTTP request headers. For RESPMOD, these are HTTP | |
4733 | response headers, but Squid currently cannot log them | |
4734 | (i.e., %http::>h will expand to "-" for RESPMOD). | |
4735 | ||
4736 | http::<h Adapted HTTP message headers sent by the ICAP | |
4737 | service to Squid (i.e., HTTP request headers in regular | |
4738 | REQMOD; HTTP response headers in RESPMOD and during | |
4739 | request satisfaction in REQMOD). | |
4740 | ||
4741 | ICAP OPTIONS transactions do not embed HTTP messages. | |
4742 | ||
4743 | Several logformat codes below deal with ICAP message bodies. An ICAP | |
4744 | message body, if any, typically includes a complete HTTP message | |
4745 | (required HTTP headers plus optional HTTP message body). When | |
4746 | computing HTTP message body size for these logformat codes, Squid | |
4747 | either includes or excludes chunked encoding overheads; see | |
4748 | code-specific documentation for details. | |
4749 | ||
4750 | For Secure ICAP services, all size-related information is currently | |
4751 | computed before/after TLS encryption/decryption, as if TLS was not | |
4752 | in use at all. | |
4753 | ||
4754 | The following format codes are also available for ICAP logs: | |
4755 | ||
4756 | icap::<A ICAP server IP address. Similar to <A. | |
4757 | ||
4758 | icap::<service_name ICAP service name from the icap_service | |
4759 | option in Squid configuration file. | |
4760 | ||
4761 | icap::ru ICAP Request-URI. Similar to ru. | |
4762 | ||
4763 | icap::rm ICAP request method (REQMOD, RESPMOD, or | |
4764 | OPTIONS). Similar to existing rm. | |
4765 | ||
4766 | icap::>st The total size of the ICAP request sent to the ICAP | |
4767 | server (ICAP headers + ICAP body), including chunking | |
4768 | metadata (if any). | |
4769 | ||
4770 | icap::<st The total size of the ICAP response received from the | |
4771 | ICAP server (ICAP headers + ICAP body), including | |
4772 | chunking metadata (if any). | |
4773 | ||
4774 | icap::<bs The size of the ICAP response body received from the | |
4775 | ICAP server, excluding chunking metadata (if any). | |
4776 | ||
4777 | icap::tr Transaction response time (in | |
4778 | milliseconds). The timer starts when | |
4779 | the ICAP transaction is created and | |
4780 | stops when the transaction is completed. | |
4781 | Similar to tr. | |
4782 | ||
4783 | icap::tio Transaction I/O time (in milliseconds). The | |
4784 | timer starts when the first ICAP request | |
4785 | byte is scheduled for sending. The timers | |
4786 | stops when the last byte of the ICAP response | |
4787 | is received. | |
4788 | ||
4789 | icap::to Transaction outcome: ICAP_ERR* for all | |
4790 | transaction errors, ICAP_OPT for OPTION | |
4791 | transactions, ICAP_ECHO for 204 | |
4792 | responses, ICAP_MOD for message | |
4793 | modification, and ICAP_SAT for request | |
4794 | satisfaction. Similar to Ss. | |
4795 | ||
4796 | icap::Hs ICAP response status code. Similar to Hs. | |
4797 | ||
4798 | icap::>h ICAP request header(s). Similar to >h. | |
4799 | ||
4800 | icap::<h ICAP response header(s). Similar to <h. | |
4801 | ||
4802 | The default ICAP log format, which can be used without an explicit | |
4803 | definition, is called icap_squid: | |
4804 | ||
4805 | logformat icap_squid %ts.%03tu %6icap::tr %>A %icap::to/%03icap::Hs %icap::<st %icap::rm %icap::ru %un -/%icap::<A - | |
4806 | ||
4807 | See also: logformat and %adapt::<last_h | |
4808 | DOC_END | |
4809 | ||
4810 | NAME: logfile_daemon | |
4811 | TYPE: string | |
4812 | DEFAULT: @DEFAULT_LOGFILED@ | |
4813 | LOC: Log::TheConfig.logfile_daemon | |
4814 | DOC_START | |
4815 | Specify the path to the logfile-writing daemon. This daemon is | |
4816 | used to write the access and store logs, if configured. | |
4817 | ||
4818 | Squid sends a number of commands to the log daemon: | |
4819 | L<data>\n - logfile data | |
4820 | R\n - rotate file | |
4821 | T\n - truncate file | |
4822 | O\n - reopen file | |
4823 | F\n - flush file | |
4824 | r<n>\n - set rotate count to <n> | |
4825 | b<n>\n - 1 = buffer output, 0 = don't buffer output | |
4826 | ||
4827 | No responses is expected. | |
4828 | DOC_END | |
4829 | ||
4830 | NAME: stats_collection | |
4831 | TYPE: acl_access | |
4832 | LOC: Config.accessList.stats_collection | |
4833 | DEFAULT: none | |
4834 | DEFAULT_DOC: Allow logging for all transactions. | |
4835 | COMMENT: allow|deny acl acl... | |
4836 | DOC_START | |
4837 | This options allows you to control which requests gets accounted | |
4838 | in performance counters. | |
4839 | ||
4840 | This clause only supports fast acl types. | |
4841 | See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details. | |
4842 | DOC_END | |
4843 | ||
4844 | NAME: cache_store_log | |
4845 | TYPE: string | |
4846 | DEFAULT: none | |
4847 | LOC: Config.Log.store | |
4848 | DOC_START | |
4849 | Logs the activities of the storage manager. Shows which | |
4850 | objects are ejected from the cache, and which objects are | |
4851 | saved and for how long. | |
4852 | There are not really utilities to analyze this data, so you can safely | |
4853 | disable it (the default). | |
4854 | ||
4855 | Store log uses modular logging outputs. See access_log for the list | |
4856 | of modules supported. | |
4857 | ||
4858 | Example: | |
4859 | cache_store_log stdio:@DEFAULT_STORE_LOG@ | |
4860 | cache_store_log daemon:@DEFAULT_STORE_LOG@ | |
4861 | DOC_END | |
4862 | ||
4863 | NAME: cache_swap_state cache_swap_log | |
4864 | TYPE: string | |
4865 | LOC: Config.Log.swap | |
4866 | DEFAULT: none | |
4867 | DEFAULT_DOC: Store the journal inside its cache_dir | |
4868 | DOC_START | |
4869 | Location for the cache "swap.state" file. This index file holds | |
4870 | the metadata of objects saved on disk. It is used to rebuild | |
4871 | the cache during startup. Normally this file resides in each | |
4872 | 'cache_dir' directory, but you may specify an alternate | |
4873 | pathname here. Note you must give a full filename, not just | |
4874 | a directory. Since this is the index for the whole object | |
4875 | list you CANNOT periodically rotate it! | |
4876 | ||
4877 | If %s can be used in the file name it will be replaced with a | |
4878 | a representation of the cache_dir name where each / is replaced | |
4879 | with '.'. This is needed to allow adding/removing cache_dir | |
4880 | lines when cache_swap_log is being used. | |
4881 | ||
4882 | If have more than one 'cache_dir', and %s is not used in the name | |
4883 | these swap logs will have names such as: | |
4884 | ||
4885 | cache_swap_log.00 | |
4886 | cache_swap_log.01 | |
4887 | cache_swap_log.02 | |
4888 | ||
4889 | The numbered extension (which is added automatically) | |
4890 | corresponds to the order of the 'cache_dir' lines in this | |
4891 | configuration file. If you change the order of the 'cache_dir' | |
4892 | lines in this file, these index files will NOT correspond to | |
4893 | the correct 'cache_dir' entry (unless you manually rename | |
4894 | them). We recommend you do NOT use this option. It is | |
4895 | better to keep these index files in each 'cache_dir' directory. | |
4896 | DOC_END | |
4897 | ||
4898 | NAME: logfile_rotate | |
4899 | TYPE: int | |
4900 | DEFAULT: 10 | |
4901 | LOC: Config.Log.rotateNumber | |
4902 | DOC_START | |
4903 | Specifies the default number of logfile rotations to make when you | |
4904 | type 'squid -k rotate'. The default is 10, which will rotate | |
4905 | with extensions 0 through 9. Setting logfile_rotate to 0 will | |
4906 | disable the file name rotation, but the logfiles are still closed | |
4907 | and re-opened. This will enable you to rename the logfiles | |
4908 | yourself just before sending the rotate signal. | |
4909 | ||
4910 | Note, from Squid-3.1 this option is only a default for cache.log, | |
4911 | that log can be rotated separately by using debug_options. | |
4912 | ||
4913 | Note, from Squid-3.6 this option is only a default for access.log | |
4914 | recorded by stdio: module. Those logs can be rotated separately by | |
4915 | using the rotate=N option on their access_log directive. | |
4916 | ||
4917 | Note, the 'squid -k rotate' command normally sends a USR1 | |
4918 | signal to the running squid process. In certain situations | |
4919 | (e.g. on Linux with Async I/O), USR1 is used for other | |
4920 | purposes, so -k rotate uses another signal. It is best to get | |
4921 | in the habit of using 'squid -k rotate' instead of 'kill -USR1 | |
4922 | <pid>'. | |
4923 | ||
4924 | DOC_END | |
4925 | ||
4926 | NAME: mime_table | |
4927 | TYPE: string | |
4928 | DEFAULT: @DEFAULT_MIME_TABLE@ | |
4929 | LOC: Config.mimeTablePathname | |
4930 | DOC_START | |
4931 | Path to Squid's icon configuration file. | |
4932 | ||
4933 | You shouldn't need to change this, but the default file contains | |
4934 | examples and formatting information if you do. | |
4935 | DOC_END | |
4936 | ||
4937 | NAME: log_mime_hdrs | |
4938 | COMMENT: on|off | |
4939 | TYPE: onoff | |
4940 | LOC: Config.onoff.log_mime_hdrs | |
4941 | DEFAULT: off | |
4942 | DOC_START | |
4943 | The Cache can record both the request and the response MIME | |
4944 | headers for each HTTP transaction. The headers are encoded | |
4945 | safely and will appear as two bracketed fields at the end of | |
4946 | the access log (for either the native or httpd-emulated log | |
4947 | formats). To enable this logging set log_mime_hdrs to 'on'. | |
4948 | DOC_END | |
4949 | ||
4950 | NAME: pid_filename | |
4951 | TYPE: string | |
4952 | DEFAULT: @DEFAULT_PID_FILE@ | |
4953 | LOC: Config.pidFilename | |
4954 | DOC_START | |
4955 | A filename to write the process-id to. To disable, enter "none". | |
4956 | DOC_END | |
4957 | ||
4958 | NAME: client_netmask | |
4959 | TYPE: address | |
4960 | LOC: Config.Addrs.client_netmask | |
4961 | DEFAULT: no_addr | |
4962 | DEFAULT_DOC: Log full client IP address | |
4963 | DOC_START | |
4964 | A netmask for client addresses in logfiles and cachemgr output. | |
4965 | Change this to protect the privacy of your cache clients. | |
4966 | A netmask of 255.255.255.0 will log all IP's in that range with | |
4967 | the last digit set to '0'. | |
4968 | DOC_END | |
4969 | ||
4970 | NAME: strip_query_terms | |
4971 | TYPE: onoff | |
4972 | LOC: Config.onoff.strip_query_terms | |
4973 | DEFAULT: on | |
4974 | DOC_START | |
4975 | By default, Squid strips query terms from requested URLs before | |
4976 | logging. This protects your user's privacy and reduces log size. | |
4977 | ||
4978 | When investigating HIT/MISS or other caching behaviour you | |
4979 | will need to disable this to see the full URL used by Squid. | |
4980 | DOC_END | |
4981 | ||
4982 | NAME: buffered_logs | |
4983 | COMMENT: on|off | |
4984 | TYPE: onoff | |
4985 | DEFAULT: off | |
4986 | LOC: Config.onoff.buffered_logs | |
4987 | DOC_START | |
4988 | Whether to write/send access_log records ASAP or accumulate them and | |
4989 | then write/send them in larger chunks. Buffering may improve | |
4990 | performance because it decreases the number of I/Os. However, | |
4991 | buffering increases the delay before log records become available to | |
4992 | the final recipient (e.g., a disk file or logging daemon) and, | |
4993 | hence, increases the risk of log records loss. | |
4994 | ||
4995 | Note that even when buffered_logs are off, Squid may have to buffer | |
4996 | records if it cannot write/send them immediately due to pending I/Os | |
4997 | (e.g., the I/O writing the previous log record) or connectivity loss. | |
4998 | ||
4999 | Currently honored by 'daemon' and 'tcp' access_log modules only. | |
5000 | DOC_END | |
5001 | ||
5002 | NAME: netdb_filename | |
5003 | TYPE: string | |
5004 | DEFAULT: stdio:@DEFAULT_NETDB_FILE@ | |
5005 | LOC: Config.netdbFilename | |
5006 | IFDEF: USE_ICMP | |
5007 | DOC_START | |
5008 | Where Squid stores it's netdb journal. | |
5009 | When enabled this journal preserves netdb state between restarts. | |
5010 | ||
5011 | To disable, enter "none". | |
5012 | DOC_END | |
5013 | ||
5014 | COMMENT_START | |
5015 | OPTIONS FOR TROUBLESHOOTING | |
5016 | ----------------------------------------------------------------------------- | |
5017 | COMMENT_END | |
5018 | ||
5019 | NAME: cache_log | |
5020 | TYPE: string | |
5021 | DEFAULT_IF_NONE: @DEFAULT_CACHE_LOG@ | |
5022 | LOC: Debug::cache_log | |
5023 | DOC_START | |
5024 | Squid administrative logging file. | |
5025 | ||
5026 | This is where general information about Squid behavior goes. You can | |
5027 | increase the amount of data logged to this file and how often it is | |
5028 | rotated with "debug_options" | |
5029 | DOC_END | |
5030 | ||
5031 | NAME: debug_options | |
5032 | TYPE: eol | |
5033 | DEFAULT: ALL,1 | |
5034 | DEFAULT_DOC: Log all critical and important messages. | |
5035 | LOC: Debug::debugOptions | |
5036 | DOC_START | |
5037 | Logging options are set as section,level where each source file | |
5038 | is assigned a unique section. Lower levels result in less | |
5039 | output, Full debugging (level 9) can result in a very large | |
5040 | log file, so be careful. | |
5041 | ||
5042 | The magic word "ALL" sets debugging levels for all sections. | |
5043 | The default is to run with "ALL,1" to record important warnings. | |
5044 | ||
5045 | The rotate=N option can be used to keep more or less of these logs | |
5046 | than would otherwise be kept by logfile_rotate. | |
5047 | For most uses a single log should be enough to monitor current | |
5048 | events affecting Squid. | |
5049 | DOC_END | |
5050 | ||
5051 | NAME: coredump_dir | |
5052 | TYPE: string | |
5053 | LOC: Config.coredump_dir | |
5054 | DEFAULT_IF_NONE: none | |
5055 | DEFAULT_DOC: Use the directory from where Squid was started. | |
5056 | DOC_START | |
5057 | By default Squid leaves core files in the directory from where | |
5058 | it was started. If you set 'coredump_dir' to a directory | |
5059 | that exists, Squid will chdir() to that directory at startup | |
5060 | and coredump files will be left there. | |
5061 | ||
5062 | NOCOMMENT_START | |
5063 | ||
5064 | # Leave coredumps in the first cache dir | |
5065 | coredump_dir @DEFAULT_SWAP_DIR@ | |
5066 | NOCOMMENT_END | |
5067 | DOC_END | |
5068 | ||
5069 | ||
5070 | COMMENT_START | |
5071 | OPTIONS FOR FTP GATEWAYING | |
5072 | ----------------------------------------------------------------------------- | |
5073 | COMMENT_END | |
5074 | ||
5075 | NAME: ftp_user | |
5076 | TYPE: string | |
5077 | DEFAULT: Squid@ | |
5078 | LOC: Config.Ftp.anon_user | |
5079 | DOC_START | |
5080 | If you want the anonymous login password to be more informative | |
5081 | (and enable the use of picky FTP servers), set this to something | |
5082 | reasonable for your domain, like wwwuser@somewhere.net | |
5083 | ||
5084 | The reason why this is domainless by default is the | |
5085 | request can be made on the behalf of a user in any domain, | |
5086 | depending on how the cache is used. | |
5087 | Some FTP server also validate the email address is valid | |
5088 | (for example perl.com). | |
5089 | DOC_END | |
5090 | ||
5091 | NAME: ftp_passive | |
5092 | TYPE: onoff | |
5093 | DEFAULT: on | |
5094 | LOC: Config.Ftp.passive | |
5095 | DOC_START | |
5096 | If your firewall does not allow Squid to use passive | |
5097 | connections, turn off this option. | |
5098 | ||
5099 | Use of ftp_epsv_all option requires this to be ON. | |
5100 | DOC_END | |
5101 | ||
5102 | NAME: ftp_epsv_all | |
5103 | TYPE: onoff | |
5104 | DEFAULT: off | |
5105 | LOC: Config.Ftp.epsv_all | |
5106 | DOC_START | |
5107 | FTP Protocol extensions permit the use of a special "EPSV ALL" command. | |
5108 | ||
5109 | NATs may be able to put the connection on a "fast path" through the | |
5110 | translator, as the EPRT command will never be used and therefore, | |
5111 | translation of the data portion of the segments will never be needed. | |
5112 | ||
5113 | When a client only expects to do two-way FTP transfers this may be | |
5114 | useful. | |
5115 | If squid finds that it must do a three-way FTP transfer after issuing | |
5116 | an EPSV ALL command, the FTP session will fail. | |
5117 | ||
5118 | If you have any doubts about this option do not use it. | |
5119 | Squid will nicely attempt all other connection methods. | |
5120 | ||
5121 | Requires ftp_passive to be ON (default) for any effect. | |
5122 | DOC_END | |
5123 | ||
5124 | NAME: ftp_epsv | |
5125 | TYPE: ftp_epsv | |
5126 | DEFAULT: none | |
5127 | LOC: Config.accessList.ftp_epsv | |
5128 | DOC_START | |
5129 | FTP Protocol extensions permit the use of a special "EPSV" command. | |
5130 | ||
5131 | NATs may be able to put the connection on a "fast path" through the | |
5132 | translator using EPSV, as the EPRT command will never be used | |
5133 | and therefore, translation of the data portion of the segments | |
5134 | will never be needed. | |
5135 | ||
5136 | EPSV is often required to interoperate with FTP servers on IPv6 | |
5137 | networks. On the other hand, it may break some IPv4 servers. | |
5138 | ||
5139 | By default, EPSV may try EPSV with any FTP server. To fine tune | |
5140 | that decision, you may restrict EPSV to certain clients or servers | |
5141 | using ACLs: | |
5142 | ||
5143 | ftp_epsv allow|deny al1 acl2 ... | |
5144 | ||
5145 | WARNING: Disabling EPSV may cause problems with external NAT and IPv6. | |
5146 | ||
5147 | Only fast ACLs are supported. | |
5148 | Requires ftp_passive to be ON (default) for any effect. | |
5149 | DOC_END | |
5150 | ||
5151 | NAME: ftp_eprt | |
5152 | TYPE: onoff | |
5153 | DEFAULT: on | |
5154 | LOC: Config.Ftp.eprt | |
5155 | DOC_START | |
5156 | FTP Protocol extensions permit the use of a special "EPRT" command. | |
5157 | ||
5158 | This extension provides a protocol neutral alternative to the | |
5159 | IPv4-only PORT command. When supported it enables active FTP data | |
5160 | channels over IPv6 and efficient NAT handling. | |
5161 | ||
5162 | Turning this OFF will prevent EPRT being attempted and will skip | |
5163 | straight to using PORT for IPv4 servers. | |
5164 | ||
5165 | Some devices are known to not handle this extension correctly and | |
5166 | may result in crashes. Devices which suport EPRT enough to fail | |
5167 | cleanly will result in Squid attempting PORT anyway. This directive | |
5168 | should only be disabled when EPRT results in device failures. | |
5169 | ||
5170 | WARNING: Doing so will convert Squid back to the old behavior with all | |
5171 | the related problems with external NAT devices/layers and IPv4-only FTP. | |
5172 | DOC_END | |
5173 | ||
5174 | NAME: ftp_sanitycheck | |
5175 | TYPE: onoff | |
5176 | DEFAULT: on | |
5177 | LOC: Config.Ftp.sanitycheck | |
5178 | DOC_START | |
5179 | For security and data integrity reasons Squid by default performs | |
5180 | sanity checks of the addresses of FTP data connections ensure the | |
5181 | data connection is to the requested server. If you need to allow | |
5182 | FTP connections to servers using another IP address for the data | |
5183 | connection turn this off. | |
5184 | DOC_END | |
5185 | ||
5186 | NAME: ftp_telnet_protocol | |
5187 | TYPE: onoff | |
5188 | DEFAULT: on | |
5189 | LOC: Config.Ftp.telnet | |
5190 | DOC_START | |
5191 | The FTP protocol is officially defined to use the telnet protocol | |
5192 | as transport channel for the control connection. However, many | |
5193 | implementations are broken and does not respect this aspect of | |
5194 | the FTP protocol. | |
5195 | ||
5196 | If you have trouble accessing files with ASCII code 255 in the | |
5197 | path or similar problems involving this ASCII code you can | |
5198 | try setting this directive to off. If that helps, report to the | |
5199 | operator of the FTP server in question that their FTP server | |
5200 | is broken and does not follow the FTP standard. | |
5201 | DOC_END | |
5202 | ||
5203 | COMMENT_START | |
5204 | OPTIONS FOR EXTERNAL SUPPORT PROGRAMS | |
5205 | ----------------------------------------------------------------------------- | |
5206 | COMMENT_END | |
5207 | ||
5208 | NAME: diskd_program | |
5209 | TYPE: string | |
5210 | DEFAULT: @DEFAULT_DISKD@ | |
5211 | LOC: Config.Program.diskd | |
5212 | DOC_START | |
5213 | Specify the location of the diskd executable. | |
5214 | Note this is only useful if you have compiled in | |
5215 | diskd as one of the store io modules. | |
5216 | DOC_END | |
5217 | ||
5218 | NAME: unlinkd_program | |
5219 | IFDEF: USE_UNLINKD | |
5220 | TYPE: string | |
5221 | DEFAULT: @DEFAULT_UNLINKD@ | |
5222 | LOC: Config.Program.unlinkd | |
5223 | DOC_START | |
5224 | Specify the location of the executable for file deletion process. | |
5225 | DOC_END | |
5226 | ||
5227 | NAME: pinger_program | |
5228 | IFDEF: USE_ICMP | |
5229 | TYPE: icmp | |
5230 | DEFAULT: @DEFAULT_PINGER@ | |
5231 | LOC: IcmpCfg | |
5232 | DOC_START | |
5233 | Specify the location of the executable for the pinger process. | |
5234 | DOC_END | |
5235 | ||
5236 | NAME: pinger_enable | |
5237 | TYPE: onoff | |
5238 | DEFAULT: on | |
5239 | LOC: IcmpCfg.enable | |
5240 | IFDEF: USE_ICMP | |
5241 | DOC_START | |
5242 | Control whether the pinger is active at run-time. | |
5243 | Enables turning ICMP pinger on and off with a simple | |
5244 | squid -k reconfigure. | |
5245 | DOC_END | |
5246 | ||
5247 | ||
5248 | COMMENT_START | |
5249 | OPTIONS FOR URL REWRITING | |
5250 | ----------------------------------------------------------------------------- | |
5251 | COMMENT_END | |
5252 | ||
5253 | NAME: url_rewrite_program redirect_program | |
5254 | TYPE: wordlist | |
5255 | LOC: Config.Program.redirect | |
5256 | DEFAULT: none | |
5257 | DOC_START | |
5258 | Specify the location of the executable URL rewriter to use. | |
5259 | Since they can perform almost any function there isn't one included. | |
5260 | ||
5261 | For each requested URL, the rewriter will receive on line with the format | |
5262 | ||
5263 | [channel-ID <SP>] URL [<SP> extras]<NL> | |
5264 | ||
5265 | See url_rewrite_extras on how to send "extras" with optional values to | |
5266 | the helper. | |
5267 | After processing the request the helper must reply using the following format: | |
5268 | ||
5269 | [channel-ID <SP>] result [<SP> kv-pairs] | |
5270 | ||
5271 | The result code can be: | |
5272 | ||
5273 | OK status=30N url="..." | |
5274 | Redirect the URL to the one supplied in 'url='. | |
5275 | 'status=' is optional and contains the status code to send | |
5276 | the client in Squids HTTP response. It must be one of the | |
5277 | HTTP redirect status codes: 301, 302, 303, 307, 308. | |
5278 | When no status is given Squid will use 302. | |
5279 | ||
5280 | OK rewrite-url="..." | |
5281 | Rewrite the URL to the one supplied in 'rewrite-url='. | |
5282 | The new URL is fetched directly by Squid and returned to | |
5283 | the client as the response to its request. | |
5284 | ||
5285 | OK | |
5286 | When neither of url= and rewrite-url= are sent Squid does | |
5287 | not change the URL. | |
5288 | ||
5289 | ERR | |
5290 | Do not change the URL. | |
5291 | ||
5292 | BH | |
5293 | An internal error occurred in the helper, preventing | |
5294 | a result being identified. The 'message=' key name is | |
5295 | reserved for delivering a log message. | |
5296 | ||
5297 | ||
5298 | In addition to the above kv-pairs Squid also understands the following | |
5299 | optional kv-pairs received from URL rewriters: | |
5300 | clt_conn_tag=TAG | |
5301 | Associates a TAG with the client TCP connection. | |
5302 | The TAG is treated as a regular annotation but persists across | |
5303 | future requests on the client connection rather than just the | |
5304 | current request. A helper may update the TAG during subsequent | |
5305 | requests be returning a new kv-pair. | |
5306 | ||
5307 | When using the concurrency= option the protocol is changed by | |
5308 | introducing a query channel tag in front of the request/response. | |
5309 | The query channel tag is a number between 0 and concurrency-1. | |
5310 | This value must be echoed back unchanged to Squid as the first part | |
5311 | of the response relating to its request. | |
5312 | ||
5313 | WARNING: URL re-writing ability should be avoided whenever possible. | |
5314 | Use the URL redirect form of response instead. | |
5315 | ||
5316 | Re-write creates a difference in the state held by the client | |
5317 | and server. Possibly causing confusion when the server response | |
5318 | contains snippets of its view state. Embeded URLs, response | |
5319 | and content Location headers, etc. are not re-written by this | |
5320 | interface. | |
5321 | ||
5322 | By default, a URL rewriter is not used. | |
5323 | DOC_END | |
5324 | ||
5325 | NAME: url_rewrite_children redirect_children | |
5326 | TYPE: HelperChildConfig | |
5327 | DEFAULT: 20 startup=0 idle=1 concurrency=0 | |
5328 | LOC: Config.redirectChildren | |
5329 | DOC_START | |
5330 | The maximum number of redirector processes to spawn. If you limit | |
5331 | it too few Squid will have to wait for them to process a backlog of | |
5332 | URLs, slowing it down. If you allow too many they will use RAM | |
5333 | and other system resources noticably. | |
5334 | ||
5335 | The startup= and idle= options allow some measure of skew in your | |
5336 | tuning. | |
5337 | ||
5338 | startup= | |
5339 | ||
5340 | Sets a minimum of how many processes are to be spawned when Squid | |
5341 | starts or reconfigures. When set to zero the first request will | |
5342 | cause spawning of the first child process to handle it. | |
5343 | ||
5344 | Starting too few will cause an initial slowdown in traffic as Squid | |
5345 | attempts to simultaneously spawn enough processes to cope. | |
5346 | ||
5347 | idle= | |
5348 | ||
5349 | Sets a minimum of how many processes Squid is to try and keep available | |
5350 | at all times. When traffic begins to rise above what the existing | |
5351 | processes can handle this many more will be spawned up to the maximum | |
5352 | configured. A minimum setting of 1 is required. | |
5353 | ||
5354 | concurrency= | |
5355 | ||
5356 | The number of requests each redirector helper can handle in | |
5357 | parallel. Defaults to 0 which indicates the redirector | |
5358 | is a old-style single threaded redirector. | |
5359 | ||
5360 | When this directive is set to a value >= 1 then the protocol | |
5361 | used to communicate with the helper is modified to include | |
5362 | an ID in front of the request/response. The ID from the request | |
5363 | must be echoed back with the response to that request. | |
5364 | ||
5365 | queue-size=N | |
5366 | ||
5367 | Sets the maximum number of queued requests to N. The default maximum | |
5368 | is 2*numberofchildren. If the queued requests exceed queue size and | |
5369 | redirector_bypass configuration option is set, then redirector is bypassed. | |
5370 | Otherwise, Squid is allowed to temporarily exceed the configured maximum, | |
5371 | marking the affected helper as "overloaded". If the helper overload lasts | |
5372 | more than 3 minutes, the action prescribed by the on-persistent-overload | |
5373 | option applies. | |
5374 | ||
5375 | on-persistent-overload=action | |
5376 | ||
5377 | Specifies Squid reaction to a new helper request arriving when the helper | |
5378 | has been overloaded for more that 3 minutes already. The number of queued | |
5379 | requests determines whether the helper is overloaded (see the queue-size | |
5380 | option). | |
5381 | ||
5382 | Two actions are supported: | |
5383 | ||
5384 | die Squid worker quits. This is the default behavior. | |
5385 | ||
5386 | ERR Squid treats the helper request as if it was | |
5387 | immediately submitted, and the helper immediately | |
5388 | replied with an ERR response. This action has no effect | |
5389 | on the already queued and in-progress helper requests. | |
5390 | DOC_END | |
5391 | ||
5392 | NAME: url_rewrite_host_header redirect_rewrites_host_header | |
5393 | TYPE: onoff | |
5394 | DEFAULT: on | |
5395 | LOC: Config.onoff.redir_rewrites_host | |
5396 | DOC_START | |
5397 | To preserve same-origin security policies in browsers and | |
5398 | prevent Host: header forgery by redirectors Squid rewrites | |
5399 | any Host: header in redirected requests. | |
5400 | ||
5401 | If you are running an accelerator this may not be a wanted | |
5402 | effect of a redirector. This directive enables you disable | |
5403 | Host: alteration in reverse-proxy traffic. | |
5404 | ||
5405 | WARNING: Entries are cached on the result of the URL rewriting | |
5406 | process, so be careful if you have domain-virtual hosts. | |
5407 | ||
5408 | WARNING: Squid and other software verifies the URL and Host | |
5409 | are matching, so be careful not to relay through other proxies | |
5410 | or inspecting firewalls with this disabled. | |
5411 | DOC_END | |
5412 | ||
5413 | NAME: url_rewrite_access redirector_access | |
5414 | TYPE: acl_access | |
5415 | DEFAULT: none | |
5416 | DEFAULT_DOC: Allow, unless rules exist in squid.conf. | |
5417 | LOC: Config.accessList.redirector | |
5418 | DOC_START | |
5419 | If defined, this access list specifies which requests are | |
5420 | sent to the redirector processes. | |
5421 | ||
5422 | This clause supports both fast and slow acl types. | |
5423 | See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details. | |
5424 | DOC_END | |
5425 | ||
5426 | NAME: url_rewrite_bypass redirector_bypass | |
5427 | TYPE: onoff | |
5428 | LOC: Config.onoff.redirector_bypass | |
5429 | DEFAULT: off | |
5430 | DOC_START | |
5431 | When this is 'on', a request will not go through the | |
5432 | redirector if all the helpers are busy. If this is 'off' and the | |
5433 | redirector queue grows too large, the action is prescribed by the | |
5434 | on-persistent-overload option. You should only enable this if the | |
5435 | redirectors are not critical to your caching system. If you use | |
5436 | redirectors for access control, and you enable this option, | |
5437 | users may have access to pages they should not | |
5438 | be allowed to request. | |
5439 | This options sets default queue-size option of the url_rewrite_children | |
5440 | to 0. | |
5441 | DOC_END | |
5442 | ||
5443 | NAME: url_rewrite_extras | |
5444 | TYPE: TokenOrQuotedString | |
5445 | LOC: Config.redirector_extras | |
5446 | DEFAULT: "%>a/%>A %un %>rm myip=%la myport=%lp" | |
5447 | DOC_START | |
5448 | Specifies a string to be append to request line format for the | |
5449 | rewriter helper. "Quoted" format values may contain spaces and | |
5450 | logformat %macros. In theory, any logformat %macro can be used. | |
5451 | In practice, a %macro expands as a dash (-) if the helper request is | |
5452 | sent before the required macro information is available to Squid. | |
5453 | DOC_END | |
5454 | ||
5455 | NAME: url_rewrite_timeout | |
5456 | TYPE: UrlHelperTimeout | |
5457 | LOC: Config.onUrlRewriteTimeout | |
5458 | DEFAULT: none | |
5459 | DEFAULT_DOC: Squid waits for the helper response forever | |
5460 | DOC_START | |
5461 | Squid times active requests to redirector. The timeout value and Squid | |
5462 | reaction to a timed out request are configurable using the following | |
5463 | format: | |
5464 | ||
5465 | url_rewrite_timeout timeout time-units on_timeout=<action> [response=<quoted-response>] | |
5466 | ||
5467 | supported timeout actions: | |
5468 | fail Squid return a ERR_GATEWAY_FAILURE error page | |
5469 | ||
5470 | bypass Do not re-write the URL | |
5471 | ||
5472 | retry Send the lookup to the helper again | |
5473 | ||
5474 | use_configured_response | |
5475 | Use the <quoted-response> as helper response | |
5476 | DOC_END | |
5477 | ||
5478 | COMMENT_START | |
5479 | OPTIONS FOR STORE ID | |
5480 | ----------------------------------------------------------------------------- | |
5481 | COMMENT_END | |
5482 | ||
5483 | NAME: store_id_program storeurl_rewrite_program | |
5484 | TYPE: wordlist | |
5485 | LOC: Config.Program.store_id | |
5486 | DEFAULT: none | |
5487 | DOC_START | |
5488 | Specify the location of the executable StoreID helper to use. | |
5489 | Since they can perform almost any function there isn't one included. | |
5490 | ||
5491 | For each requested URL, the helper will receive one line with the format | |
5492 | ||
5493 | [channel-ID <SP>] URL [<SP> extras]<NL> | |
5494 | ||
5495 | ||
5496 | After processing the request the helper must reply using the following format: | |
5497 | ||
5498 | [channel-ID <SP>] result [<SP> kv-pairs] | |
5499 | ||
5500 | The result code can be: | |
5501 | ||
5502 | OK store-id="..." | |
5503 | Use the StoreID supplied in 'store-id='. | |
5504 | ||
5505 | ERR | |
5506 | The default is to use HTTP request URL as the store ID. | |
5507 | ||
5508 | BH | |
5509 | An internal error occured in the helper, preventing | |
5510 | a result being identified. | |
5511 | ||
5512 | In addition to the above kv-pairs Squid also understands the following | |
5513 | optional kv-pairs received from URL rewriters: | |
5514 | clt_conn_tag=TAG | |
5515 | Associates a TAG with the client TCP connection. | |
5516 | Please see url_rewrite_program related documentation for this | |
5517 | kv-pair | |
5518 | ||
5519 | Helper programs should be prepared to receive and possibly ignore | |
5520 | additional whitespace-separated tokens on each input line. | |
5521 | ||
5522 | When using the concurrency= option the protocol is changed by | |
5523 | introducing a query channel tag in front of the request/response. | |
5524 | The query channel tag is a number between 0 and concurrency-1. | |
5525 | This value must be echoed back unchanged to Squid as the first part | |
5526 | of the response relating to its request. | |
5527 | ||
5528 | NOTE: when using StoreID refresh_pattern will apply to the StoreID | |
5529 | returned from the helper and not the URL. | |
5530 | ||
5531 | WARNING: Wrong StoreID value returned by a careless helper may result | |
5532 | in the wrong cached response returned to the user. | |
5533 | ||
5534 | By default, a StoreID helper is not used. | |
5535 | DOC_END | |
5536 | ||
5537 | NAME: store_id_extras | |
5538 | TYPE: TokenOrQuotedString | |
5539 | LOC: Config.storeId_extras | |
5540 | DEFAULT: "%>a/%>A %un %>rm myip=%la myport=%lp" | |
5541 | DOC_START | |
5542 | Specifies a string to be append to request line format for the | |
5543 | StoreId helper. "Quoted" format values may contain spaces and | |
5544 | logformat %macros. In theory, any logformat %macro can be used. | |
5545 | In practice, a %macro expands as a dash (-) if the helper request is | |
5546 | sent before the required macro information is available to Squid. | |
5547 | DOC_END | |
5548 | ||
5549 | NAME: store_id_children storeurl_rewrite_children | |
5550 | TYPE: HelperChildConfig | |
5551 | DEFAULT: 20 startup=0 idle=1 concurrency=0 | |
5552 | LOC: Config.storeIdChildren | |
5553 | DOC_START | |
5554 | The maximum number of StoreID helper processes to spawn. If you limit | |
5555 | it too few Squid will have to wait for them to process a backlog of | |
5556 | requests, slowing it down. If you allow too many they will use RAM | |
5557 | and other system resources noticably. | |
5558 | ||
5559 | The startup= and idle= options allow some measure of skew in your | |
5560 | tuning. | |
5561 | ||
5562 | startup= | |
5563 | ||
5564 | Sets a minimum of how many processes are to be spawned when Squid | |
5565 | starts or reconfigures. When set to zero the first request will | |
5566 | cause spawning of the first child process to handle it. | |
5567 | ||
5568 | Starting too few will cause an initial slowdown in traffic as Squid | |
5569 | attempts to simultaneously spawn enough processes to cope. | |
5570 | ||
5571 | idle= | |
5572 | ||
5573 | Sets a minimum of how many processes Squid is to try and keep available | |
5574 | at all times. When traffic begins to rise above what the existing | |
5575 | processes can handle this many more will be spawned up to the maximum | |
5576 | configured. A minimum setting of 1 is required. | |
5577 | ||
5578 | concurrency= | |
5579 | ||
5580 | The number of requests each storeID helper can handle in | |
5581 | parallel. Defaults to 0 which indicates the helper | |
5582 | is a old-style single threaded program. | |
5583 | ||
5584 | When this directive is set to a value >= 1 then the protocol | |
5585 | used to communicate with the helper is modified to include | |
5586 | an ID in front of the request/response. The ID from the request | |
5587 | must be echoed back with the response to that request. | |
5588 | ||
5589 | queue-size=N | |
5590 | ||
5591 | Sets the maximum number of queued requests to N. The default maximum | |
5592 | is 2*numberofchildren. If the queued requests exceed queue size and | |
5593 | redirector_bypass configuration option is set, then redirector is bypassed. | |
5594 | Otherwise, Squid is allowed to temporarily exceed the configured maximum, | |
5595 | marking the affected helper as "overloaded". If the helper overload lasts | |
5596 | more than 3 minutes, the action prescribed by the on-persistent-overload | |
5597 | option applies. | |
5598 | ||
5599 | on-persistent-overload=action | |
5600 | ||
5601 | Specifies Squid reaction to a new helper request arriving when the helper | |
5602 | has been overloaded for more that 3 minutes already. The number of queued | |
5603 | requests determines whether the helper is overloaded (see the queue-size | |
5604 | option). | |
5605 | ||
5606 | Two actions are supported: | |
5607 | ||
5608 | die Squid worker quits. This is the default behavior. | |
5609 | ||
5610 | ERR Squid treats the helper request as if it was | |
5611 | immediately submitted, and the helper immediately | |
5612 | replied with an ERR response. This action has no effect | |
5613 | on the already queued and in-progress helper requests. | |
5614 | DOC_END | |
5615 | ||
5616 | NAME: store_id_access storeurl_rewrite_access | |
5617 | TYPE: acl_access | |
5618 | DEFAULT: none | |
5619 | DEFAULT_DOC: Allow, unless rules exist in squid.conf. | |
5620 | LOC: Config.accessList.store_id | |
5621 | DOC_START | |
5622 | If defined, this access list specifies which requests are | |
5623 | sent to the StoreID processes. By default all requests | |
5624 | are sent. | |
5625 | ||
5626 | This clause supports both fast and slow acl types. | |
5627 | See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details. | |
5628 | DOC_END | |
5629 | ||
5630 | NAME: store_id_bypass storeurl_rewrite_bypass | |
5631 | TYPE: onoff | |
5632 | LOC: Config.onoff.store_id_bypass | |
5633 | DEFAULT: on | |
5634 | DOC_START | |
5635 | When this is 'on', a request will not go through the | |
5636 | helper if all helpers are busy. If this is 'off' and the helper | |
5637 | queue grows too large, the action is prescribed by the | |
5638 | on-persistent-overload option. You should only enable this if the | |
5639 | helpers are not critical to your caching system. If you use | |
5640 | helpers for critical caching components, and you enable this | |
5641 | option, users may not get objects from cache. | |
5642 | This options sets default queue-size option of the store_id_children | |
5643 | to 0. | |
5644 | DOC_END | |
5645 | ||
5646 | COMMENT_START | |
5647 | OPTIONS FOR TUNING THE CACHE | |
5648 | ----------------------------------------------------------------------------- | |
5649 | COMMENT_END | |
5650 | ||
5651 | NAME: cache no_cache | |
5652 | TYPE: acl_access | |
5653 | DEFAULT: none | |
5654 | DEFAULT_DOC: By default, this directive is unused and has no effect. | |
5655 | LOC: Config.accessList.noCache | |
5656 | DOC_START | |
5657 | Requests denied by this directive will not be served from the cache | |
5658 | and their responses will not be stored in the cache. This directive | |
5659 | has no effect on other transactions and on already cached responses. | |
5660 | ||
5661 | This clause supports both fast and slow acl types. | |
5662 | See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details. | |
5663 | ||
5664 | This and the two other similar caching directives listed below are | |
5665 | checked at different transaction processing stages, have different | |
5666 | access to response information, affect different cache operations, | |
5667 | and differ in slow ACLs support: | |
5668 | ||
5669 | * cache: Checked before Squid makes a hit/miss determination. | |
5670 | No access to reply information! | |
5671 | Denies both serving a hit and storing a miss. | |
5672 | Supports both fast and slow ACLs. | |
5673 | * send_hit: Checked after a hit was detected. | |
5674 | Has access to reply (hit) information. | |
5675 | Denies serving a hit only. | |
5676 | Supports fast ACLs only. | |
5677 | * store_miss: Checked before storing a cachable miss. | |
5678 | Has access to reply (miss) information. | |
5679 | Denies storing a miss only. | |
5680 | Supports fast ACLs only. | |
5681 | ||
5682 | If you are not sure which of the three directives to use, apply the | |
5683 | following decision logic: | |
5684 | ||
5685 | * If your ACL(s) are of slow type _and_ need response info, redesign. | |
5686 | Squid does not support that particular combination at this time. | |
5687 | Otherwise: | |
5688 | * If your directive ACL(s) are of slow type, use "cache"; and/or | |
5689 | * if your directive ACL(s) need no response info, use "cache". | |
5690 | Otherwise: | |
5691 | * If you do not want the response cached, use store_miss; and/or | |
5692 | * if you do not want a hit on a cached response, use send_hit. | |
5693 | DOC_END | |
5694 | ||
5695 | NAME: send_hit | |
5696 | TYPE: acl_access | |
5697 | DEFAULT: none | |
5698 | DEFAULT_DOC: By default, this directive is unused and has no effect. | |
5699 | LOC: Config.accessList.sendHit | |
5700 | DOC_START | |
5701 | Responses denied by this directive will not be served from the cache | |
5702 | (but may still be cached, see store_miss). This directive has no | |
5703 | effect on the responses it allows and on the cached objects. | |
5704 | ||
5705 | Please see the "cache" directive for a summary of differences among | |
5706 | store_miss, send_hit, and cache directives. | |
5707 | ||
5708 | Unlike the "cache" directive, send_hit only supports fast acl | |
5709 | types. See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details. | |
5710 | ||
5711 | For example: | |
5712 | ||
5713 | # apply custom Store ID mapping to some URLs | |
5714 | acl MapMe dstdomain .c.example.com | |
5715 | store_id_program ... | |
5716 | store_id_access allow MapMe | |
5717 | ||
5718 | # but prevent caching of special responses | |
5719 | # such as 302 redirects that cause StoreID loops | |
5720 | acl Ordinary http_status 200-299 | |
5721 | store_miss deny MapMe !Ordinary | |
5722 | ||
5723 | # and do not serve any previously stored special responses | |
5724 | # from the cache (in case they were already cached before | |
5725 | # the above store_miss rule was in effect). | |
5726 | send_hit deny MapMe !Ordinary | |
5727 | DOC_END | |
5728 | ||
5729 | NAME: store_miss | |
5730 | TYPE: acl_access | |
5731 | DEFAULT: none | |
5732 | DEFAULT_DOC: By default, this directive is unused and has no effect. | |
5733 | LOC: Config.accessList.storeMiss | |
5734 | DOC_START | |
5735 | Responses denied by this directive will not be cached (but may still | |
5736 | be served from the cache, see send_hit). This directive has no | |
5737 | effect on the responses it allows and on the already cached responses. | |
5738 | ||
5739 | Please see the "cache" directive for a summary of differences among | |
5740 | store_miss, send_hit, and cache directives. See the | |
5741 | send_hit directive for a usage example. | |
5742 | ||
5743 | Unlike the "cache" directive, store_miss only supports fast acl | |
5744 | types. See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details. | |
5745 | DOC_END | |
5746 | ||
5747 | NAME: max_stale | |
5748 | COMMENT: time-units | |
5749 | TYPE: time_t | |
5750 | LOC: Config.maxStale | |
5751 | DEFAULT: 1 week | |
5752 | DOC_START | |
5753 | This option puts an upper limit on how stale content Squid | |
5754 | will serve from the cache if cache validation fails. | |
5755 | Can be overriden by the refresh_pattern max-stale option. | |
5756 | DOC_END | |
5757 | ||
5758 | NAME: refresh_pattern | |
5759 | TYPE: refreshpattern | |
5760 | LOC: Config.Refresh | |
5761 | DEFAULT: none | |
5762 | DOC_START | |
5763 | usage: refresh_pattern [-i] regex min percent max [options] | |
5764 | ||
5765 | By default, regular expressions are CASE-SENSITIVE. To make | |
5766 | them case-insensitive, use the -i option. | |
5767 | ||
5768 | 'Min' is the time (in minutes) an object without an explicit | |
5769 | expiry time should be considered fresh. The recommended | |
5770 | value is 0, any higher values may cause dynamic applications | |
5771 | to be erroneously cached unless the application designer | |
5772 | has taken the appropriate actions. | |
5773 | ||
5774 | 'Percent' is a percentage of the objects age (time since last | |
5775 | modification age) an object without explicit expiry time | |
5776 | will be considered fresh. | |
5777 | ||
5778 | 'Max' is an upper limit on how long objects without an explicit | |
5779 | expiry time will be considered fresh. The value is also used | |
5780 | to form Cache-Control: max-age header for a request sent from | |
5781 | Squid to origin/parent. | |
5782 | ||
5783 | options: override-expire | |
5784 | override-lastmod | |
5785 | reload-into-ims | |
5786 | ignore-reload | |
5787 | ignore-no-store | |
5788 | ignore-private | |
5789 | max-stale=NN | |
5790 | refresh-ims | |
5791 | store-stale | |
5792 | ||
5793 | override-expire enforces min age even if the server | |
5794 | sent an explicit expiry time (e.g., with the | |
5795 | Expires: header or Cache-Control: max-age). Doing this | |
5796 | VIOLATES the HTTP standard. Enabling this feature | |
5797 | could make you liable for problems which it causes. | |
5798 | ||
5799 | Note: override-expire does not enforce staleness - it only extends | |
5800 | freshness / min. If the server returns a Expires time which | |
5801 | is longer than your max time, Squid will still consider | |
5802 | the object fresh for that period of time. | |
5803 | ||
5804 | override-lastmod enforces min age even on objects | |
5805 | that were modified recently. | |
5806 | ||
5807 | reload-into-ims changes a client no-cache or ``reload'' | |
5808 | request for a cached entry into a conditional request using | |
5809 | If-Modified-Since and/or If-None-Match headers, provided the | |
5810 | cached entry has a Last-Modified and/or a strong ETag header. | |
5811 | Doing this VIOLATES the HTTP standard. Enabling this feature | |
5812 | could make you liable for problems which it causes. | |
5813 | ||
5814 | ignore-reload ignores a client no-cache or ``reload'' | |
5815 | header. Doing this VIOLATES the HTTP standard. Enabling | |
5816 | this feature could make you liable for problems which | |
5817 | it causes. | |
5818 | ||
5819 | ignore-no-store ignores any ``Cache-control: no-store'' | |
5820 | headers received from a server. Doing this VIOLATES | |
5821 | the HTTP standard. Enabling this feature could make you | |
5822 | liable for problems which it causes. | |
5823 | ||
5824 | ignore-private ignores any ``Cache-control: private'' | |
5825 | headers received from a server. Doing this VIOLATES | |
5826 | the HTTP standard. Enabling this feature could make you | |
5827 | liable for problems which it causes. | |
5828 | ||
5829 | refresh-ims causes squid to contact the origin server | |
5830 | when a client issues an If-Modified-Since request. This | |
5831 | ensures that the client will receive an updated version | |
5832 | if one is available. | |
5833 | ||
5834 | store-stale stores responses even if they don't have explicit | |
5835 | freshness or a validator (i.e., Last-Modified or an ETag) | |
5836 | present, or if they're already stale. By default, Squid will | |
5837 | not cache such responses because they usually can't be | |
5838 | reused. Note that such responses will be stale by default. | |
5839 | ||
5840 | max-stale=NN provide a maximum staleness factor. Squid won't | |
5841 | serve objects more stale than this even if it failed to | |
5842 | validate the object. Default: use the max_stale global limit. | |
5843 | ||
5844 | Basically a cached object is: | |
5845 | ||
5846 | FRESH if expire > now, else STALE | |
5847 | STALE if age > max | |
5848 | FRESH if lm-factor < percent, else STALE | |
5849 | FRESH if age < min | |
5850 | else STALE | |
5851 | ||
5852 | The refresh_pattern lines are checked in the order listed here. | |
5853 | The first entry which matches is used. If none of the entries | |
5854 | match the default will be used. | |
5855 | ||
5856 | Note, you must uncomment all the default lines if you want | |
5857 | to change one. The default setting is only active if none is | |
5858 | used. | |
5859 | ||
5860 | NOCOMMENT_START | |
5861 | ||
5862 | # | |
5863 | # Add any of your own refresh_pattern entries above these. | |
5864 | # | |
5865 | refresh_pattern ^ftp: 1440 20% 10080 | |
5866 | refresh_pattern ^gopher: 1440 0% 1440 | |
5867 | refresh_pattern -i (/cgi-bin/|\?) 0 0% 0 | |
5868 | refresh_pattern . 0 20% 4320 | |
5869 | NOCOMMENT_END | |
5870 | DOC_END | |
5871 | ||
5872 | NAME: quick_abort_min | |
5873 | COMMENT: (KB) | |
5874 | TYPE: kb_int64_t | |
5875 | DEFAULT: 16 KB | |
5876 | LOC: Config.quickAbort.min | |
5877 | DOC_NONE | |
5878 | ||
5879 | NAME: quick_abort_max | |
5880 | COMMENT: (KB) | |
5881 | TYPE: kb_int64_t | |
5882 | DEFAULT: 16 KB | |
5883 | LOC: Config.quickAbort.max | |
5884 | DOC_NONE | |
5885 | ||
5886 | NAME: quick_abort_pct | |
5887 | COMMENT: (percent) | |
5888 | TYPE: int | |
5889 | DEFAULT: 95 | |
5890 | LOC: Config.quickAbort.pct | |
5891 | DOC_START | |
5892 | The cache by default continues downloading aborted requests | |
5893 | which are almost completed (less than 16 KB remaining). This | |
5894 | may be undesirable on slow (e.g. SLIP) links and/or very busy | |
5895 | caches. Impatient users may tie up file descriptors and | |
5896 | bandwidth by repeatedly requesting and immediately aborting | |
5897 | downloads. | |
5898 | ||
5899 | When the user aborts a request, Squid will check the | |
5900 | quick_abort values to the amount of data transferred until | |
5901 | then. | |
5902 | ||
5903 | If the transfer has less than 'quick_abort_min' KB remaining, | |
5904 | it will finish the retrieval. | |
5905 | ||
5906 | If the transfer has more than 'quick_abort_max' KB remaining, | |
5907 | it will abort the retrieval. | |
5908 | ||
5909 | If more than 'quick_abort_pct' of the transfer has completed, | |
5910 | it will finish the retrieval. | |
5911 | ||
5912 | If you do not want any retrieval to continue after the client | |
5913 | has aborted, set both 'quick_abort_min' and 'quick_abort_max' | |
5914 | to '0 KB'. | |
5915 | ||
5916 | If you want retrievals to always continue if they are being | |
5917 | cached set 'quick_abort_min' to '-1 KB'. | |
5918 | DOC_END | |
5919 | ||
5920 | NAME: read_ahead_gap | |
5921 | COMMENT: buffer-size | |
5922 | TYPE: b_int64_t | |
5923 | LOC: Config.readAheadGap | |
5924 | DEFAULT: 16 KB | |
5925 | DOC_START | |
5926 | The amount of data the cache will buffer ahead of what has been | |
5927 | sent to the client when retrieving an object from another server. | |
5928 | DOC_END | |
5929 | ||
5930 | NAME: negative_ttl | |
5931 | IFDEF: USE_HTTP_VIOLATIONS | |
5932 | COMMENT: time-units | |
5933 | TYPE: time_t | |
5934 | LOC: Config.negativeTtl | |
5935 | DEFAULT: 0 seconds | |
5936 | DOC_START | |
5937 | Set the Default Time-to-Live (TTL) for failed requests. | |
5938 | Certain types of failures (such as "connection refused" and | |
5939 | "404 Not Found") are able to be negatively-cached for a short time. | |
5940 | Modern web servers should provide Expires: header, however if they | |
5941 | do not this can provide a minimum TTL. | |
5942 | The default is not to cache errors with unknown expiry details. | |
5943 | ||
5944 | Note that this is different from negative caching of DNS lookups. | |
5945 | ||
5946 | WARNING: Doing this VIOLATES the HTTP standard. Enabling | |
5947 | this feature could make you liable for problems which it | |
5948 | causes. | |
5949 | DOC_END | |
5950 | ||
5951 | NAME: positive_dns_ttl | |
5952 | COMMENT: time-units | |
5953 | TYPE: time_t | |
5954 | LOC: Config.positiveDnsTtl | |
5955 | DEFAULT: 6 hours | |
5956 | DOC_START | |
5957 | Upper limit on how long Squid will cache positive DNS responses. | |
5958 | Default is 6 hours (360 minutes). This directive must be set | |
5959 | larger than negative_dns_ttl. | |
5960 | DOC_END | |
5961 | ||
5962 | NAME: negative_dns_ttl | |
5963 | COMMENT: time-units | |
5964 | TYPE: time_t | |
5965 | LOC: Config.negativeDnsTtl | |
5966 | DEFAULT: 1 minutes | |
5967 | DOC_START | |
5968 | Time-to-Live (TTL) for negative caching of failed DNS lookups. | |
5969 | This also sets the lower cache limit on positive lookups. | |
5970 | Minimum value is 1 second, and it is not recommendable to go | |
5971 | much below 10 seconds. | |
5972 | DOC_END | |
5973 | ||
5974 | NAME: range_offset_limit | |
5975 | COMMENT: size [acl acl...] | |
5976 | TYPE: acl_b_size_t | |
5977 | LOC: Config.rangeOffsetLimit | |
5978 | DEFAULT: none | |
5979 | DOC_START | |
5980 | usage: (size) [units] [[!]aclname] | |
5981 | ||
5982 | Sets an upper limit on how far (number of bytes) into the file | |
5983 | a Range request may be to cause Squid to prefetch the whole file. | |
5984 | If beyond this limit, Squid forwards the Range request as it is and | |
5985 | the result is NOT cached. | |
5986 | ||
5987 | This is to stop a far ahead range request (lets say start at 17MB) | |
5988 | from making Squid fetch the whole object up to that point before | |
5989 | sending anything to the client. | |
5990 | ||
5991 | Multiple range_offset_limit lines may be specified, and they will | |
5992 | be searched from top to bottom on each request until a match is found. | |
5993 | The first match found will be used. If no line matches a request, the | |
5994 | default limit of 0 bytes will be used. | |
5995 | ||
5996 | 'size' is the limit specified as a number of units. | |
5997 | ||
5998 | 'units' specifies whether to use bytes, KB, MB, etc. | |
5999 | If no units are specified bytes are assumed. | |
6000 | ||
6001 | A size of 0 causes Squid to never fetch more than the | |
6002 | client requested. (default) | |
6003 | ||
6004 | A size of 'none' causes Squid to always fetch the object from the | |
6005 | beginning so it may cache the result. (2.0 style) | |
6006 | ||
6007 | 'aclname' is the name of a defined ACL. | |
6008 | ||
6009 | NP: Using 'none' as the byte value here will override any quick_abort settings | |
6010 | that may otherwise apply to the range request. The range request will | |
6011 | be fully fetched from start to finish regardless of the client | |
6012 | actions. This affects bandwidth usage. | |
6013 | DOC_END | |
6014 | ||
6015 | NAME: minimum_expiry_time | |
6016 | COMMENT: (seconds) | |
6017 | TYPE: time_t | |
6018 | LOC: Config.minimum_expiry_time | |
6019 | DEFAULT: 60 seconds | |
6020 | DOC_START | |
6021 | The minimum caching time according to (Expires - Date) | |
6022 | headers Squid honors if the object can't be revalidated. | |
6023 | The default is 60 seconds. | |
6024 | ||
6025 | In reverse proxy environments it might be desirable to honor | |
6026 | shorter object lifetimes. It is most likely better to make | |
6027 | your server return a meaningful Last-Modified header however. | |
6028 | ||
6029 | In ESI environments where page fragments often have short | |
6030 | lifetimes, this will often be best set to 0. | |
6031 | DOC_END | |
6032 | ||
6033 | NAME: store_avg_object_size | |
6034 | COMMENT: (bytes) | |
6035 | TYPE: b_int64_t | |
6036 | DEFAULT: 13 KB | |
6037 | LOC: Config.Store.avgObjectSize | |
6038 | DOC_START | |
6039 | Average object size, used to estimate number of objects your | |
6040 | cache can hold. The default is 13 KB. | |
6041 | ||
6042 | This is used to pre-seed the cache index memory allocation to | |
6043 | reduce expensive reallocate operations while handling clients | |
6044 | traffic. Too-large values may result in memory allocation during | |
6045 | peak traffic, too-small values will result in wasted memory. | |
6046 | ||
6047 | Check the cache manager 'info' report metrics for the real | |
6048 | object sizes seen by your Squid before tuning this. | |
6049 | DOC_END | |
6050 | ||
6051 | NAME: store_objects_per_bucket | |
6052 | TYPE: int | |
6053 | DEFAULT: 20 | |
6054 | LOC: Config.Store.objectsPerBucket | |
6055 | DOC_START | |
6056 | Target number of objects per bucket in the store hash table. | |
6057 | Lowering this value increases the total number of buckets and | |
6058 | also the storage maintenance rate. The default is 20. | |
6059 | DOC_END | |
6060 | ||
6061 | COMMENT_START | |
6062 | HTTP OPTIONS | |
6063 | ----------------------------------------------------------------------------- | |
6064 | COMMENT_END | |
6065 | ||
6066 | NAME: request_header_max_size | |
6067 | COMMENT: (KB) | |
6068 | TYPE: b_size_t | |
6069 | DEFAULT: 64 KB | |
6070 | LOC: Config.maxRequestHeaderSize | |
6071 | DOC_START | |
6072 | This specifies the maximum size for HTTP headers in a request. | |
6073 | Request headers are usually relatively small (about 512 bytes). | |
6074 | Placing a limit on the request header size will catch certain | |
6075 | bugs (for example with persistent connections) and possibly | |
6076 | buffer-overflow or denial-of-service attacks. | |
6077 | DOC_END | |
6078 | ||
6079 | NAME: reply_header_max_size | |
6080 | COMMENT: (KB) | |
6081 | TYPE: b_size_t | |
6082 | DEFAULT: 64 KB | |
6083 | LOC: Config.maxReplyHeaderSize | |
6084 | DOC_START | |
6085 | This specifies the maximum size for HTTP headers in a reply. | |
6086 | Reply headers are usually relatively small (about 512 bytes). | |
6087 | Placing a limit on the reply header size will catch certain | |
6088 | bugs (for example with persistent connections) and possibly | |
6089 | buffer-overflow or denial-of-service attacks. | |
6090 | DOC_END | |
6091 | ||
6092 | NAME: request_body_max_size | |
6093 | COMMENT: (bytes) | |
6094 | TYPE: b_int64_t | |
6095 | DEFAULT: 0 KB | |
6096 | DEFAULT_DOC: No limit. | |
6097 | LOC: Config.maxRequestBodySize | |
6098 | DOC_START | |
6099 | This specifies the maximum size for an HTTP request body. | |
6100 | In other words, the maximum size of a PUT/POST request. | |
6101 | A user who attempts to send a request with a body larger | |
6102 | than this limit receives an "Invalid Request" error message. | |
6103 | If you set this parameter to a zero (the default), there will | |
6104 | be no limit imposed. | |
6105 | ||
6106 | See also client_request_buffer_max_size for an alternative | |
6107 | limitation on client uploads which can be configured. | |
6108 | DOC_END | |
6109 | ||
6110 | NAME: client_request_buffer_max_size | |
6111 | COMMENT: (bytes) | |
6112 | TYPE: b_size_t | |
6113 | DEFAULT: 512 KB | |
6114 | LOC: Config.maxRequestBufferSize | |
6115 | DOC_START | |
6116 | This specifies the maximum buffer size of a client request. | |
6117 | It prevents squid eating too much memory when somebody uploads | |
6118 | a large file. | |
6119 | DOC_END | |
6120 | ||
6121 | NAME: broken_posts | |
6122 | IFDEF: USE_HTTP_VIOLATIONS | |
6123 | TYPE: acl_access | |
6124 | DEFAULT: none | |
6125 | DEFAULT_DOC: Obey RFC 2616. | |
6126 | LOC: Config.accessList.brokenPosts | |
6127 | DOC_START | |
6128 | A list of ACL elements which, if matched, causes Squid to send | |
6129 | an extra CRLF pair after the body of a PUT/POST request. | |
6130 | ||
6131 | Some HTTP servers has broken implementations of PUT/POST, | |
6132 | and rely on an extra CRLF pair sent by some WWW clients. | |
6133 | ||
6134 | Quote from RFC2616 section 4.1 on this matter: | |
6135 | ||
6136 | Note: certain buggy HTTP/1.0 client implementations generate an | |
6137 | extra CRLF's after a POST request. To restate what is explicitly | |
6138 | forbidden by the BNF, an HTTP/1.1 client must not preface or follow | |
6139 | a request with an extra CRLF. | |
6140 | ||
6141 | This clause only supports fast acl types. | |
6142 | See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details. | |
6143 | ||
6144 | Example: | |
6145 | acl buggy_server url_regex ^http://.... | |
6146 | broken_posts allow buggy_server | |
6147 | DOC_END | |
6148 | ||
6149 | NAME: adaptation_uses_indirect_client icap_uses_indirect_client | |
6150 | COMMENT: on|off | |
6151 | TYPE: onoff | |
6152 | IFDEF: FOLLOW_X_FORWARDED_FOR&&USE_ADAPTATION | |
6153 | DEFAULT: on | |
6154 | LOC: Adaptation::Config::use_indirect_client | |
6155 | DOC_START | |
6156 | Controls whether the indirect client IP address (instead of the direct | |
6157 | client IP address) is passed to adaptation services. | |
6158 | ||
6159 | See also: follow_x_forwarded_for adaptation_send_client_ip | |
6160 | DOC_END | |
6161 | ||
6162 | NAME: via | |
6163 | IFDEF: USE_HTTP_VIOLATIONS | |
6164 | COMMENT: on|off | |
6165 | TYPE: onoff | |
6166 | DEFAULT: on | |
6167 | LOC: Config.onoff.via | |
6168 | DOC_START | |
6169 | If set (default), Squid will include a Via header in requests and | |
6170 | replies as required by RFC2616. | |
6171 | DOC_END | |
6172 | ||
6173 | NAME: vary_ignore_expire | |
6174 | COMMENT: on|off | |
6175 | TYPE: onoff | |
6176 | LOC: Config.onoff.vary_ignore_expire | |
6177 | DEFAULT: off | |
6178 | DOC_START | |
6179 | Many HTTP servers supporting Vary gives such objects | |
6180 | immediate expiry time with no cache-control header | |
6181 | when requested by a HTTP/1.0 client. This option | |
6182 | enables Squid to ignore such expiry times until | |
6183 | HTTP/1.1 is fully implemented. | |
6184 | ||
6185 | WARNING: If turned on this may eventually cause some | |
6186 | varying objects not intended for caching to get cached. | |
6187 | DOC_END | |
6188 | ||
6189 | NAME: request_entities | |
6190 | TYPE: onoff | |
6191 | LOC: Config.onoff.request_entities | |
6192 | DEFAULT: off | |
6193 | DOC_START | |
6194 | Squid defaults to deny GET and HEAD requests with request entities, | |
6195 | as the meaning of such requests are undefined in the HTTP standard | |
6196 | even if not explicitly forbidden. | |
6197 | ||
6198 | Set this directive to on if you have clients which insists | |
6199 | on sending request entities in GET or HEAD requests. But be warned | |
6200 | that there is server software (both proxies and web servers) which | |
6201 | can fail to properly process this kind of request which may make you | |
6202 | vulnerable to cache pollution attacks if enabled. | |
6203 | DOC_END | |
6204 | ||
6205 | NAME: request_header_access | |
6206 | IFDEF: USE_HTTP_VIOLATIONS | |
6207 | TYPE: http_header_access | |
6208 | LOC: Config.request_header_access | |
6209 | DEFAULT: none | |
6210 | DEFAULT_DOC: No limits. | |
6211 | DOC_START | |
6212 | Usage: request_header_access header_name allow|deny [!]aclname ... | |
6213 | ||
6214 | WARNING: Doing this VIOLATES the HTTP standard. Enabling | |
6215 | this feature could make you liable for problems which it | |
6216 | causes. | |
6217 | ||
6218 | This option replaces the old 'anonymize_headers' and the | |
6219 | older 'http_anonymizer' option with something that is much | |
6220 | more configurable. A list of ACLs for each header name allows | |
6221 | removal of specific header fields under specific conditions. | |
6222 | ||
6223 | This option only applies to outgoing HTTP request headers (i.e., | |
6224 | headers sent by Squid to the next HTTP hop such as a cache peer | |
6225 | or an origin server). The option has no effect during cache hit | |
6226 | detection. The equivalent adaptation vectoring point in ICAP | |
6227 | terminology is post-cache REQMOD. | |
6228 | ||
6229 | The option is applied to individual outgoing request header | |
6230 | fields. For each request header field F, Squid uses the first | |
6231 | qualifying sets of request_header_access rules: | |
6232 | ||
6233 | 1. Rules with header_name equal to F's name. | |
6234 | 2. Rules with header_name 'Other', provided F's name is not | |
6235 | on the hard-coded list of commonly used HTTP header names. | |
6236 | 3. Rules with header_name 'All'. | |
6237 | ||
6238 | Within that qualifying rule set, rule ACLs are checked as usual. | |
6239 | If ACLs of an "allow" rule match, the header field is allowed to | |
6240 | go through as is. If ACLs of a "deny" rule match, the header is | |
6241 | removed and request_header_replace is then checked to identify | |
6242 | if the removed header has a replacement. If no rules within the | |
6243 | set have matching ACLs, the header field is left as is. | |
6244 | ||
6245 | For example, to achieve the same behavior as the old | |
6246 | 'http_anonymizer standard' option, you should use: | |
6247 | ||
6248 | request_header_access From deny all | |
6249 | request_header_access Referer deny all | |
6250 | request_header_access User-Agent deny all | |
6251 | ||
6252 | Or, to reproduce the old 'http_anonymizer paranoid' feature | |
6253 | you should use: | |
6254 | ||
6255 | request_header_access Authorization allow all | |
6256 | request_header_access Proxy-Authorization allow all | |
6257 | request_header_access Cache-Control allow all | |
6258 | request_header_access Content-Length allow all | |
6259 | request_header_access Content-Type allow all | |
6260 | request_header_access Date allow all | |
6261 | request_header_access Host allow all | |
6262 | request_header_access If-Modified-Since allow all | |
6263 | request_header_access Pragma allow all | |
6264 | request_header_access Accept allow all | |
6265 | request_header_access Accept-Charset allow all | |
6266 | request_header_access Accept-Encoding allow all | |
6267 | request_header_access Accept-Language allow all | |
6268 | request_header_access Connection allow all | |
6269 | request_header_access All deny all | |
6270 | ||
6271 | HTTP reply headers are controlled with the reply_header_access directive. | |
6272 | ||
6273 | By default, all headers are allowed (no anonymizing is performed). | |
6274 | DOC_END | |
6275 | ||
6276 | NAME: reply_header_access | |
6277 | IFDEF: USE_HTTP_VIOLATIONS | |
6278 | TYPE: http_header_access | |
6279 | LOC: Config.reply_header_access | |
6280 | DEFAULT: none | |
6281 | DEFAULT_DOC: No limits. | |
6282 | DOC_START | |
6283 | Usage: reply_header_access header_name allow|deny [!]aclname ... | |
6284 | ||
6285 | WARNING: Doing this VIOLATES the HTTP standard. Enabling | |
6286 | this feature could make you liable for problems which it | |
6287 | causes. | |
6288 | ||
6289 | This option only applies to reply headers, i.e., from the | |
6290 | server to the client. | |
6291 | ||
6292 | This is the same as request_header_access, but in the other | |
6293 | direction. Please see request_header_access for detailed | |
6294 | documentation. | |
6295 | ||
6296 | For example, to achieve the same behavior as the old | |
6297 | 'http_anonymizer standard' option, you should use: | |
6298 | ||
6299 | reply_header_access Server deny all | |
6300 | reply_header_access WWW-Authenticate deny all | |
6301 | reply_header_access Link deny all | |
6302 | ||
6303 | Or, to reproduce the old 'http_anonymizer paranoid' feature | |
6304 | you should use: | |
6305 | ||
6306 | reply_header_access Allow allow all | |
6307 | reply_header_access WWW-Authenticate allow all | |
6308 | reply_header_access Proxy-Authenticate allow all | |
6309 | reply_header_access Cache-Control allow all | |
6310 | reply_header_access Content-Encoding allow all | |
6311 | reply_header_access Content-Length allow all | |
6312 | reply_header_access Content-Type allow all | |
6313 | reply_header_access Date allow all | |
6314 | reply_header_access Expires allow all | |
6315 | reply_header_access Last-Modified allow all | |
6316 | reply_header_access Location allow all | |
6317 | reply_header_access Pragma allow all | |
6318 | reply_header_access Content-Language allow all | |
6319 | reply_header_access Retry-After allow all | |
6320 | reply_header_access Title allow all | |
6321 | reply_header_access Content-Disposition allow all | |
6322 | reply_header_access Connection allow all | |
6323 | reply_header_access All deny all | |
6324 | ||
6325 | HTTP request headers are controlled with the request_header_access directive. | |
6326 | ||
6327 | By default, all headers are allowed (no anonymizing is | |
6328 | performed). | |
6329 | DOC_END | |
6330 | ||
6331 | NAME: request_header_replace header_replace | |
6332 | IFDEF: USE_HTTP_VIOLATIONS | |
6333 | TYPE: http_header_replace | |
6334 | LOC: Config.request_header_access | |
6335 | DEFAULT: none | |
6336 | DOC_START | |
6337 | Usage: request_header_replace header_name message | |
6338 | Example: request_header_replace User-Agent Nutscrape/1.0 (CP/M; 8-bit) | |
6339 | ||
6340 | This option allows you to change the contents of headers | |
6341 | denied with request_header_access above, by replacing them | |
6342 | with some fixed string. | |
6343 | ||
6344 | This only applies to request headers, not reply headers. | |
6345 | ||
6346 | By default, headers are removed if denied. | |
6347 | DOC_END | |
6348 | ||
6349 | NAME: reply_header_replace | |
6350 | IFDEF: USE_HTTP_VIOLATIONS | |
6351 | TYPE: http_header_replace | |
6352 | LOC: Config.reply_header_access | |
6353 | DEFAULT: none | |
6354 | DOC_START | |
6355 | Usage: reply_header_replace header_name message | |
6356 | Example: reply_header_replace Server Foo/1.0 | |
6357 | ||
6358 | This option allows you to change the contents of headers | |
6359 | denied with reply_header_access above, by replacing them | |
6360 | with some fixed string. | |
6361 | ||
6362 | This only applies to reply headers, not request headers. | |
6363 | ||
6364 | By default, headers are removed if denied. | |
6365 | DOC_END | |
6366 | ||
6367 | NAME: request_header_add | |
6368 | TYPE: HeaderWithAclList | |
6369 | LOC: Config.request_header_add | |
6370 | DEFAULT: none | |
6371 | DOC_START | |
6372 | Usage: request_header_add field-name field-value [ acl ... ] | |
6373 | Example: request_header_add X-Client-CA "CA=%ssl::>cert_issuer" all | |
6374 | ||
6375 | This option adds header fields to outgoing HTTP requests (i.e., | |
6376 | request headers sent by Squid to the next HTTP hop such as a | |
6377 | cache peer or an origin server). The option has no effect during | |
6378 | cache hit detection. The equivalent adaptation vectoring point | |
6379 | in ICAP terminology is post-cache REQMOD. | |
6380 | ||
6381 | Field-name is a token specifying an HTTP header name. If a | |
6382 | standard HTTP header name is used, Squid does not check whether | |
6383 | the new header conflicts with any existing headers or violates | |
6384 | HTTP rules. If the request to be modified already contains a | |
6385 | field with the same name, the old field is preserved but the | |
6386 | header field values are not merged. | |
6387 | ||
6388 | Field-value is either a token or a quoted string. If quoted | |
6389 | string format is used, then the surrounding quotes are removed | |
6390 | while escape sequences and %macros are processed. | |
6391 | ||
6392 | One or more Squid ACLs may be specified to restrict header | |
6393 | injection to matching requests. As always in squid.conf, all | |
6394 | ACLs in the ACL list must be satisfied for the insertion to | |
6395 | happen. The request_header_add supports fast ACLs only. | |
6396 | ||
6397 | See also: reply_header_add. | |
6398 | DOC_END | |
6399 | ||
6400 | NAME: reply_header_add | |
6401 | TYPE: HeaderWithAclList | |
6402 | LOC: Config.reply_header_add | |
6403 | DEFAULT: none | |
6404 | DOC_START | |
6405 | Usage: reply_header_add field-name field-value [ acl ... ] | |
6406 | Example: reply_header_add X-Client-CA "CA=%ssl::>cert_issuer" all | |
6407 | ||
6408 | This option adds header fields to outgoing HTTP responses (i.e., response | |
6409 | headers delivered by Squid to the client). This option has no effect on | |
6410 | cache hit detection. The equivalent adaptation vectoring point in | |
6411 | ICAP terminology is post-cache RESPMOD. This option does not apply to | |
6412 | successful CONNECT replies. | |
6413 | ||
6414 | Field-name is a token specifying an HTTP header name. If a | |
6415 | standard HTTP header name is used, Squid does not check whether | |
6416 | the new header conflicts with any existing headers or violates | |
6417 | HTTP rules. If the response to be modified already contains a | |
6418 | field with the same name, the old field is preserved but the | |
6419 | header field values are not merged. | |
6420 | ||
6421 | Field-value is either a token or a quoted string. If quoted | |
6422 | string format is used, then the surrounding quotes are removed | |
6423 | while escape sequences and %macros are processed. | |
6424 | ||
6425 | One or more Squid ACLs may be specified to restrict header | |
6426 | injection to matching responses. As always in squid.conf, all | |
6427 | ACLs in the ACL list must be satisfied for the insertion to | |
6428 | happen. The reply_header_add option supports fast ACLs only. | |
6429 | ||
6430 | See also: request_header_add. | |
6431 | DOC_END | |
6432 | ||
6433 | NAME: note | |
6434 | TYPE: note | |
6435 | LOC: Config.notes | |
6436 | DEFAULT: none | |
6437 | DOC_START | |
6438 | This option used to log custom information about the master | |
6439 | transaction. For example, an admin may configure Squid to log | |
6440 | which "user group" the transaction belongs to, where "user group" | |
6441 | will be determined based on a set of ACLs and not [just] | |
6442 | authentication information. | |
6443 | Values of key/value pairs can be logged using %{key}note macros: | |
6444 | ||
6445 | note key value acl ... | |
6446 | logformat myFormat ... %{key}note ... | |
6447 | DOC_END | |
6448 | ||
6449 | NAME: relaxed_header_parser | |
6450 | COMMENT: on|off|warn | |
6451 | TYPE: tristate | |
6452 | LOC: Config.onoff.relaxed_header_parser | |
6453 | DEFAULT: on | |
6454 | DOC_START | |
6455 | In the default "on" setting Squid accepts certain forms | |
6456 | of non-compliant HTTP messages where it is unambiguous | |
6457 | what the sending application intended even if the message | |
6458 | is not correctly formatted. The messages is then normalized | |
6459 | to the correct form when forwarded by Squid. | |
6460 | ||
6461 | If set to "warn" then a warning will be emitted in cache.log | |
6462 | each time such HTTP error is encountered. | |
6463 | ||
6464 | If set to "off" then such HTTP errors will cause the request | |
6465 | or response to be rejected. | |
6466 | DOC_END | |
6467 | ||
6468 | NAME: collapsed_forwarding | |
6469 | COMMENT: (on|off) | |
6470 | TYPE: onoff | |
6471 | LOC: Config.onoff.collapsed_forwarding | |
6472 | DEFAULT: off | |
6473 | DOC_START | |
6474 | This option controls whether Squid is allowed to merge multiple | |
6475 | potentially cachable requests for the same URI before Squid knows | |
6476 | whether the response is going to be cachable. | |
6477 | ||
6478 | When enabled, instead of forwarding each concurrent request for | |
6479 | the same URL, Squid just sends the first of them. The other, so | |
6480 | called "collapsed" requests, wait for the response to the first | |
6481 | request and, if it happens to be cachable, use that response. | |
6482 | Here, "concurrent requests" means "received after the first | |
6483 | request headers were parsed and before the corresponding response | |
6484 | headers were parsed". | |
6485 | ||
6486 | This feature is disabled by default: enabling collapsed | |
6487 | forwarding needlessly delays forwarding requests that look | |
6488 | cachable (when they are collapsed) but then need to be forwarded | |
6489 | individually anyway because they end up being for uncachable | |
6490 | content. However, in some cases, such as acceleration of highly | |
6491 | cachable content with periodic or grouped expiration times, the | |
6492 | gains from collapsing [large volumes of simultaneous refresh | |
6493 | requests] outweigh losses from such delays. | |
6494 | ||
6495 | Squid collapses two kinds of requests: regular client requests | |
6496 | received on one of the listening ports and internal "cache | |
6497 | revalidation" requests which are triggered by those regular | |
6498 | requests hitting a stale cached object. Revalidation collapsing | |
6499 | is currently disabled for Squid instances containing SMP-aware | |
6500 | disk or memory caches and for Vary-controlled cached objects. | |
6501 | DOC_END | |
6502 | ||
6503 | NAME: collapsed_forwarding_shared_entries_limit | |
6504 | COMMENT: (number of entries) | |
6505 | TYPE: int64_t | |
6506 | LOC: Config.collapsed_forwarding_shared_entries_limit | |
6507 | DEFAULT: 16384 | |
6508 | DOC_START | |
6509 | This limits the size of a table used for sharing information | |
6510 | about collapsible entries among SMP workers. Limiting sharing | |
6511 | too much results in cache content duplication and missed | |
6512 | collapsing opportunities. Using excessively large values | |
6513 | wastes shared memory. | |
6514 | ||
6515 | The limit should be significantly larger then the number of | |
6516 | concurrent collapsible entries one wants to share. For a cache | |
6517 | that handles less than 5000 concurrent requests, the default | |
6518 | setting of 16384 should be plenty. | |
6519 | ||
6520 | If the limit is set to zero, it disables sharing of collapsed | |
6521 | forwarding between SMP workers. | |
6522 | DOC_END | |
6523 | ||
6524 | COMMENT_START | |
6525 | TIMEOUTS | |
6526 | ----------------------------------------------------------------------------- | |
6527 | COMMENT_END | |
6528 | ||
6529 | NAME: forward_timeout | |
6530 | COMMENT: time-units | |
6531 | TYPE: time_t | |
6532 | LOC: Config.Timeout.forward | |
6533 | DEFAULT: 4 minutes | |
6534 | DOC_START | |
6535 | This parameter specifies how long Squid should at most attempt in | |
6536 | finding a forwarding path for the request before giving up. | |
6537 | DOC_END | |
6538 | ||
6539 | NAME: connect_timeout | |
6540 | COMMENT: time-units | |
6541 | TYPE: time_t | |
6542 | LOC: Config.Timeout.connect | |
6543 | DEFAULT: 1 minute | |
6544 | DOC_START | |
6545 | This parameter specifies how long to wait for the TCP connect to | |
6546 | the requested server or peer to complete before Squid should | |
6547 | attempt to find another path where to forward the request. | |
6548 | DOC_END | |
6549 | ||
6550 | NAME: peer_connect_timeout | |
6551 | COMMENT: time-units | |
6552 | TYPE: time_t | |
6553 | LOC: Config.Timeout.peer_connect | |
6554 | DEFAULT: 30 seconds | |
6555 | DOC_START | |
6556 | This parameter specifies how long to wait for a pending TCP | |
6557 | connection to a peer cache. The default is 30 seconds. You | |
6558 | may also set different timeout values for individual neighbors | |
6559 | with the 'connect-timeout' option on a 'cache_peer' line. | |
6560 | DOC_END | |
6561 | ||
6562 | NAME: read_timeout | |
6563 | COMMENT: time-units | |
6564 | TYPE: time_t | |
6565 | LOC: Config.Timeout.read | |
6566 | DEFAULT: 15 minutes | |
6567 | DOC_START | |
6568 | Applied on peer server connections. | |
6569 | ||
6570 | After each successful read(), the timeout will be extended by this | |
6571 | amount. If no data is read again after this amount of time, | |
6572 | the request is aborted and logged with ERR_READ_TIMEOUT. | |
6573 | ||
6574 | The default is 15 minutes. | |
6575 | DOC_END | |
6576 | ||
6577 | NAME: write_timeout | |
6578 | COMMENT: time-units | |
6579 | TYPE: time_t | |
6580 | LOC: Config.Timeout.write | |
6581 | DEFAULT: 15 minutes | |
6582 | DOC_START | |
6583 | This timeout is tracked for all connections that have data | |
6584 | available for writing and are waiting for the socket to become | |
6585 | ready. After each successful write, the timeout is extended by | |
6586 | the configured amount. If Squid has data to write but the | |
6587 | connection is not ready for the configured duration, the | |
6588 | transaction associated with the connection is terminated. The | |
6589 | default is 15 minutes. | |
6590 | DOC_END | |
6591 | ||
6592 | NAME: request_timeout | |
6593 | TYPE: time_t | |
6594 | LOC: Config.Timeout.request | |
6595 | DEFAULT: 5 minutes | |
6596 | DOC_START | |
6597 | How long to wait for complete HTTP request headers after initial | |
6598 | connection establishment. | |
6599 | DOC_END | |
6600 | ||
6601 | NAME: request_start_timeout | |
6602 | TYPE: time_t | |
6603 | LOC: Config.Timeout.request_start_timeout | |
6604 | DEFAULT: 5 minutes | |
6605 | DOC_START | |
6606 | How long to wait for the first request byte after initial | |
6607 | connection establishment. | |
6608 | DOC_END | |
6609 | ||
6610 | NAME: client_idle_pconn_timeout persistent_request_timeout | |
6611 | TYPE: time_t | |
6612 | LOC: Config.Timeout.clientIdlePconn | |
6613 | DEFAULT: 2 minutes | |
6614 | DOC_START | |
6615 | How long to wait for the next HTTP request on a persistent | |
6616 | client connection after the previous request completes. | |
6617 | DOC_END | |
6618 | ||
6619 | NAME: ftp_client_idle_timeout | |
6620 | TYPE: time_t | |
6621 | LOC: Config.Timeout.ftpClientIdle | |
6622 | DEFAULT: 30 minutes | |
6623 | DOC_START | |
6624 | How long to wait for an FTP request on a connection to Squid ftp_port. | |
6625 | Many FTP clients do not deal with idle connection closures well, | |
6626 | necessitating a longer default timeout than client_idle_pconn_timeout | |
6627 | used for incoming HTTP requests. | |
6628 | DOC_END | |
6629 | ||
6630 | NAME: client_lifetime | |
6631 | COMMENT: time-units | |
6632 | TYPE: time_t | |
6633 | LOC: Config.Timeout.lifetime | |
6634 | DEFAULT: 1 day | |
6635 | DOC_START | |
6636 | The maximum amount of time a client (browser) is allowed to | |
6637 | remain connected to the cache process. This protects the Cache | |
6638 | from having a lot of sockets (and hence file descriptors) tied up | |
6639 | in a CLOSE_WAIT state from remote clients that go away without | |
6640 | properly shutting down (either because of a network failure or | |
6641 | because of a poor client implementation). The default is one | |
6642 | day, 1440 minutes. | |
6643 | ||
6644 | NOTE: The default value is intended to be much larger than any | |
6645 | client would ever need to be connected to your cache. You | |
6646 | should probably change client_lifetime only as a last resort. | |
6647 | If you seem to have many client connections tying up | |
6648 | filedescriptors, we recommend first tuning the read_timeout, | |
6649 | request_timeout, persistent_request_timeout and quick_abort values. | |
6650 | DOC_END | |
6651 | ||
6652 | NAME: pconn_lifetime | |
6653 | COMMENT: time-units | |
6654 | TYPE: time_t | |
6655 | LOC: Config.Timeout.pconnLifetime | |
6656 | DEFAULT: 0 seconds | |
6657 | DOC_START | |
6658 | Desired maximum lifetime of a persistent connection. | |
6659 | When set, Squid will close a now-idle persistent connection that | |
6660 | exceeded configured lifetime instead of moving the connection into | |
6661 | the idle connection pool (or equivalent). No effect on ongoing/active | |
6662 | transactions. Connection lifetime is the time period from the | |
6663 | connection acceptance or opening time until "now". | |
6664 | ||
6665 | This limit is useful in environments with long-lived connections | |
6666 | where Squid configuration or environmental factors change during a | |
6667 | single connection lifetime. If unrestricted, some connections may | |
6668 | last for hours and even days, ignoring those changes that should | |
6669 | have affected their behavior or their existence. | |
6670 | ||
6671 | Currently, a new lifetime value supplied via Squid reconfiguration | |
6672 | has no effect on already idle connections unless they become busy. | |
6673 | ||
6674 | When set to '0' this limit is not used. | |
6675 | DOC_END | |
6676 | ||
6677 | NAME: half_closed_clients | |
6678 | TYPE: onoff | |
6679 | LOC: Config.onoff.half_closed_clients | |
6680 | DEFAULT: off | |
6681 | DOC_START | |
6682 | Some clients may shutdown the sending side of their TCP | |
6683 | connections, while leaving their receiving sides open. Sometimes, | |
6684 | Squid can not tell the difference between a half-closed and a | |
6685 | fully-closed TCP connection. | |
6686 | ||
6687 | By default, Squid will immediately close client connections when | |
6688 | read(2) returns "no more data to read." | |
6689 | ||
6690 | Change this option to 'on' and Squid will keep open connections | |
6691 | until a read(2) or write(2) on the socket returns an error. | |
6692 | This may show some benefits for reverse proxies. But if not | |
6693 | it is recommended to leave OFF. | |
6694 | DOC_END | |
6695 | ||
6696 | NAME: server_idle_pconn_timeout pconn_timeout | |
6697 | TYPE: time_t | |
6698 | LOC: Config.Timeout.serverIdlePconn | |
6699 | DEFAULT: 1 minute | |
6700 | DOC_START | |
6701 | Timeout for idle persistent connections to servers and other | |
6702 | proxies. | |
6703 | DOC_END | |
6704 | ||
6705 | NAME: ident_timeout | |
6706 | TYPE: time_t | |
6707 | IFDEF: USE_IDENT | |
6708 | LOC: Ident::TheConfig.timeout | |
6709 | DEFAULT: 10 seconds | |
6710 | DOC_START | |
6711 | Maximum time to wait for IDENT lookups to complete. | |
6712 | ||
6713 | If this is too high, and you enabled IDENT lookups from untrusted | |
6714 | users, you might be susceptible to denial-of-service by having | |
6715 | many ident requests going at once. | |
6716 | DOC_END | |
6717 | ||
6718 | NAME: shutdown_lifetime | |
6719 | COMMENT: time-units | |
6720 | TYPE: time_t | |
6721 | LOC: Config.shutdownLifetime | |
6722 | DEFAULT: 30 seconds | |
6723 | DOC_START | |
6724 | When SIGTERM or SIGHUP is received, the cache is put into | |
6725 | "shutdown pending" mode until all active sockets are closed. | |
6726 | This value is the lifetime to set for all open descriptors | |
6727 | during shutdown mode. Any active clients after this many | |
6728 | seconds will receive a 'timeout' message. | |
6729 | DOC_END | |
6730 | ||
6731 | COMMENT_START | |
6732 | ADMINISTRATIVE PARAMETERS | |
6733 | ----------------------------------------------------------------------------- | |
6734 | COMMENT_END | |
6735 | ||
6736 | NAME: cache_mgr | |
6737 | TYPE: string | |
6738 | DEFAULT: webmaster | |
6739 | LOC: Config.adminEmail | |
6740 | DOC_START | |
6741 | Email-address of local cache manager who will receive | |
6742 | mail if the cache dies. The default is "webmaster". | |
6743 | DOC_END | |
6744 | ||
6745 | NAME: mail_from | |
6746 | TYPE: string | |
6747 | DEFAULT: none | |
6748 | LOC: Config.EmailFrom | |
6749 | DOC_START | |
6750 | From: email-address for mail sent when the cache dies. | |
6751 | The default is to use 'squid@unique_hostname'. | |
6752 | ||
6753 | See also: unique_hostname directive. | |
6754 | DOC_END | |
6755 | ||
6756 | NAME: mail_program | |
6757 | TYPE: eol | |
6758 | DEFAULT: mail | |
6759 | LOC: Config.EmailProgram | |
6760 | DOC_START | |
6761 | Email program used to send mail if the cache dies. | |
6762 | The default is "mail". The specified program must comply | |
6763 | with the standard Unix mail syntax: | |
6764 | mail-program recipient < mailfile | |
6765 | ||
6766 | Optional command line options can be specified. | |
6767 | DOC_END | |
6768 | ||
6769 | NAME: cache_effective_user | |
6770 | TYPE: string | |
6771 | DEFAULT: @DEFAULT_CACHE_EFFECTIVE_USER@ | |
6772 | LOC: Config.effectiveUser | |
6773 | DOC_START | |
6774 | If you start Squid as root, it will change its effective/real | |
6775 | UID/GID to the user specified below. The default is to change | |
6776 | to UID of @DEFAULT_CACHE_EFFECTIVE_USER@. | |
6777 | see also; cache_effective_group | |
6778 | DOC_END | |
6779 | ||
6780 | NAME: cache_effective_group | |
6781 | TYPE: string | |
6782 | DEFAULT: none | |
6783 | DEFAULT_DOC: Use system group memberships of the cache_effective_user account | |
6784 | LOC: Config.effectiveGroup | |
6785 | DOC_START | |
6786 | Squid sets the GID to the effective user's default group ID | |
6787 | (taken from the password file) and supplementary group list | |
6788 | from the groups membership. | |
6789 | ||
6790 | If you want Squid to run with a specific GID regardless of | |
6791 | the group memberships of the effective user then set this | |
6792 | to the group (or GID) you want Squid to run as. When set | |
6793 | all other group privileges of the effective user are ignored | |
6794 | and only this GID is effective. If Squid is not started as | |
6795 | root the user starting Squid MUST be member of the specified | |
6796 | group. | |
6797 | ||
6798 | This option is not recommended by the Squid Team. | |
6799 | Our preference is for administrators to configure a secure | |
6800 | user account for squid with UID/GID matching system policies. | |
6801 | DOC_END | |
6802 | ||
6803 | NAME: httpd_suppress_version_string | |
6804 | COMMENT: on|off | |
6805 | TYPE: onoff | |
6806 | DEFAULT: off | |
6807 | LOC: Config.onoff.httpd_suppress_version_string | |
6808 | DOC_START | |
6809 | Suppress Squid version string info in HTTP headers and HTML error pages. | |
6810 | DOC_END | |
6811 | ||
6812 | NAME: visible_hostname | |
6813 | TYPE: string | |
6814 | LOC: Config.visibleHostname | |
6815 | DEFAULT: none | |
6816 | DEFAULT_DOC: Automatically detect the system host name | |
6817 | DOC_START | |
6818 | If you want to present a special hostname in error messages, etc, | |
6819 | define this. Otherwise, the return value of gethostname() | |
6820 | will be used. If you have multiple caches in a cluster and | |
6821 | get errors about IP-forwarding you must set them to have individual | |
6822 | names with this setting. | |
6823 | DOC_END | |
6824 | ||
6825 | NAME: unique_hostname | |
6826 | TYPE: string | |
6827 | LOC: Config.uniqueHostname | |
6828 | DEFAULT: none | |
6829 | DEFAULT_DOC: Copy the value from visible_hostname | |
6830 | DOC_START | |
6831 | If you want to have multiple machines with the same | |
6832 | 'visible_hostname' you must give each machine a different | |
6833 | 'unique_hostname' so forwarding loops can be detected. | |
6834 | DOC_END | |
6835 | ||
6836 | NAME: hostname_aliases | |
6837 | TYPE: wordlist | |
6838 | LOC: Config.hostnameAliases | |
6839 | DEFAULT: none | |
6840 | DOC_START | |
6841 | A list of other DNS names your cache has. | |
6842 | DOC_END | |
6843 | ||
6844 | NAME: umask | |
6845 | TYPE: int | |
6846 | LOC: Config.umask | |
6847 | DEFAULT: 027 | |
6848 | DOC_START | |
6849 | Minimum umask which should be enforced while the proxy | |
6850 | is running, in addition to the umask set at startup. | |
6851 | ||
6852 | For a traditional octal representation of umasks, start | |
6853 | your value with 0. | |
6854 | DOC_END | |
6855 | ||
6856 | COMMENT_START | |
6857 | OPTIONS FOR THE CACHE REGISTRATION SERVICE | |
6858 | ----------------------------------------------------------------------------- | |
6859 | ||
6860 | This section contains parameters for the (optional) cache | |
6861 | announcement service. This service is provided to help | |
6862 | cache administrators locate one another in order to join or | |
6863 | create cache hierarchies. | |
6864 | ||
6865 | An 'announcement' message is sent (via UDP) to the registration | |
6866 | service by Squid. By default, the announcement message is NOT | |
6867 | SENT unless you enable it with 'announce_period' below. | |
6868 | ||
6869 | The announcement message includes your hostname, plus the | |
6870 | following information from this configuration file: | |
6871 | ||
6872 | http_port | |
6873 | icp_port | |
6874 | cache_mgr | |
6875 | ||
6876 | All current information is processed regularly and made | |
6877 | available on the Web at http://www.ircache.net/Cache/Tracker/. | |
6878 | COMMENT_END | |
6879 | ||
6880 | NAME: announce_period | |
6881 | TYPE: time_t | |
6882 | LOC: Config.Announce.period | |
6883 | DEFAULT: 0 | |
6884 | DEFAULT_DOC: Announcement messages disabled. | |
6885 | DOC_START | |
6886 | This is how frequently to send cache announcements. | |
6887 | ||
6888 | To enable announcing your cache, just set an announce period. | |
6889 | ||
6890 | Example: | |
6891 | announce_period 1 day | |
6892 | DOC_END | |
6893 | ||
6894 | NAME: announce_host | |
6895 | TYPE: string | |
6896 | DEFAULT: tracker.ircache.net | |
6897 | LOC: Config.Announce.host | |
6898 | DOC_START | |
6899 | Set the hostname where announce registration messages will be sent. | |
6900 | ||
6901 | See also announce_port and announce_file | |
6902 | DOC_END | |
6903 | ||
6904 | NAME: announce_file | |
6905 | TYPE: string | |
6906 | DEFAULT: none | |
6907 | LOC: Config.Announce.file | |
6908 | DOC_START | |
6909 | The contents of this file will be included in the announce | |
6910 | registration messages. | |
6911 | DOC_END | |
6912 | ||
6913 | NAME: announce_port | |
6914 | TYPE: u_short | |
6915 | DEFAULT: 3131 | |
6916 | LOC: Config.Announce.port | |
6917 | DOC_START | |
6918 | Set the port where announce registration messages will be sent. | |
6919 | ||
6920 | See also announce_host and announce_file | |
6921 | DOC_END | |
6922 | ||
6923 | COMMENT_START | |
6924 | HTTPD-ACCELERATOR OPTIONS | |
6925 | ----------------------------------------------------------------------------- | |
6926 | COMMENT_END | |
6927 | ||
6928 | NAME: httpd_accel_surrogate_id | |
6929 | TYPE: string | |
6930 | DEFAULT: none | |
6931 | DEFAULT_DOC: visible_hostname is used if no specific ID is set. | |
6932 | LOC: Config.Accel.surrogate_id | |
6933 | DOC_START | |
6934 | Surrogates (http://www.esi.org/architecture_spec_1.0.html) | |
6935 | need an identification token to allow control targeting. Because | |
6936 | a farm of surrogates may all perform the same tasks, they may share | |
6937 | an identification token. | |
6938 | DOC_END | |
6939 | ||
6940 | NAME: http_accel_surrogate_remote | |
6941 | COMMENT: on|off | |
6942 | TYPE: onoff | |
6943 | DEFAULT: off | |
6944 | LOC: Config.onoff.surrogate_is_remote | |
6945 | DOC_START | |
6946 | Remote surrogates (such as those in a CDN) honour the header | |
6947 | "Surrogate-Control: no-store-remote". | |
6948 | ||
6949 | Set this to on to have squid behave as a remote surrogate. | |
6950 | DOC_END | |
6951 | ||
6952 | NAME: esi_parser | |
6953 | IFDEF: USE_SQUID_ESI | |
6954 | COMMENT: libxml2|expat|custom | |
6955 | TYPE: string | |
6956 | LOC: ESIParser::Type | |
6957 | DEFAULT: custom | |
6958 | DOC_START | |
6959 | ESI markup is not strictly XML compatible. The custom ESI parser | |
6960 | will give higher performance, but cannot handle non ASCII character | |
6961 | encodings. | |
6962 | DOC_END | |
6963 | ||
6964 | COMMENT_START | |
6965 | DELAY POOL PARAMETERS | |
6966 | ----------------------------------------------------------------------------- | |
6967 | COMMENT_END | |
6968 | ||
6969 | NAME: delay_pools | |
6970 | TYPE: delay_pool_count | |
6971 | DEFAULT: 0 | |
6972 | IFDEF: USE_DELAY_POOLS | |
6973 | LOC: Config.Delay | |
6974 | DOC_START | |
6975 | This represents the number of delay pools to be used. For example, | |
6976 | if you have one class 2 delay pool and one class 3 delays pool, you | |
6977 | have a total of 2 delay pools. | |
6978 | ||
6979 | See also delay_parameters, delay_class, delay_access for pool | |
6980 | configuration details. | |
6981 | DOC_END | |
6982 | ||
6983 | NAME: delay_class | |
6984 | TYPE: delay_pool_class | |
6985 | DEFAULT: none | |
6986 | IFDEF: USE_DELAY_POOLS | |
6987 | LOC: Config.Delay | |
6988 | DOC_START | |
6989 | This defines the class of each delay pool. There must be exactly one | |
6990 | delay_class line for each delay pool. For example, to define two | |
6991 | delay pools, one of class 2 and one of class 3, the settings above | |
6992 | and here would be: | |
6993 | ||
6994 | Example: | |
6995 | delay_pools 4 # 4 delay pools | |
6996 | delay_class 1 2 # pool 1 is a class 2 pool | |
6997 | delay_class 2 3 # pool 2 is a class 3 pool | |
6998 | delay_class 3 4 # pool 3 is a class 4 pool | |
6999 | delay_class 4 5 # pool 4 is a class 5 pool | |
7000 | ||
7001 | The delay pool classes are: | |
7002 | ||
7003 | class 1 Everything is limited by a single aggregate | |
7004 | bucket. | |
7005 | ||
7006 | class 2 Everything is limited by a single aggregate | |
7007 | bucket as well as an "individual" bucket chosen | |
7008 | from bits 25 through 32 of the IPv4 address. | |
7009 | ||
7010 | class 3 Everything is limited by a single aggregate | |
7011 | bucket as well as a "network" bucket chosen | |
7012 | from bits 17 through 24 of the IP address and a | |
7013 | "individual" bucket chosen from bits 17 through | |
7014 | 32 of the IPv4 address. | |
7015 | ||
7016 | class 4 Everything in a class 3 delay pool, with an | |
7017 | additional limit on a per user basis. This | |
7018 | only takes effect if the username is established | |
7019 | in advance - by forcing authentication in your | |
7020 | http_access rules. | |
7021 | ||
7022 | class 5 Requests are grouped according their tag (see | |
7023 | external_acl's tag= reply). | |
7024 | ||
7025 | ||
7026 | Each pool also requires a delay_parameters directive to configure the pool size | |
7027 | and speed limits used whenever the pool is applied to a request. Along with | |
7028 | a set of delay_access directives to determine when it is used. | |
7029 | ||
7030 | NOTE: If an IP address is a.b.c.d | |
7031 | -> bits 25 through 32 are "d" | |
7032 | -> bits 17 through 24 are "c" | |
7033 | -> bits 17 through 32 are "c * 256 + d" | |
7034 | ||
7035 | NOTE-2: Due to the use of bitmasks in class 2,3,4 pools they only apply to | |
7036 | IPv4 traffic. Class 1 and 5 pools may be used with IPv6 traffic. | |
7037 | ||
7038 | This clause only supports fast acl types. | |
7039 | See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details. | |
7040 | ||
7041 | See also delay_parameters and delay_access. | |
7042 | DOC_END | |
7043 | ||
7044 | NAME: delay_access | |
7045 | TYPE: delay_pool_access | |
7046 | DEFAULT: none | |
7047 | DEFAULT_DOC: Deny using the pool, unless allow rules exist in squid.conf for the pool. | |
7048 | IFDEF: USE_DELAY_POOLS | |
7049 | LOC: Config.Delay | |
7050 | DOC_START | |
7051 | This is used to determine which delay pool a request falls into. | |
7052 | ||
7053 | delay_access is sorted per pool and the matching starts with pool 1, | |
7054 | then pool 2, ..., and finally pool N. The first delay pool where the | |
7055 | request is allowed is selected for the request. If it does not allow | |
7056 | the request to any pool then the request is not delayed (default). | |
7057 | ||
7058 | For example, if you want some_big_clients in delay | |
7059 | pool 1 and lotsa_little_clients in delay pool 2: | |
7060 | ||
7061 | delay_access 1 allow some_big_clients | |
7062 | delay_access 1 deny all | |
7063 | delay_access 2 allow lotsa_little_clients | |
7064 | delay_access 2 deny all | |
7065 | delay_access 3 allow authenticated_clients | |
7066 | ||
7067 | See also delay_parameters and delay_class. | |
7068 | ||
7069 | DOC_END | |
7070 | ||
7071 | NAME: delay_parameters | |
7072 | TYPE: delay_pool_rates | |
7073 | DEFAULT: none | |
7074 | IFDEF: USE_DELAY_POOLS | |
7075 | LOC: Config.Delay | |
7076 | DOC_START | |
7077 | This defines the parameters for a delay pool. Each delay pool has | |
7078 | a number of "buckets" associated with it, as explained in the | |
7079 | description of delay_class. | |
7080 | ||
7081 | For a class 1 delay pool, the syntax is: | |
7082 | delay_class pool 1 | |
7083 | delay_parameters pool aggregate | |
7084 | ||
7085 | For a class 2 delay pool: | |
7086 | delay_class pool 2 | |
7087 | delay_parameters pool aggregate individual | |
7088 | ||
7089 | For a class 3 delay pool: | |
7090 | delay_class pool 3 | |
7091 | delay_parameters pool aggregate network individual | |
7092 | ||
7093 | For a class 4 delay pool: | |
7094 | delay_class pool 4 | |
7095 | delay_parameters pool aggregate network individual user | |
7096 | ||
7097 | For a class 5 delay pool: | |
7098 | delay_class pool 5 | |
7099 | delay_parameters pool tagrate | |
7100 | ||
7101 | The option variables are: | |
7102 | ||
7103 | pool a pool number - ie, a number between 1 and the | |
7104 | number specified in delay_pools as used in | |
7105 | delay_class lines. | |
7106 | ||
7107 | aggregate the speed limit parameters for the aggregate bucket | |
7108 | (class 1, 2, 3). | |
7109 | ||
7110 | individual the speed limit parameters for the individual | |
7111 | buckets (class 2, 3). | |
7112 | ||
7113 | network the speed limit parameters for the network buckets | |
7114 | (class 3). | |
7115 | ||
7116 | user the speed limit parameters for the user buckets | |
7117 | (class 4). | |
7118 | ||
7119 | tagrate the speed limit parameters for the tag buckets | |
7120 | (class 5). | |
7121 | ||
7122 | A pair of delay parameters is written restore/maximum, where restore is | |
7123 | the number of bytes (not bits - modem and network speeds are usually | |
7124 | quoted in bits) per second placed into the bucket, and maximum is the | |
7125 | maximum number of bytes which can be in the bucket at any time. | |
7126 | ||
7127 | There must be one delay_parameters line for each delay pool. | |
7128 | ||
7129 | ||
7130 | For example, if delay pool number 1 is a class 2 delay pool as in the | |
7131 | above example, and is being used to strictly limit each host to 64Kbit/sec | |
7132 | (plus overheads), with no overall limit, the line is: | |
7133 | ||
7134 | delay_parameters 1 none 8000/8000 | |
7135 | ||
7136 | Note that 8 x 8K Byte/sec -> 64K bit/sec. | |
7137 | ||
7138 | Note that the word 'none' is used to represent no limit. | |
7139 | ||
7140 | ||
7141 | And, if delay pool number 2 is a class 3 delay pool as in the above | |
7142 | example, and you want to limit it to a total of 256Kbit/sec (strict limit) | |
7143 | with each 8-bit network permitted 64Kbit/sec (strict limit) and each | |
7144 | individual host permitted 4800bit/sec with a bucket maximum size of 64Kbits | |
7145 | to permit a decent web page to be downloaded at a decent speed | |
7146 | (if the network is not being limited due to overuse) but slow down | |
7147 | large downloads more significantly: | |
7148 | ||
7149 | delay_parameters 2 32000/32000 8000/8000 600/8000 | |
7150 | ||
7151 | Note that 8 x 32K Byte/sec -> 256K bit/sec. | |
7152 | 8 x 8K Byte/sec -> 64K bit/sec. | |
7153 | 8 x 600 Byte/sec -> 4800 bit/sec. | |
7154 | ||
7155 | ||
7156 | Finally, for a class 4 delay pool as in the example - each user will | |
7157 | be limited to 128Kbits/sec no matter how many workstations they are logged into.: | |
7158 | ||
7159 | delay_parameters 4 32000/32000 8000/8000 600/64000 16000/16000 | |
7160 | ||
7161 | ||
7162 | See also delay_class and delay_access. | |
7163 | ||
7164 | DOC_END | |
7165 | ||
7166 | NAME: delay_initial_bucket_level | |
7167 | COMMENT: (percent, 0-100) | |
7168 | TYPE: u_short | |
7169 | DEFAULT: 50 | |
7170 | IFDEF: USE_DELAY_POOLS | |
7171 | LOC: Config.Delay.initial | |
7172 | DOC_START | |
7173 | The initial bucket percentage is used to determine how much is put | |
7174 | in each bucket when squid starts, is reconfigured, or first notices | |
7175 | a host accessing it (in class 2 and class 3, individual hosts and | |
7176 | networks only have buckets associated with them once they have been | |
7177 | "seen" by squid). | |
7178 | DOC_END | |
7179 | ||
7180 | COMMENT_START | |
7181 | CLIENT DELAY POOL PARAMETERS | |
7182 | ----------------------------------------------------------------------------- | |
7183 | COMMENT_END | |
7184 | ||
7185 | NAME: client_delay_pools | |
7186 | TYPE: client_delay_pool_count | |
7187 | DEFAULT: 0 | |
7188 | IFDEF: USE_DELAY_POOLS | |
7189 | LOC: Config.ClientDelay | |
7190 | DOC_START | |
7191 | This option specifies the number of client delay pools used. It must | |
7192 | preceed other client_delay_* options. | |
7193 | ||
7194 | Example: | |
7195 | client_delay_pools 2 | |
7196 | ||
7197 | See also client_delay_parameters and client_delay_access. | |
7198 | DOC_END | |
7199 | ||
7200 | NAME: client_delay_initial_bucket_level | |
7201 | COMMENT: (percent, 0-no_limit) | |
7202 | TYPE: u_short | |
7203 | DEFAULT: 50 | |
7204 | IFDEF: USE_DELAY_POOLS | |
7205 | LOC: Config.ClientDelay.initial | |
7206 | DOC_START | |
7207 | This option determines the initial bucket size as a percentage of | |
7208 | max_bucket_size from client_delay_parameters. Buckets are created | |
7209 | at the time of the "first" connection from the matching IP. Idle | |
7210 | buckets are periodically deleted up. | |
7211 | ||
7212 | You can specify more than 100 percent but note that such "oversized" | |
7213 | buckets are not refilled until their size goes down to max_bucket_size | |
7214 | from client_delay_parameters. | |
7215 | ||
7216 | Example: | |
7217 | client_delay_initial_bucket_level 50 | |
7218 | DOC_END | |
7219 | ||
7220 | NAME: client_delay_parameters | |
7221 | TYPE: client_delay_pool_rates | |
7222 | DEFAULT: none | |
7223 | IFDEF: USE_DELAY_POOLS | |
7224 | LOC: Config.ClientDelay | |
7225 | DOC_START | |
7226 | ||
7227 | This option configures client-side bandwidth limits using the | |
7228 | following format: | |
7229 | ||
7230 | client_delay_parameters pool speed_limit max_bucket_size | |
7231 | ||
7232 | pool is an integer ID used for client_delay_access matching. | |
7233 | ||
7234 | speed_limit is bytes added to the bucket per second. | |
7235 | ||
7236 | max_bucket_size is the maximum size of a bucket, enforced after any | |
7237 | speed_limit additions. | |
7238 | ||
7239 | Please see the delay_parameters option for more information and | |
7240 | examples. | |
7241 | ||
7242 | Example: | |
7243 | client_delay_parameters 1 1024 2048 | |
7244 | client_delay_parameters 2 51200 16384 | |
7245 | ||
7246 | See also client_delay_access. | |
7247 | ||
7248 | DOC_END | |
7249 | ||
7250 | NAME: client_delay_access | |
7251 | TYPE: client_delay_pool_access | |
7252 | DEFAULT: none | |
7253 | DEFAULT_DOC: Deny use of the pool, unless allow rules exist in squid.conf for the pool. | |
7254 | IFDEF: USE_DELAY_POOLS | |
7255 | LOC: Config.ClientDelay | |
7256 | DOC_START | |
7257 | This option determines the client-side delay pool for the | |
7258 | request: | |
7259 | ||
7260 | client_delay_access pool_ID allow|deny acl_name | |
7261 | ||
7262 | All client_delay_access options are checked in their pool ID | |
7263 | order, starting with pool 1. The first checked pool with allowed | |
7264 | request is selected for the request. If no ACL matches or there | |
7265 | are no client_delay_access options, the request bandwidth is not | |
7266 | limited. | |
7267 | ||
7268 | The ACL-selected pool is then used to find the | |
7269 | client_delay_parameters for the request. Client-side pools are | |
7270 | not used to aggregate clients. Clients are always aggregated | |
7271 | based on their source IP addresses (one bucket per source IP). | |
7272 | ||
7273 | This clause only supports fast acl types. | |
7274 | See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details. | |
7275 | Additionally, only the client TCP connection details are available. | |
7276 | ACLs testing HTTP properties will not work. | |
7277 | ||
7278 | Please see delay_access for more examples. | |
7279 | ||
7280 | Example: | |
7281 | client_delay_access 1 allow low_rate_network | |
7282 | client_delay_access 2 allow vips_network | |
7283 | ||
7284 | ||
7285 | See also client_delay_parameters and client_delay_pools. | |
7286 | DOC_END | |
7287 | ||
7288 | NAME: response_delay_pool | |
7289 | TYPE: response_delay_pool_parameters | |
7290 | DEFAULT: none | |
7291 | IFDEF: USE_DELAY_POOLS | |
7292 | LOC: Config.MessageDelay | |
7293 | DOC_START | |
7294 | This option configures client response bandwidth limits using the | |
7295 | following format: | |
7296 | ||
7297 | response_delay_pool name [option=value] ... | |
7298 | ||
7299 | name the response delay pool name | |
7300 | ||
7301 | available options: | |
7302 | ||
7303 | individual-restore The speed limit of an individual | |
7304 | bucket(bytes/s). To be used in conjunction | |
7305 | with 'individual-maximum'. | |
7306 | ||
7307 | individual-maximum The maximum number of bytes which can | |
7308 | be placed into the individual bucket. To be used | |
7309 | in conjunction with 'individual-restore'. | |
7310 | ||
7311 | aggregate-restore The speed limit for the aggregate | |
7312 | bucket(bytes/s). To be used in conjunction with | |
7313 | 'aggregate-maximum'. | |
7314 | ||
7315 | aggregate-maximum The maximum number of bytes which can | |
7316 | be placed into the aggregate bucket. To be used | |
7317 | in conjunction with 'aggregate-restore'. | |
7318 | ||
7319 | initial-bucket-level The initial bucket size as a percentage | |
7320 | of individual-maximum. | |
7321 | ||
7322 | Individual and(or) aggregate bucket options may not be specified, | |
7323 | meaning no individual and(or) aggregate speed limitation. | |
7324 | See also response_delay_pool_access and delay_parameters for | |
7325 | terminology details. | |
7326 | DOC_END | |
7327 | ||
7328 | NAME: response_delay_pool_access | |
7329 | TYPE: response_delay_pool_access | |
7330 | DEFAULT: none | |
7331 | DEFAULT_DOC: Deny use of the pool, unless allow rules exist in squid.conf for the pool. | |
7332 | IFDEF: USE_DELAY_POOLS | |
7333 | LOC: Config.MessageDelay | |
7334 | DOC_START | |
7335 | Determines whether a specific named response delay pool is used | |
7336 | for the transaction. The syntax for this directive is: | |
7337 | ||
7338 | response_delay_pool_access pool_name allow|deny acl_name | |
7339 | ||
7340 | All response_delay_pool_access options are checked in the order | |
7341 | they appear in this configuration file. The first rule with a | |
7342 | matching ACL wins. If (and only if) an "allow" rule won, Squid | |
7343 | assigns the response to the corresponding named delay pool. | |
7344 | DOC_END | |
7345 | ||
7346 | COMMENT_START | |
7347 | WCCPv1 AND WCCPv2 CONFIGURATION OPTIONS | |
7348 | ----------------------------------------------------------------------------- | |
7349 | COMMENT_END | |
7350 | ||
7351 | NAME: wccp_router | |
7352 | TYPE: address | |
7353 | LOC: Config.Wccp.router | |
7354 | DEFAULT: any_addr | |
7355 | DEFAULT_DOC: WCCP disabled. | |
7356 | IFDEF: USE_WCCP | |
7357 | DOC_START | |
7358 | Use this option to define your WCCP ``home'' router for | |
7359 | Squid. | |
7360 | ||
7361 | wccp_router supports a single WCCP(v1) router | |
7362 | ||
7363 | wccp2_router supports multiple WCCPv2 routers | |
7364 | ||
7365 | only one of the two may be used at the same time and defines | |
7366 | which version of WCCP to use. | |
7367 | DOC_END | |
7368 | ||
7369 | NAME: wccp2_router | |
7370 | TYPE: IpAddress_list | |
7371 | LOC: Config.Wccp2.router | |
7372 | DEFAULT: none | |
7373 | DEFAULT_DOC: WCCPv2 disabled. | |
7374 | IFDEF: USE_WCCPv2 | |
7375 | DOC_START | |
7376 | Use this option to define your WCCP ``home'' router for | |
7377 | Squid. | |
7378 | ||
7379 | wccp_router supports a single WCCP(v1) router | |
7380 | ||
7381 | wccp2_router supports multiple WCCPv2 routers | |
7382 | ||
7383 | only one of the two may be used at the same time and defines | |
7384 | which version of WCCP to use. | |
7385 | DOC_END | |
7386 | ||
7387 | NAME: wccp_version | |
7388 | TYPE: int | |
7389 | LOC: Config.Wccp.version | |
7390 | DEFAULT: 4 | |
7391 | IFDEF: USE_WCCP | |
7392 | DOC_START | |
7393 | This directive is only relevant if you need to set up WCCP(v1) | |
7394 | to some very old and end-of-life Cisco routers. In all other | |
7395 | setups it must be left unset or at the default setting. | |
7396 | It defines an internal version in the WCCP(v1) protocol, | |
7397 | with version 4 being the officially documented protocol. | |
7398 | ||
7399 | According to some users, Cisco IOS 11.2 and earlier only | |
7400 | support WCCP version 3. If you're using that or an earlier | |
7401 | version of IOS, you may need to change this value to 3, otherwise | |
7402 | do not specify this parameter. | |
7403 | DOC_END | |
7404 | ||
7405 | NAME: wccp2_rebuild_wait | |
7406 | TYPE: onoff | |
7407 | LOC: Config.Wccp2.rebuildwait | |
7408 | DEFAULT: on | |
7409 | IFDEF: USE_WCCPv2 | |
7410 | DOC_START | |
7411 | If this is enabled Squid will wait for the cache dir rebuild to finish | |
7412 | before sending the first wccp2 HereIAm packet | |
7413 | DOC_END | |
7414 | ||
7415 | NAME: wccp2_forwarding_method | |
7416 | TYPE: wccp2_method | |
7417 | LOC: Config.Wccp2.forwarding_method | |
7418 | DEFAULT: gre | |
7419 | IFDEF: USE_WCCPv2 | |
7420 | DOC_START | |
7421 | WCCP2 allows the setting of forwarding methods between the | |
7422 | router/switch and the cache. Valid values are as follows: | |
7423 | ||
7424 | gre - GRE encapsulation (forward the packet in a GRE/WCCP tunnel) | |
7425 | l2 - L2 redirect (forward the packet using Layer 2/MAC rewriting) | |
7426 | ||
7427 | Currently (as of IOS 12.4) cisco routers only support GRE. | |
7428 | Cisco switches only support the L2 redirect assignment method. | |
7429 | DOC_END | |
7430 | ||
7431 | NAME: wccp2_return_method | |
7432 | TYPE: wccp2_method | |
7433 | LOC: Config.Wccp2.return_method | |
7434 | DEFAULT: gre | |
7435 | IFDEF: USE_WCCPv2 | |
7436 | DOC_START | |
7437 | WCCP2 allows the setting of return methods between the | |
7438 | router/switch and the cache for packets that the cache | |
7439 | decides not to handle. Valid values are as follows: | |
7440 | ||
7441 | gre - GRE encapsulation (forward the packet in a GRE/WCCP tunnel) | |
7442 | l2 - L2 redirect (forward the packet using Layer 2/MAC rewriting) | |
7443 | ||
7444 | Currently (as of IOS 12.4) cisco routers only support GRE. | |
7445 | Cisco switches only support the L2 redirect assignment. | |
7446 | ||
7447 | If the "ip wccp redirect exclude in" command has been | |
7448 | enabled on the cache interface, then it is still safe for | |
7449 | the proxy server to use a l2 redirect method even if this | |
7450 | option is set to GRE. | |
7451 | DOC_END | |
7452 | ||
7453 | NAME: wccp2_assignment_method | |
7454 | TYPE: wccp2_amethod | |
7455 | LOC: Config.Wccp2.assignment_method | |
7456 | DEFAULT: hash | |
7457 | IFDEF: USE_WCCPv2 | |
7458 | DOC_START | |
7459 | WCCP2 allows the setting of methods to assign the WCCP hash | |
7460 | Valid values are as follows: | |
7461 | ||
7462 | hash - Hash assignment | |
7463 | mask - Mask assignment | |
7464 | ||
7465 | As a general rule, cisco routers support the hash assignment method | |
7466 | and cisco switches support the mask assignment method. | |
7467 | DOC_END | |
7468 | ||
7469 | NAME: wccp2_service | |
7470 | TYPE: wccp2_service | |
7471 | LOC: Config.Wccp2.info | |
7472 | DEFAULT_IF_NONE: standard 0 | |
7473 | DEFAULT_DOC: Use the 'web-cache' standard service. | |
7474 | IFDEF: USE_WCCPv2 | |
7475 | DOC_START | |
7476 | WCCP2 allows for multiple traffic services. There are two | |
7477 | types: "standard" and "dynamic". The standard type defines | |
7478 | one service id - http (id 0). The dynamic service ids can be from | |
7479 | 51 to 255 inclusive. In order to use a dynamic service id | |
7480 | one must define the type of traffic to be redirected; this is done | |
7481 | using the wccp2_service_info option. | |
7482 | ||
7483 | The "standard" type does not require a wccp2_service_info option, | |
7484 | just specifying the service id will suffice. | |
7485 | ||
7486 | MD5 service authentication can be enabled by adding | |
7487 | "password=<password>" to the end of this service declaration. | |
7488 | ||
7489 | Examples: | |
7490 | ||
7491 | wccp2_service standard 0 # for the 'web-cache' standard service | |
7492 | wccp2_service dynamic 80 # a dynamic service type which will be | |
7493 | # fleshed out with subsequent options. | |
7494 | wccp2_service standard 0 password=foo | |
7495 | DOC_END | |
7496 | ||
7497 | NAME: wccp2_service_info | |
7498 | TYPE: wccp2_service_info | |
7499 | LOC: Config.Wccp2.info | |
7500 | DEFAULT: none | |
7501 | IFDEF: USE_WCCPv2 | |
7502 | DOC_START | |
7503 | Dynamic WCCPv2 services require further information to define the | |
7504 | traffic you wish to have diverted. | |
7505 | ||
7506 | The format is: | |
7507 | ||
7508 | wccp2_service_info <id> protocol=<protocol> flags=<flag>,<flag>.. | |
7509 | priority=<priority> ports=<port>,<port>.. | |
7510 | ||
7511 | The relevant WCCPv2 flags: | |
7512 | + src_ip_hash, dst_ip_hash | |
7513 | + source_port_hash, dst_port_hash | |
7514 | + src_ip_alt_hash, dst_ip_alt_hash | |
7515 | + src_port_alt_hash, dst_port_alt_hash | |
7516 | + ports_source | |
7517 | ||
7518 | The port list can be one to eight entries. | |
7519 | ||
7520 | Example: | |
7521 | ||
7522 | wccp2_service_info 80 protocol=tcp flags=src_ip_hash,ports_source | |
7523 | priority=240 ports=80 | |
7524 | ||
7525 | Note: the service id must have been defined by a previous | |
7526 | 'wccp2_service dynamic <id>' entry. | |
7527 | DOC_END | |
7528 | ||
7529 | NAME: wccp2_weight | |
7530 | TYPE: int | |
7531 | LOC: Config.Wccp2.weight | |
7532 | DEFAULT: 10000 | |
7533 | IFDEF: USE_WCCPv2 | |
7534 | DOC_START | |
7535 | Each cache server gets assigned a set of the destination | |
7536 | hash proportional to their weight. | |
7537 | DOC_END | |
7538 | ||
7539 | NAME: wccp_address | |
7540 | TYPE: address | |
7541 | LOC: Config.Wccp.address | |
7542 | DEFAULT: 0.0.0.0 | |
7543 | DEFAULT_DOC: Address selected by the operating system. | |
7544 | IFDEF: USE_WCCP | |
7545 | DOC_START | |
7546 | Use this option if you require WCCPv2 to use a specific | |
7547 | interface address. | |
7548 | ||
7549 | The default behavior is to not bind to any specific address. | |
7550 | DOC_END | |
7551 | ||
7552 | NAME: wccp2_address | |
7553 | TYPE: address | |
7554 | LOC: Config.Wccp2.address | |
7555 | DEFAULT: 0.0.0.0 | |
7556 | DEFAULT_DOC: Address selected by the operating system. | |
7557 | IFDEF: USE_WCCPv2 | |
7558 | DOC_START | |
7559 | Use this option if you require WCCP to use a specific | |
7560 | interface address. | |
7561 | ||
7562 | The default behavior is to not bind to any specific address. | |
7563 | DOC_END | |
7564 | ||
7565 | COMMENT_START | |
7566 | PERSISTENT CONNECTION HANDLING | |
7567 | ----------------------------------------------------------------------------- | |
7568 | ||
7569 | Also see "pconn_timeout" in the TIMEOUTS section | |
7570 | COMMENT_END | |
7571 | ||
7572 | NAME: client_persistent_connections | |
7573 | TYPE: onoff | |
7574 | LOC: Config.onoff.client_pconns | |
7575 | DEFAULT: on | |
7576 | DOC_START | |
7577 | Persistent connection support for clients. | |
7578 | Squid uses persistent connections (when allowed). You can use | |
7579 | this option to disable persistent connections with clients. | |
7580 | DOC_END | |
7581 | ||
7582 | NAME: server_persistent_connections | |
7583 | TYPE: onoff | |
7584 | LOC: Config.onoff.server_pconns | |
7585 | DEFAULT: on | |
7586 | DOC_START | |
7587 | Persistent connection support for servers. | |
7588 | Squid uses persistent connections (when allowed). You can use | |
7589 | this option to disable persistent connections with servers. | |
7590 | DOC_END | |
7591 | ||
7592 | NAME: persistent_connection_after_error | |
7593 | TYPE: onoff | |
7594 | LOC: Config.onoff.error_pconns | |
7595 | DEFAULT: on | |
7596 | DOC_START | |
7597 | With this directive the use of persistent connections after | |
7598 | HTTP errors can be disabled. Useful if you have clients | |
7599 | who fail to handle errors on persistent connections proper. | |
7600 | DOC_END | |
7601 | ||
7602 | NAME: detect_broken_pconn | |
7603 | TYPE: onoff | |
7604 | LOC: Config.onoff.detect_broken_server_pconns | |
7605 | DEFAULT: off | |
7606 | DOC_START | |
7607 | Some servers have been found to incorrectly signal the use | |
7608 | of HTTP/1.0 persistent connections even on replies not | |
7609 | compatible, causing significant delays. This server problem | |
7610 | has mostly been seen on redirects. | |
7611 | ||
7612 | By enabling this directive Squid attempts to detect such | |
7613 | broken replies and automatically assume the reply is finished | |
7614 | after 10 seconds timeout. | |
7615 | DOC_END | |
7616 | ||
7617 | COMMENT_START | |
7618 | CACHE DIGEST OPTIONS | |
7619 | ----------------------------------------------------------------------------- | |
7620 | COMMENT_END | |
7621 | ||
7622 | NAME: digest_generation | |
7623 | IFDEF: USE_CACHE_DIGESTS | |
7624 | TYPE: onoff | |
7625 | LOC: Config.onoff.digest_generation | |
7626 | DEFAULT: on | |
7627 | DOC_START | |
7628 | This controls whether the server will generate a Cache Digest | |
7629 | of its contents. By default, Cache Digest generation is | |
7630 | enabled if Squid is compiled with --enable-cache-digests defined. | |
7631 | DOC_END | |
7632 | ||
7633 | NAME: digest_bits_per_entry | |
7634 | IFDEF: USE_CACHE_DIGESTS | |
7635 | TYPE: int | |
7636 | LOC: Config.digest.bits_per_entry | |
7637 | DEFAULT: 5 | |
7638 | DOC_START | |
7639 | This is the number of bits of the server's Cache Digest which | |
7640 | will be associated with the Digest entry for a given HTTP | |
7641 | Method and URL (public key) combination. The default is 5. | |
7642 | DOC_END | |
7643 | ||
7644 | NAME: digest_rebuild_period | |
7645 | IFDEF: USE_CACHE_DIGESTS | |
7646 | COMMENT: (seconds) | |
7647 | TYPE: time_t | |
7648 | LOC: Config.digest.rebuild_period | |
7649 | DEFAULT: 1 hour | |
7650 | DOC_START | |
7651 | This is the wait time between Cache Digest rebuilds. | |
7652 | DOC_END | |
7653 | ||
7654 | NAME: digest_rewrite_period | |
7655 | COMMENT: (seconds) | |
7656 | IFDEF: USE_CACHE_DIGESTS | |
7657 | TYPE: time_t | |
7658 | LOC: Config.digest.rewrite_period | |
7659 | DEFAULT: 1 hour | |
7660 | DOC_START | |
7661 | This is the wait time between Cache Digest writes to | |
7662 | disk. | |
7663 | DOC_END | |
7664 | ||
7665 | NAME: digest_swapout_chunk_size | |
7666 | COMMENT: (bytes) | |
7667 | TYPE: b_size_t | |
7668 | IFDEF: USE_CACHE_DIGESTS | |
7669 | LOC: Config.digest.swapout_chunk_size | |
7670 | DEFAULT: 4096 bytes | |
7671 | DOC_START | |
7672 | This is the number of bytes of the Cache Digest to write to | |
7673 | disk at a time. It defaults to 4096 bytes (4KB), the Squid | |
7674 | default swap page. | |
7675 | DOC_END | |
7676 | ||
7677 | NAME: digest_rebuild_chunk_percentage | |
7678 | COMMENT: (percent, 0-100) | |
7679 | IFDEF: USE_CACHE_DIGESTS | |
7680 | TYPE: int | |
7681 | LOC: Config.digest.rebuild_chunk_percentage | |
7682 | DEFAULT: 10 | |
7683 | DOC_START | |
7684 | This is the percentage of the Cache Digest to be scanned at a | |
7685 | time. By default it is set to 10% of the Cache Digest. | |
7686 | DOC_END | |
7687 | ||
7688 | COMMENT_START | |
7689 | SNMP OPTIONS | |
7690 | ----------------------------------------------------------------------------- | |
7691 | COMMENT_END | |
7692 | ||
7693 | NAME: snmp_port | |
7694 | TYPE: u_short | |
7695 | LOC: Config.Port.snmp | |
7696 | DEFAULT: 0 | |
7697 | DEFAULT_DOC: SNMP disabled. | |
7698 | IFDEF: SQUID_SNMP | |
7699 | DOC_START | |
7700 | The port number where Squid listens for SNMP requests. To enable | |
7701 | SNMP support set this to a suitable port number. Port number | |
7702 | 3401 is often used for the Squid SNMP agent. By default it's | |
7703 | set to "0" (disabled) | |
7704 | ||
7705 | Example: | |
7706 | snmp_port 3401 | |
7707 | DOC_END | |
7708 | ||
7709 | NAME: snmp_access | |
7710 | TYPE: acl_access | |
7711 | LOC: Config.accessList.snmp | |
7712 | DEFAULT: none | |
7713 | DEFAULT_DOC: Deny, unless rules exist in squid.conf. | |
7714 | IFDEF: SQUID_SNMP | |
7715 | DOC_START | |
7716 | Allowing or denying access to the SNMP port. | |
7717 | ||
7718 | All access to the agent is denied by default. | |
7719 | usage: | |
7720 | ||
7721 | snmp_access allow|deny [!]aclname ... | |
7722 | ||
7723 | This clause only supports fast acl types. | |
7724 | See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details. | |
7725 | ||
7726 | Example: | |
7727 | snmp_access allow snmppublic localhost | |
7728 | snmp_access deny all | |
7729 | DOC_END | |
7730 | ||
7731 | NAME: snmp_incoming_address | |
7732 | TYPE: address | |
7733 | LOC: Config.Addrs.snmp_incoming | |
7734 | DEFAULT: any_addr | |
7735 | DEFAULT_DOC: Accept SNMP packets from all machine interfaces. | |
7736 | IFDEF: SQUID_SNMP | |
7737 | DOC_START | |
7738 | Just like 'udp_incoming_address', but for the SNMP port. | |
7739 | ||
7740 | snmp_incoming_address is used for the SNMP socket receiving | |
7741 | messages from SNMP agents. | |
7742 | ||
7743 | The default snmp_incoming_address is to listen on all | |
7744 | available network interfaces. | |
7745 | DOC_END | |
7746 | ||
7747 | NAME: snmp_outgoing_address | |
7748 | TYPE: address | |
7749 | LOC: Config.Addrs.snmp_outgoing | |
7750 | DEFAULT: no_addr | |
7751 | DEFAULT_DOC: Use snmp_incoming_address or an address selected by the operating system. | |
7752 | IFDEF: SQUID_SNMP | |
7753 | DOC_START | |
7754 | Just like 'udp_outgoing_address', but for the SNMP port. | |
7755 | ||
7756 | snmp_outgoing_address is used for SNMP packets returned to SNMP | |
7757 | agents. | |
7758 | ||
7759 | If snmp_outgoing_address is not set it will use the same socket | |
7760 | as snmp_incoming_address. Only change this if you want to have | |
7761 | SNMP replies sent using another address than where this Squid | |
7762 | listens for SNMP queries. | |
7763 | ||
7764 | NOTE, snmp_incoming_address and snmp_outgoing_address can not have | |
7765 | the same value since they both use the same port. | |
7766 | DOC_END | |
7767 | ||
7768 | COMMENT_START | |
7769 | ICP OPTIONS | |
7770 | ----------------------------------------------------------------------------- | |
7771 | COMMENT_END | |
7772 | ||
7773 | NAME: icp_port udp_port | |
7774 | TYPE: u_short | |
7775 | DEFAULT: 0 | |
7776 | DEFAULT_DOC: ICP disabled. | |
7777 | LOC: Config.Port.icp | |
7778 | DOC_START | |
7779 | The port number where Squid sends and receives ICP queries to | |
7780 | and from neighbor caches. The standard UDP port for ICP is 3130. | |
7781 | ||
7782 | Example: | |
7783 | icp_port @DEFAULT_ICP_PORT@ | |
7784 | DOC_END | |
7785 | ||
7786 | NAME: htcp_port | |
7787 | IFDEF: USE_HTCP | |
7788 | TYPE: u_short | |
7789 | DEFAULT: 0 | |
7790 | DEFAULT_DOC: HTCP disabled. | |
7791 | LOC: Config.Port.htcp | |
7792 | DOC_START | |
7793 | The port number where Squid sends and receives HTCP queries to | |
7794 | and from neighbor caches. To turn it on you want to set it to | |
7795 | 4827. | |
7796 | ||
7797 | Example: | |
7798 | htcp_port 4827 | |
7799 | DOC_END | |
7800 | ||
7801 | NAME: log_icp_queries | |
7802 | COMMENT: on|off | |
7803 | TYPE: onoff | |
7804 | DEFAULT: on | |
7805 | LOC: Config.onoff.log_udp | |
7806 | DOC_START | |
7807 | If set, ICP queries are logged to access.log. You may wish | |
7808 | do disable this if your ICP load is VERY high to speed things | |
7809 | up or to simplify log analysis. | |
7810 | DOC_END | |
7811 | ||
7812 | NAME: udp_incoming_address | |
7813 | TYPE: address | |
7814 | LOC:Config.Addrs.udp_incoming | |
7815 | DEFAULT: any_addr | |
7816 | DEFAULT_DOC: Accept packets from all machine interfaces. | |
7817 | DOC_START | |
7818 | udp_incoming_address is used for UDP packets received from other | |
7819 | caches. | |
7820 | ||
7821 | The default behavior is to not bind to any specific address. | |
7822 | ||
7823 | Only change this if you want to have all UDP queries received on | |
7824 | a specific interface/address. | |
7825 | ||
7826 | NOTE: udp_incoming_address is used by the ICP, HTCP, and DNS | |
7827 | modules. Altering it will affect all of them in the same manner. | |
7828 | ||
7829 | see also; udp_outgoing_address | |
7830 | ||
7831 | NOTE, udp_incoming_address and udp_outgoing_address can not | |
7832 | have the same value since they both use the same port. | |
7833 | DOC_END | |
7834 | ||
7835 | NAME: udp_outgoing_address | |
7836 | TYPE: address | |
7837 | LOC: Config.Addrs.udp_outgoing | |
7838 | DEFAULT: no_addr | |
7839 | DEFAULT_DOC: Use udp_incoming_address or an address selected by the operating system. | |
7840 | DOC_START | |
7841 | udp_outgoing_address is used for UDP packets sent out to other | |
7842 | caches. | |
7843 | ||
7844 | The default behavior is to not bind to any specific address. | |
7845 | ||
7846 | Instead it will use the same socket as udp_incoming_address. | |
7847 | Only change this if you want to have UDP queries sent using another | |
7848 | address than where this Squid listens for UDP queries from other | |
7849 | caches. | |
7850 | ||
7851 | NOTE: udp_outgoing_address is used by the ICP, HTCP, and DNS | |
7852 | modules. Altering it will affect all of them in the same manner. | |
7853 | ||
7854 | see also; udp_incoming_address | |
7855 | ||
7856 | NOTE, udp_incoming_address and udp_outgoing_address can not | |
7857 | have the same value since they both use the same port. | |
7858 | DOC_END | |
7859 | ||
7860 | NAME: icp_hit_stale | |
7861 | COMMENT: on|off | |
7862 | TYPE: onoff | |
7863 | DEFAULT: off | |
7864 | LOC: Config.onoff.icp_hit_stale | |
7865 | DOC_START | |
7866 | If you want to return ICP_HIT for stale cache objects, set this | |
7867 | option to 'on'. If you have sibling relationships with caches | |
7868 | in other administrative domains, this should be 'off'. If you only | |
7869 | have sibling relationships with caches under your control, | |
7870 | it is probably okay to set this to 'on'. | |
7871 | If set to 'on', your siblings should use the option "allow-miss" | |
7872 | on their cache_peer lines for connecting to you. | |
7873 | DOC_END | |
7874 | ||
7875 | NAME: minimum_direct_hops | |
7876 | TYPE: int | |
7877 | DEFAULT: 4 | |
7878 | LOC: Config.minDirectHops | |
7879 | DOC_START | |
7880 | If using the ICMP pinging stuff, do direct fetches for sites | |
7881 | which are no more than this many hops away. | |
7882 | DOC_END | |
7883 | ||
7884 | NAME: minimum_direct_rtt | |
7885 | COMMENT: (msec) | |
7886 | TYPE: int | |
7887 | DEFAULT: 400 | |
7888 | LOC: Config.minDirectRtt | |
7889 | DOC_START | |
7890 | If using the ICMP pinging stuff, do direct fetches for sites | |
7891 | which are no more than this many rtt milliseconds away. | |
7892 | DOC_END | |
7893 | ||
7894 | NAME: netdb_low | |
7895 | TYPE: int | |
7896 | DEFAULT: 900 | |
7897 | LOC: Config.Netdb.low | |
7898 | DOC_START | |
7899 | The low water mark for the ICMP measurement database. | |
7900 | ||
7901 | Note: high watermark controlled by netdb_high directive. | |
7902 | ||
7903 | These watermarks are counts, not percents. The defaults are | |
7904 | (low) 900 and (high) 1000. When the high water mark is | |
7905 | reached, database entries will be deleted until the low | |
7906 | mark is reached. | |
7907 | DOC_END | |
7908 | ||
7909 | NAME: netdb_high | |
7910 | TYPE: int | |
7911 | DEFAULT: 1000 | |
7912 | LOC: Config.Netdb.high | |
7913 | DOC_START | |
7914 | The high water mark for the ICMP measurement database. | |
7915 | ||
7916 | Note: low watermark controlled by netdb_low directive. | |
7917 | ||
7918 | These watermarks are counts, not percents. The defaults are | |
7919 | (low) 900 and (high) 1000. When the high water mark is | |
7920 | reached, database entries will be deleted until the low | |
7921 | mark is reached. | |
7922 | DOC_END | |
7923 | ||
7924 | NAME: netdb_ping_period | |
7925 | TYPE: time_t | |
7926 | LOC: Config.Netdb.period | |
7927 | DEFAULT: 5 minutes | |
7928 | DOC_START | |
7929 | The minimum period for measuring a site. There will be at | |
7930 | least this much delay between successive pings to the same | |
7931 | network. The default is five minutes. | |
7932 | DOC_END | |
7933 | ||
7934 | NAME: query_icmp | |
7935 | COMMENT: on|off | |
7936 | TYPE: onoff | |
7937 | DEFAULT: off | |
7938 | LOC: Config.onoff.query_icmp | |
7939 | DOC_START | |
7940 | If you want to ask your peers to include ICMP data in their ICP | |
7941 | replies, enable this option. | |
7942 | ||
7943 | If your peer has configured Squid (during compilation) with | |
7944 | '--enable-icmp' that peer will send ICMP pings to origin server | |
7945 | sites of the URLs it receives. If you enable this option the | |
7946 | ICP replies from that peer will include the ICMP data (if available). | |
7947 | Then, when choosing a parent cache, Squid will choose the parent with | |
7948 | the minimal RTT to the origin server. When this happens, the | |
7949 | hierarchy field of the access.log will be | |
7950 | "CLOSEST_PARENT_MISS". This option is off by default. | |
7951 | DOC_END | |
7952 | ||
7953 | NAME: test_reachability | |
7954 | COMMENT: on|off | |
7955 | TYPE: onoff | |
7956 | DEFAULT: off | |
7957 | LOC: Config.onoff.test_reachability | |
7958 | DOC_START | |
7959 | When this is 'on', ICP MISS replies will be ICP_MISS_NOFETCH | |
7960 | instead of ICP_MISS if the target host is NOT in the ICMP | |
7961 | database, or has a zero RTT. | |
7962 | DOC_END | |
7963 | ||
7964 | NAME: icp_query_timeout | |
7965 | COMMENT: (msec) | |
7966 | DEFAULT: 0 | |
7967 | DEFAULT_DOC: Dynamic detection. | |
7968 | TYPE: int | |
7969 | LOC: Config.Timeout.icp_query | |
7970 | DOC_START | |
7971 | Normally Squid will automatically determine an optimal ICP | |
7972 | query timeout value based on the round-trip-time of recent ICP | |
7973 | queries. If you want to override the value determined by | |
7974 | Squid, set this 'icp_query_timeout' to a non-zero value. This | |
7975 | value is specified in MILLISECONDS, so, to use a 2-second | |
7976 | timeout (the old default), you would write: | |
7977 | ||
7978 | icp_query_timeout 2000 | |
7979 | DOC_END | |
7980 | ||
7981 | NAME: maximum_icp_query_timeout | |
7982 | COMMENT: (msec) | |
7983 | DEFAULT: 2000 | |
7984 | TYPE: int | |
7985 | LOC: Config.Timeout.icp_query_max | |
7986 | DOC_START | |
7987 | Normally the ICP query timeout is determined dynamically. But | |
7988 | sometimes it can lead to very large values (say 5 seconds). | |
7989 | Use this option to put an upper limit on the dynamic timeout | |
7990 | value. Do NOT use this option to always use a fixed (instead | |
7991 | of a dynamic) timeout value. To set a fixed timeout see the | |
7992 | 'icp_query_timeout' directive. | |
7993 | DOC_END | |
7994 | ||
7995 | NAME: minimum_icp_query_timeout | |
7996 | COMMENT: (msec) | |
7997 | DEFAULT: 5 | |
7998 | TYPE: int | |
7999 | LOC: Config.Timeout.icp_query_min | |
8000 | DOC_START | |
8001 | Normally the ICP query timeout is determined dynamically. But | |
8002 | sometimes it can lead to very small timeouts, even lower than | |
8003 | the normal latency variance on your link due to traffic. | |
8004 | Use this option to put an lower limit on the dynamic timeout | |
8005 | value. Do NOT use this option to always use a fixed (instead | |
8006 | of a dynamic) timeout value. To set a fixed timeout see the | |
8007 | 'icp_query_timeout' directive. | |
8008 | DOC_END | |
8009 | ||
8010 | NAME: background_ping_rate | |
8011 | COMMENT: time-units | |
8012 | TYPE: time_t | |
8013 | DEFAULT: 10 seconds | |
8014 | LOC: Config.backgroundPingRate | |
8015 | DOC_START | |
8016 | Controls how often the ICP pings are sent to siblings that | |
8017 | have background-ping set. | |
8018 | DOC_END | |
8019 | ||
8020 | COMMENT_START | |
8021 | MULTICAST ICP OPTIONS | |
8022 | ----------------------------------------------------------------------------- | |
8023 | COMMENT_END | |
8024 | ||
8025 | NAME: mcast_groups | |
8026 | TYPE: wordlist | |
8027 | LOC: Config.mcast_group_list | |
8028 | DEFAULT: none | |
8029 | DOC_START | |
8030 | This tag specifies a list of multicast groups which your server | |
8031 | should join to receive multicasted ICP queries. | |
8032 | ||
8033 | NOTE! Be very careful what you put here! Be sure you | |
8034 | understand the difference between an ICP _query_ and an ICP | |
8035 | _reply_. This option is to be set only if you want to RECEIVE | |
8036 | multicast queries. Do NOT set this option to SEND multicast | |
8037 | ICP (use cache_peer for that). ICP replies are always sent via | |
8038 | unicast, so this option does not affect whether or not you will | |
8039 | receive replies from multicast group members. | |
8040 | ||
8041 | You must be very careful to NOT use a multicast address which | |
8042 | is already in use by another group of caches. | |
8043 | ||
8044 | If you are unsure about multicast, please read the Multicast | |
8045 | chapter in the Squid FAQ (http://www.squid-cache.org/FAQ/). | |
8046 | ||
8047 | Usage: mcast_groups 239.128.16.128 224.0.1.20 | |
8048 | ||
8049 | By default, Squid doesn't listen on any multicast groups. | |
8050 | DOC_END | |
8051 | ||
8052 | NAME: mcast_miss_addr | |
8053 | IFDEF: MULTICAST_MISS_STREAM | |
8054 | TYPE: address | |
8055 | LOC: Config.mcast_miss.addr | |
8056 | DEFAULT: no_addr | |
8057 | DEFAULT_DOC: disabled. | |
8058 | DOC_START | |
8059 | If you enable this option, every "cache miss" URL will | |
8060 | be sent out on the specified multicast address. | |
8061 | ||
8062 | Do not enable this option unless you are are absolutely | |
8063 | certain you understand what you are doing. | |
8064 | DOC_END | |
8065 | ||
8066 | NAME: mcast_miss_ttl | |
8067 | IFDEF: MULTICAST_MISS_STREAM | |
8068 | TYPE: u_short | |
8069 | LOC: Config.mcast_miss.ttl | |
8070 | DEFAULT: 16 | |
8071 | DOC_START | |
8072 | This is the time-to-live value for packets multicasted | |
8073 | when multicasting off cache miss URLs is enabled. By | |
8074 | default this is set to 'site scope', i.e. 16. | |
8075 | DOC_END | |
8076 | ||
8077 | NAME: mcast_miss_port | |
8078 | IFDEF: MULTICAST_MISS_STREAM | |
8079 | TYPE: u_short | |
8080 | LOC: Config.mcast_miss.port | |
8081 | DEFAULT: 3135 | |
8082 | DOC_START | |
8083 | This is the port number to be used in conjunction with | |
8084 | 'mcast_miss_addr'. | |
8085 | DOC_END | |
8086 | ||
8087 | NAME: mcast_miss_encode_key | |
8088 | IFDEF: MULTICAST_MISS_STREAM | |
8089 | TYPE: string | |
8090 | LOC: Config.mcast_miss.encode_key | |
8091 | DEFAULT: XXXXXXXXXXXXXXXX | |
8092 | DOC_START | |
8093 | The URLs that are sent in the multicast miss stream are | |
8094 | encrypted. This is the encryption key. | |
8095 | DOC_END | |
8096 | ||
8097 | NAME: mcast_icp_query_timeout | |
8098 | COMMENT: (msec) | |
8099 | DEFAULT: 2000 | |
8100 | TYPE: int | |
8101 | LOC: Config.Timeout.mcast_icp_query | |
8102 | DOC_START | |
8103 | For multicast peers, Squid regularly sends out ICP "probes" to | |
8104 | count how many other peers are listening on the given multicast | |
8105 | address. This value specifies how long Squid should wait to | |
8106 | count all the replies. The default is 2000 msec, or 2 | |
8107 | seconds. | |
8108 | DOC_END | |
8109 | ||
8110 | COMMENT_START | |
8111 | INTERNAL ICON OPTIONS | |
8112 | ----------------------------------------------------------------------------- | |
8113 | COMMENT_END | |
8114 | ||
8115 | NAME: icon_directory | |
8116 | TYPE: string | |
8117 | LOC: Config.icons.directory | |
8118 | DEFAULT: @DEFAULT_ICON_DIR@ | |
8119 | DOC_START | |
8120 | Where the icons are stored. These are normally kept in | |
8121 | @DEFAULT_ICON_DIR@ | |
8122 | DOC_END | |
8123 | ||
8124 | NAME: global_internal_static | |
8125 | TYPE: onoff | |
8126 | LOC: Config.onoff.global_internal_static | |
8127 | DEFAULT: on | |
8128 | DOC_START | |
8129 | This directive controls is Squid should intercept all requests for | |
8130 | /squid-internal-static/ no matter which host the URL is requesting | |
8131 | (default on setting), or if nothing special should be done for | |
8132 | such URLs (off setting). The purpose of this directive is to make | |
8133 | icons etc work better in complex cache hierarchies where it may | |
8134 | not always be possible for all corners in the cache mesh to reach | |
8135 | the server generating a directory listing. | |
8136 | DOC_END | |
8137 | ||
8138 | NAME: short_icon_urls | |
8139 | TYPE: onoff | |
8140 | LOC: Config.icons.use_short_names | |
8141 | DEFAULT: on | |
8142 | DOC_START | |
8143 | If this is enabled Squid will use short URLs for icons. | |
8144 | If disabled it will revert to the old behavior of including | |
8145 | it's own name and port in the URL. | |
8146 | ||
8147 | If you run a complex cache hierarchy with a mix of Squid and | |
8148 | other proxies you may need to disable this directive. | |
8149 | DOC_END | |
8150 | ||
8151 | COMMENT_START | |
8152 | ERROR PAGE OPTIONS | |
8153 | ----------------------------------------------------------------------------- | |
8154 | COMMENT_END | |
8155 | ||
8156 | NAME: error_directory | |
8157 | TYPE: string | |
8158 | LOC: Config.errorDirectory | |
8159 | DEFAULT: none | |
8160 | DEFAULT_DOC: Send error pages in the clients preferred language | |
8161 | DOC_START | |
8162 | If you wish to create your own versions of the default | |
8163 | error files to customize them to suit your company copy | |
8164 | the error/template files to another directory and point | |
8165 | this tag at them. | |
8166 | ||
8167 | WARNING: This option will disable multi-language support | |
8168 | on error pages if used. | |
8169 | ||
8170 | The squid developers are interested in making squid available in | |
8171 | a wide variety of languages. If you are making translations for a | |
8172 | language that Squid does not currently provide please consider | |
8173 | contributing your translation back to the project. | |
8174 | http://wiki.squid-cache.org/Translations | |
8175 | ||
8176 | The squid developers working on translations are happy to supply drop-in | |
8177 | translated error files in exchange for any new language contributions. | |
8178 | DOC_END | |
8179 | ||
8180 | NAME: error_default_language | |
8181 | IFDEF: USE_ERR_LOCALES | |
8182 | TYPE: string | |
8183 | LOC: Config.errorDefaultLanguage | |
8184 | DEFAULT: none | |
8185 | DEFAULT_DOC: Generate English language pages. | |
8186 | DOC_START | |
8187 | Set the default language which squid will send error pages in | |
8188 | if no existing translation matches the clients language | |
8189 | preferences. | |
8190 | ||
8191 | If unset (default) generic English will be used. | |
8192 | ||
8193 | The squid developers are interested in making squid available in | |
8194 | a wide variety of languages. If you are interested in making | |
8195 | translations for any language see the squid wiki for details. | |
8196 | http://wiki.squid-cache.org/Translations | |
8197 | DOC_END | |
8198 | ||
8199 | NAME: error_log_languages | |
8200 | IFDEF: USE_ERR_LOCALES | |
8201 | TYPE: onoff | |
8202 | LOC: Config.errorLogMissingLanguages | |
8203 | DEFAULT: on | |
8204 | DOC_START | |
8205 | Log to cache.log what languages users are attempting to | |
8206 | auto-negotiate for translations. | |
8207 | ||
8208 | Successful negotiations are not logged. Only failures | |
8209 | have meaning to indicate that Squid may need an upgrade | |
8210 | of its error page translations. | |
8211 | DOC_END | |
8212 | ||
8213 | NAME: err_page_stylesheet | |
8214 | TYPE: string | |
8215 | LOC: Config.errorStylesheet | |
8216 | DEFAULT: @DEFAULT_CONFIG_DIR@/errorpage.css | |
8217 | DOC_START | |
8218 | CSS Stylesheet to pattern the display of Squid default error pages. | |
8219 | ||
8220 | For information on CSS see http://www.w3.org/Style/CSS/ | |
8221 | DOC_END | |
8222 | ||
8223 | NAME: err_html_text | |
8224 | TYPE: eol | |
8225 | LOC: Config.errHtmlText | |
8226 | DEFAULT: none | |
8227 | DOC_START | |
8228 | HTML text to include in error messages. Make this a "mailto" | |
8229 | URL to your admin address, or maybe just a link to your | |
8230 | organizations Web page. | |
8231 | ||
8232 | To include this in your error messages, you must rewrite | |
8233 | the error template files (found in the "errors" directory). | |
8234 | Wherever you want the 'err_html_text' line to appear, | |
8235 | insert a %L tag in the error template file. | |
8236 | DOC_END | |
8237 | ||
8238 | NAME: email_err_data | |
8239 | COMMENT: on|off | |
8240 | TYPE: onoff | |
8241 | LOC: Config.onoff.emailErrData | |
8242 | DEFAULT: on | |
8243 | DOC_START | |
8244 | If enabled, information about the occurred error will be | |
8245 | included in the mailto links of the ERR pages (if %W is set) | |
8246 | so that the email body contains the data. | |
8247 | Syntax is <A HREF="mailto:%w%W">%w</A> | |
8248 | DOC_END | |
8249 | ||
8250 | NAME: deny_info | |
8251 | TYPE: denyinfo | |
8252 | LOC: Config.denyInfoList | |
8253 | DEFAULT: none | |
8254 | DOC_START | |
8255 | Usage: deny_info err_page_name acl | |
8256 | or deny_info http://... acl | |
8257 | or deny_info TCP_RESET acl | |
8258 | ||
8259 | This can be used to return a ERR_ page for requests which | |
8260 | do not pass the 'http_access' rules. Squid remembers the last | |
8261 | acl it evaluated in http_access, and if a 'deny_info' line exists | |
8262 | for that ACL Squid returns a corresponding error page. | |
8263 | ||
8264 | The acl is typically the last acl on the http_access deny line which | |
8265 | denied access. The exceptions to this rule are: | |
8266 | - When Squid needs to request authentication credentials. It's then | |
8267 | the first authentication related acl encountered | |
8268 | - When none of the http_access lines matches. It's then the last | |
8269 | acl processed on the last http_access line. | |
8270 | - When the decision to deny access was made by an adaptation service, | |
8271 | the acl name is the corresponding eCAP or ICAP service_name. | |
8272 | ||
8273 | NP: If providing your own custom error pages with error_directory | |
8274 | you may also specify them by your custom file name: | |
8275 | Example: deny_info ERR_CUSTOM_ACCESS_DENIED bad_guys | |
8276 | ||
8277 | By defaut Squid will send "403 Forbidden". A different 4xx or 5xx | |
8278 | may be specified by prefixing the file name with the code and a colon. | |
8279 | e.g. 404:ERR_CUSTOM_ACCESS_DENIED | |
8280 | ||
8281 | Alternatively you can tell Squid to reset the TCP connection | |
8282 | by specifying TCP_RESET. | |
8283 | ||
8284 | Or you can specify an error URL or URL pattern. The browsers will | |
8285 | get redirected to the specified URL after formatting tags have | |
8286 | been replaced. Redirect will be done with 302 or 307 according to | |
8287 | HTTP/1.1 specs. A different 3xx code may be specified by prefixing | |
8288 | the URL. e.g. 303:http://example.com/ | |
8289 | ||
8290 | URL FORMAT TAGS: | |
8291 | %a - username (if available. Password NOT included) | |
8292 | %B - FTP path URL | |
8293 | %e - Error number | |
8294 | %E - Error description | |
8295 | %h - Squid hostname | |
8296 | %H - Request domain name | |
8297 | %i - Client IP Address | |
8298 | %M - Request Method | |
8299 | %O - Unescaped message result from external ACL helper | |
8300 | %o - Message result from external ACL helper | |
8301 | %p - Request Port number | |
8302 | %P - Request Protocol name | |
8303 | %R - Request URL path | |
8304 | %T - Timestamp in RFC 1123 format | |
8305 | %U - Full canonical URL from client | |
8306 | (HTTPS URLs terminate with *) | |
8307 | %u - Full canonical URL from client | |
8308 | %w - Admin email from squid.conf | |
8309 | %x - Error name | |
8310 | %% - Literal percent (%) code | |
8311 | ||
8312 | DOC_END | |
8313 | ||
8314 | COMMENT_START | |
8315 | OPTIONS INFLUENCING REQUEST FORWARDING | |
8316 | ----------------------------------------------------------------------------- | |
8317 | COMMENT_END | |
8318 | ||
8319 | NAME: nonhierarchical_direct | |
8320 | TYPE: onoff | |
8321 | LOC: Config.onoff.nonhierarchical_direct | |
8322 | DEFAULT: on | |
8323 | DOC_START | |
8324 | By default, Squid will send any non-hierarchical requests | |
8325 | (not cacheable request type) direct to origin servers. | |
8326 | ||
8327 | When this is set to "off", Squid will prefer to send these | |
8328 | requests to parents. | |
8329 | ||
8330 | Note that in most configurations, by turning this off you will only | |
8331 | add latency to these request without any improvement in global hit | |
8332 | ratio. | |
8333 | ||
8334 | This option only sets a preference. If the parent is unavailable a | |
8335 | direct connection to the origin server may still be attempted. To | |
8336 | completely prevent direct connections use never_direct. | |
8337 | DOC_END | |
8338 | ||
8339 | NAME: prefer_direct | |
8340 | TYPE: onoff | |
8341 | LOC: Config.onoff.prefer_direct | |
8342 | DEFAULT: off | |
8343 | DOC_START | |
8344 | Normally Squid tries to use parents for most requests. If you for some | |
8345 | reason like it to first try going direct and only use a parent if | |
8346 | going direct fails set this to on. | |
8347 | ||
8348 | By combining nonhierarchical_direct off and prefer_direct on you | |
8349 | can set up Squid to use a parent as a backup path if going direct | |
8350 | fails. | |
8351 | ||
8352 | Note: If you want Squid to use parents for all requests see | |
8353 | the never_direct directive. prefer_direct only modifies how Squid | |
8354 | acts on cacheable requests. | |
8355 | DOC_END | |
8356 | ||
8357 | NAME: cache_miss_revalidate | |
8358 | COMMENT: on|off | |
8359 | TYPE: onoff | |
8360 | DEFAULT: on | |
8361 | LOC: Config.onoff.cache_miss_revalidate | |
8362 | DOC_START | |
8363 | RFC 7232 defines a conditional request mechanism to prevent | |
8364 | response objects being unnecessarily transferred over the network. | |
8365 | If that mechanism is used by the client and a cache MISS occurs | |
8366 | it can prevent new cache entries being created. | |
8367 | ||
8368 | This option determines whether Squid on cache MISS will pass the | |
8369 | client revalidation request to the server or tries to fetch new | |
8370 | content for caching. It can be useful while the cache is mostly | |
8371 | empty to more quickly have the cache populated by generating | |
8372 | non-conditional GETs. | |
8373 | ||
8374 | When set to 'on' (default), Squid will pass all client If-* headers | |
8375 | to the server. This permits server responses without a cacheable | |
8376 | payload to be delivered and on MISS no new cache entry is created. | |
8377 | ||
8378 | When set to 'off' and if the request is cacheable, Squid will | |
8379 | remove the clients If-Modified-Since and If-None-Match headers from | |
8380 | the request sent to the server. This requests a 200 status response | |
8381 | from the server to create a new cache entry with. | |
8382 | DOC_END | |
8383 | ||
8384 | NAME: always_direct | |
8385 | TYPE: acl_access | |
8386 | LOC: Config.accessList.AlwaysDirect | |
8387 | DEFAULT: none | |
8388 | DEFAULT_DOC: Prevent any cache_peer being used for this request. | |
8389 | DOC_START | |
8390 | Usage: always_direct allow|deny [!]aclname ... | |
8391 | ||
8392 | Here you can use ACL elements to specify requests which should | |
8393 | ALWAYS be forwarded by Squid to the origin servers without using | |
8394 | any peers. For example, to always directly forward requests for | |
8395 | local servers ignoring any parents or siblings you may have use | |
8396 | something like: | |
8397 | ||
8398 | acl local-servers dstdomain my.domain.net | |
8399 | always_direct allow local-servers | |
8400 | ||
8401 | To always forward FTP requests directly, use | |
8402 | ||
8403 | acl FTP proto FTP | |
8404 | always_direct allow FTP | |
8405 | ||
8406 | NOTE: There is a similar, but opposite option named | |
8407 | 'never_direct'. You need to be aware that "always_direct deny | |
8408 | foo" is NOT the same thing as "never_direct allow foo". You | |
8409 | may need to use a deny rule to exclude a more-specific case of | |
8410 | some other rule. Example: | |
8411 | ||
8412 | acl local-external dstdomain external.foo.net | |
8413 | acl local-servers dstdomain .foo.net | |
8414 | always_direct deny local-external | |
8415 | always_direct allow local-servers | |
8416 | ||
8417 | NOTE: If your goal is to make the client forward the request | |
8418 | directly to the origin server bypassing Squid then this needs | |
8419 | to be done in the client configuration. Squid configuration | |
8420 | can only tell Squid how Squid should fetch the object. | |
8421 | ||
8422 | NOTE: This directive is not related to caching. The replies | |
8423 | is cached as usual even if you use always_direct. To not cache | |
8424 | the replies see the 'cache' directive. | |
8425 | ||
8426 | This clause supports both fast and slow acl types. | |
8427 | See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details. | |
8428 | DOC_END | |
8429 | ||
8430 | NAME: never_direct | |
8431 | TYPE: acl_access | |
8432 | LOC: Config.accessList.NeverDirect | |
8433 | DEFAULT: none | |
8434 | DEFAULT_DOC: Allow DNS results to be used for this request. | |
8435 | DOC_START | |
8436 | Usage: never_direct allow|deny [!]aclname ... | |
8437 | ||
8438 | never_direct is the opposite of always_direct. Please read | |
8439 | the description for always_direct if you have not already. | |
8440 | ||
8441 | With 'never_direct' you can use ACL elements to specify | |
8442 | requests which should NEVER be forwarded directly to origin | |
8443 | servers. For example, to force the use of a proxy for all | |
8444 | requests, except those in your local domain use something like: | |
8445 | ||
8446 | acl local-servers dstdomain .foo.net | |
8447 | never_direct deny local-servers | |
8448 | never_direct allow all | |
8449 | ||
8450 | or if Squid is inside a firewall and there are local intranet | |
8451 | servers inside the firewall use something like: | |
8452 | ||
8453 | acl local-intranet dstdomain .foo.net | |
8454 | acl local-external dstdomain external.foo.net | |
8455 | always_direct deny local-external | |
8456 | always_direct allow local-intranet | |
8457 | never_direct allow all | |
8458 | ||
8459 | This clause supports both fast and slow acl types. | |
8460 | See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details. | |
8461 | DOC_END | |
8462 | ||
8463 | COMMENT_START | |
8464 | ADVANCED NETWORKING OPTIONS | |
8465 | ----------------------------------------------------------------------------- | |
8466 | COMMENT_END | |
8467 | ||
8468 | NAME: incoming_udp_average incoming_icp_average | |
8469 | TYPE: int | |
8470 | DEFAULT: 6 | |
8471 | LOC: Config.comm_incoming.udp.average | |
8472 | DOC_START | |
8473 | Heavy voodoo here. I can't even believe you are reading this. | |
8474 | Are you crazy? Don't even think about adjusting these unless | |
8475 | you understand the algorithms in comm_select.c first! | |
8476 | DOC_END | |
8477 | ||
8478 | NAME: incoming_tcp_average incoming_http_average | |
8479 | TYPE: int | |
8480 | DEFAULT: 4 | |
8481 | LOC: Config.comm_incoming.tcp.average | |
8482 | DOC_START | |
8483 | Heavy voodoo here. I can't even believe you are reading this. | |
8484 | Are you crazy? Don't even think about adjusting these unless | |
8485 | you understand the algorithms in comm_select.c first! | |
8486 | DOC_END | |
8487 | ||
8488 | NAME: incoming_dns_average | |
8489 | TYPE: int | |
8490 | DEFAULT: 4 | |
8491 | LOC: Config.comm_incoming.dns.average | |
8492 | DOC_START | |
8493 | Heavy voodoo here. I can't even believe you are reading this. | |
8494 | Are you crazy? Don't even think about adjusting these unless | |
8495 | you understand the algorithms in comm_select.c first! | |
8496 | DOC_END | |
8497 | ||
8498 | NAME: min_udp_poll_cnt min_icp_poll_cnt | |
8499 | TYPE: int | |
8500 | DEFAULT: 8 | |
8501 | LOC: Config.comm_incoming.udp.min_poll | |
8502 | DOC_START | |
8503 | Heavy voodoo here. I can't even believe you are reading this. | |
8504 | Are you crazy? Don't even think about adjusting these unless | |
8505 | you understand the algorithms in comm_select.c first! | |
8506 | DOC_END | |
8507 | ||
8508 | NAME: min_dns_poll_cnt | |
8509 | TYPE: int | |
8510 | DEFAULT: 8 | |
8511 | LOC: Config.comm_incoming.dns.min_poll | |
8512 | DOC_START | |
8513 | Heavy voodoo here. I can't even believe you are reading this. | |
8514 | Are you crazy? Don't even think about adjusting these unless | |
8515 | you understand the algorithms in comm_select.c first! | |
8516 | DOC_END | |
8517 | ||
8518 | NAME: min_tcp_poll_cnt min_http_poll_cnt | |
8519 | TYPE: int | |
8520 | DEFAULT: 8 | |
8521 | LOC: Config.comm_incoming.tcp.min_poll | |
8522 | DOC_START | |
8523 | Heavy voodoo here. I can't even believe you are reading this. | |
8524 | Are you crazy? Don't even think about adjusting these unless | |
8525 | you understand the algorithms in comm_select.c first! | |
8526 | DOC_END | |
8527 | ||
8528 | NAME: accept_filter | |
8529 | TYPE: string | |
8530 | DEFAULT: none | |
8531 | LOC: Config.accept_filter | |
8532 | DOC_START | |
8533 | FreeBSD: | |
8534 | ||
8535 | The name of an accept(2) filter to install on Squid's | |
8536 | listen socket(s). This feature is perhaps specific to | |
8537 | FreeBSD and requires support in the kernel. | |
8538 | ||
8539 | The 'httpready' filter delays delivering new connections | |
8540 | to Squid until a full HTTP request has been received. | |
8541 | See the accf_http(9) man page for details. | |
8542 | ||
8543 | The 'dataready' filter delays delivering new connections | |
8544 | to Squid until there is some data to process. | |
8545 | See the accf_dataready(9) man page for details. | |
8546 | ||
8547 | Linux: | |
8548 | ||
8549 | The 'data' filter delays delivering of new connections | |
8550 | to Squid until there is some data to process by TCP_ACCEPT_DEFER. | |
8551 | You may optionally specify a number of seconds to wait by | |
8552 | 'data=N' where N is the number of seconds. Defaults to 30 | |
8553 | if not specified. See the tcp(7) man page for details. | |
8554 | EXAMPLE: | |
8555 | # FreeBSD | |
8556 | accept_filter httpready | |
8557 | # Linux | |
8558 | accept_filter data | |
8559 | DOC_END | |
8560 | ||
8561 | NAME: client_ip_max_connections | |
8562 | TYPE: int | |
8563 | LOC: Config.client_ip_max_connections | |
8564 | DEFAULT: -1 | |
8565 | DEFAULT_DOC: No limit. | |
8566 | DOC_START | |
8567 | Set an absolute limit on the number of connections a single | |
8568 | client IP can use. Any more than this and Squid will begin to drop | |
8569 | new connections from the client until it closes some links. | |
8570 | ||
8571 | Note that this is a global limit. It affects all HTTP, HTCP, Gopher and FTP | |
8572 | connections from the client. For finer control use the ACL access controls. | |
8573 | ||
8574 | Requires client_db to be enabled (the default). | |
8575 | ||
8576 | WARNING: This may noticably slow down traffic received via external proxies | |
8577 | or NAT devices and cause them to rebound error messages back to their clients. | |
8578 | DOC_END | |
8579 | ||
8580 | NAME: tcp_recv_bufsize | |
8581 | COMMENT: (bytes) | |
8582 | TYPE: b_size_t | |
8583 | DEFAULT: 0 bytes | |
8584 | DEFAULT_DOC: Use operating system TCP defaults. | |
8585 | LOC: Config.tcpRcvBufsz | |
8586 | DOC_START | |
8587 | Size of receive buffer to set for TCP sockets. Probably just | |
8588 | as easy to change your kernel's default. | |
8589 | Omit from squid.conf to use the default buffer size. | |
8590 | DOC_END | |
8591 | ||
8592 | COMMENT_START | |
8593 | ICAP OPTIONS | |
8594 | ----------------------------------------------------------------------------- | |
8595 | COMMENT_END | |
8596 | ||
8597 | NAME: icap_enable | |
8598 | TYPE: onoff | |
8599 | IFDEF: ICAP_CLIENT | |
8600 | COMMENT: on|off | |
8601 | LOC: Adaptation::Icap::TheConfig.onoff | |
8602 | DEFAULT: off | |
8603 | DOC_START | |
8604 | If you want to enable the ICAP module support, set this to on. | |
8605 | DOC_END | |
8606 | ||
8607 | NAME: icap_connect_timeout | |
8608 | TYPE: time_t | |
8609 | DEFAULT: none | |
8610 | LOC: Adaptation::Icap::TheConfig.connect_timeout_raw | |
8611 | IFDEF: ICAP_CLIENT | |
8612 | DOC_START | |
8613 | This parameter specifies how long to wait for the TCP connect to | |
8614 | the requested ICAP server to complete before giving up and either | |
8615 | terminating the HTTP transaction or bypassing the failure. | |
8616 | ||
8617 | The default for optional services is peer_connect_timeout. | |
8618 | The default for essential services is connect_timeout. | |
8619 | If this option is explicitly set, its value applies to all services. | |
8620 | DOC_END | |
8621 | ||
8622 | NAME: icap_io_timeout | |
8623 | COMMENT: time-units | |
8624 | TYPE: time_t | |
8625 | DEFAULT: none | |
8626 | DEFAULT_DOC: Use read_timeout. | |
8627 | LOC: Adaptation::Icap::TheConfig.io_timeout_raw | |
8628 | IFDEF: ICAP_CLIENT | |
8629 | DOC_START | |
8630 | This parameter specifies how long to wait for an I/O activity on | |
8631 | an established, active ICAP connection before giving up and | |
8632 | either terminating the HTTP transaction or bypassing the | |
8633 | failure. | |
8634 | DOC_END | |
8635 | ||
8636 | NAME: icap_service_failure_limit | |
8637 | COMMENT: limit [in memory-depth time-units] | |
8638 | TYPE: icap_service_failure_limit | |
8639 | IFDEF: ICAP_CLIENT | |
8640 | LOC: Adaptation::Icap::TheConfig | |
8641 | DEFAULT: 10 | |
8642 | DOC_START | |
8643 | The limit specifies the number of failures that Squid tolerates | |
8644 | when establishing a new TCP connection with an ICAP service. If | |
8645 | the number of failures exceeds the limit, the ICAP service is | |
8646 | not used for new ICAP requests until it is time to refresh its | |
8647 | OPTIONS. | |
8648 | ||
8649 | A negative value disables the limit. Without the limit, an ICAP | |
8650 | service will not be considered down due to connectivity failures | |
8651 | between ICAP OPTIONS requests. | |
8652 | ||
8653 | Squid forgets ICAP service failures older than the specified | |
8654 | value of memory-depth. The memory fading algorithm | |
8655 | is approximate because Squid does not remember individual | |
8656 | errors but groups them instead, splitting the option | |
8657 | value into ten time slots of equal length. | |
8658 | ||
8659 | When memory-depth is 0 and by default this option has no | |
8660 | effect on service failure expiration. | |
8661 | ||
8662 | Squid always forgets failures when updating service settings | |
8663 | using an ICAP OPTIONS transaction, regardless of this option | |
8664 | setting. | |
8665 | ||
8666 | For example, | |
8667 | # suspend service usage after 10 failures in 5 seconds: | |
8668 | icap_service_failure_limit 10 in 5 seconds | |
8669 | DOC_END | |
8670 | ||
8671 | NAME: icap_service_revival_delay | |
8672 | TYPE: int | |
8673 | IFDEF: ICAP_CLIENT | |
8674 | LOC: Adaptation::Icap::TheConfig.service_revival_delay | |
8675 | DEFAULT: 180 | |
8676 | DOC_START | |
8677 | The delay specifies the number of seconds to wait after an ICAP | |
8678 | OPTIONS request failure before requesting the options again. The | |
8679 | failed ICAP service is considered "down" until fresh OPTIONS are | |
8680 | fetched. | |
8681 | ||
8682 | The actual delay cannot be smaller than the hardcoded minimum | |
8683 | delay of 30 seconds. | |
8684 | DOC_END | |
8685 | ||
8686 | NAME: icap_preview_enable | |
8687 | TYPE: onoff | |
8688 | IFDEF: ICAP_CLIENT | |
8689 | COMMENT: on|off | |
8690 | LOC: Adaptation::Icap::TheConfig.preview_enable | |
8691 | DEFAULT: on | |
8692 | DOC_START | |
8693 | The ICAP Preview feature allows the ICAP server to handle the | |
8694 | HTTP message by looking only at the beginning of the message body | |
8695 | or even without receiving the body at all. In some environments, | |
8696 | previews greatly speedup ICAP processing. | |
8697 | ||
8698 | During an ICAP OPTIONS transaction, the server may tell Squid what | |
8699 | HTTP messages should be previewed and how big the preview should be. | |
8700 | Squid will not use Preview if the server did not request one. | |
8701 | ||
8702 | To disable ICAP Preview for all ICAP services, regardless of | |
8703 | individual ICAP server OPTIONS responses, set this option to "off". | |
8704 | Example: | |
8705 | icap_preview_enable off | |
8706 | DOC_END | |
8707 | ||
8708 | NAME: icap_preview_size | |
8709 | TYPE: int | |
8710 | IFDEF: ICAP_CLIENT | |
8711 | LOC: Adaptation::Icap::TheConfig.preview_size | |
8712 | DEFAULT: -1 | |
8713 | DEFAULT_DOC: No preview sent. | |
8714 | DOC_START | |
8715 | The default size of preview data to be sent to the ICAP server. | |
8716 | This value might be overwritten on a per server basis by OPTIONS requests. | |
8717 | DOC_END | |
8718 | ||
8719 | NAME: icap_206_enable | |
8720 | TYPE: onoff | |
8721 | IFDEF: ICAP_CLIENT | |
8722 | COMMENT: on|off | |
8723 | LOC: Adaptation::Icap::TheConfig.allow206_enable | |
8724 | DEFAULT: on | |
8725 | DOC_START | |
8726 | 206 (Partial Content) responses is an ICAP extension that allows the | |
8727 | ICAP agents to optionally combine adapted and original HTTP message | |
8728 | content. The decision to combine is postponed until the end of the | |
8729 | ICAP response. Squid supports Partial Content extension by default. | |
8730 | ||
8731 | Activation of the Partial Content extension is negotiated with each | |
8732 | ICAP service during OPTIONS exchange. Most ICAP servers should handle | |
8733 | negotation correctly even if they do not support the extension, but | |
8734 | some might fail. To disable Partial Content support for all ICAP | |
8735 | services and to avoid any negotiation, set this option to "off". | |
8736 | ||
8737 | Example: | |
8738 | icap_206_enable off | |
8739 | DOC_END | |
8740 | ||
8741 | NAME: icap_default_options_ttl | |
8742 | TYPE: int | |
8743 | IFDEF: ICAP_CLIENT | |
8744 | LOC: Adaptation::Icap::TheConfig.default_options_ttl | |
8745 | DEFAULT: 60 | |
8746 | DOC_START | |
8747 | The default TTL value for ICAP OPTIONS responses that don't have | |
8748 | an Options-TTL header. | |
8749 | DOC_END | |
8750 | ||
8751 | NAME: icap_persistent_connections | |
8752 | TYPE: onoff | |
8753 | IFDEF: ICAP_CLIENT | |
8754 | COMMENT: on|off | |
8755 | LOC: Adaptation::Icap::TheConfig.reuse_connections | |
8756 | DEFAULT: on | |
8757 | DOC_START | |
8758 | Whether or not Squid should use persistent connections to | |
8759 | an ICAP server. | |
8760 | DOC_END | |
8761 | ||
8762 | NAME: adaptation_send_client_ip icap_send_client_ip | |
8763 | TYPE: onoff | |
8764 | IFDEF: USE_ADAPTATION | |
8765 | COMMENT: on|off | |
8766 | LOC: Adaptation::Config::send_client_ip | |
8767 | DEFAULT: off | |
8768 | DOC_START | |
8769 | If enabled, Squid shares HTTP client IP information with adaptation | |
8770 | services. For ICAP, Squid adds the X-Client-IP header to ICAP requests. | |
8771 | For eCAP, Squid sets the libecap::metaClientIp transaction option. | |
8772 | ||
8773 | See also: adaptation_uses_indirect_client | |
8774 | DOC_END | |
8775 | ||
8776 | NAME: adaptation_send_username icap_send_client_username | |
8777 | TYPE: onoff | |
8778 | IFDEF: USE_ADAPTATION | |
8779 | COMMENT: on|off | |
8780 | LOC: Adaptation::Config::send_username | |
8781 | DEFAULT: off | |
8782 | DOC_START | |
8783 | This sends authenticated HTTP client username (if available) to | |
8784 | the adaptation service. | |
8785 | ||
8786 | For ICAP, the username value is encoded based on the | |
8787 | icap_client_username_encode option and is sent using the header | |
8788 | specified by the icap_client_username_header option. | |
8789 | DOC_END | |
8790 | ||
8791 | NAME: icap_client_username_header | |
8792 | TYPE: string | |
8793 | IFDEF: ICAP_CLIENT | |
8794 | LOC: Adaptation::Icap::TheConfig.client_username_header | |
8795 | DEFAULT: X-Client-Username | |
8796 | DOC_START | |
8797 | ICAP request header name to use for adaptation_send_username. | |
8798 | DOC_END | |
8799 | ||
8800 | NAME: icap_client_username_encode | |
8801 | TYPE: onoff | |
8802 | IFDEF: ICAP_CLIENT | |
8803 | COMMENT: on|off | |
8804 | LOC: Adaptation::Icap::TheConfig.client_username_encode | |
8805 | DEFAULT: off | |
8806 | DOC_START | |
8807 | Whether to base64 encode the authenticated client username. | |
8808 | DOC_END | |
8809 | ||
8810 | NAME: icap_service | |
8811 | TYPE: icap_service_type | |
8812 | IFDEF: ICAP_CLIENT | |
8813 | LOC: Adaptation::Icap::TheConfig | |
8814 | DEFAULT: none | |
8815 | DOC_START | |
8816 | Defines a single ICAP service using the following format: | |
8817 | ||
8818 | icap_service id vectoring_point uri [option ...] | |
8819 | ||
8820 | id: ID | |
8821 | an opaque identifier or name which is used to direct traffic to | |
8822 | this specific service. Must be unique among all adaptation | |
8823 | services in squid.conf. | |
8824 | ||
8825 | vectoring_point: reqmod_precache|reqmod_postcache|respmod_precache|respmod_postcache | |
8826 | This specifies at which point of transaction processing the | |
8827 | ICAP service should be activated. *_postcache vectoring points | |
8828 | are not yet supported. | |
8829 | ||
8830 | uri: icap://servername:port/servicepath | |
8831 | ICAP server and service location. | |
8832 | icaps://servername:port/servicepath | |
8833 | The "icap:" URI scheme is used for traditional ICAP server and | |
8834 | service location (default port is 1344, connections are not | |
8835 | encrypted). The "icaps:" URI scheme is for Secure ICAP | |
8836 | services that use SSL/TLS-encrypted ICAP connections (by | |
8837 | default, on port 11344). | |
8838 | ||
8839 | ICAP does not allow a single service to handle both REQMOD and RESPMOD | |
8840 | transactions. Squid does not enforce that requirement. You can specify | |
8841 | services with the same service_url and different vectoring_points. You | |
8842 | can even specify multiple identical services as long as their | |
8843 | service_names differ. | |
8844 | ||
8845 | To activate a service, use the adaptation_access directive. To group | |
8846 | services, use adaptation_service_chain and adaptation_service_set. | |
8847 | ||
8848 | Service options are separated by white space. ICAP services support | |
8849 | the following name=value options: | |
8850 | ||
8851 | bypass=on|off|1|0 | |
8852 | If set to 'on' or '1', the ICAP service is treated as | |
8853 | optional. If the service cannot be reached or malfunctions, | |
8854 | Squid will try to ignore any errors and process the message as | |
8855 | if the service was not enabled. No all ICAP errors can be | |
8856 | bypassed. If set to 0, the ICAP service is treated as | |
8857 | essential and all ICAP errors will result in an error page | |
8858 | returned to the HTTP client. | |
8859 | ||
8860 | Bypass is off by default: services are treated as essential. | |
8861 | ||
8862 | routing=on|off|1|0 | |
8863 | If set to 'on' or '1', the ICAP service is allowed to | |
8864 | dynamically change the current message adaptation plan by | |
8865 | returning a chain of services to be used next. The services | |
8866 | are specified using the X-Next-Services ICAP response header | |
8867 | value, formatted as a comma-separated list of service names. | |
8868 | Each named service should be configured in squid.conf. Other | |
8869 | services are ignored. An empty X-Next-Services value results | |
8870 | in an empty plan which ends the current adaptation. | |
8871 | ||
8872 | Dynamic adaptation plan may cross or cover multiple supported | |
8873 | vectoring points in their natural processing order. | |
8874 | ||
8875 | Routing is not allowed by default: the ICAP X-Next-Services | |
8876 | response header is ignored. | |
8877 | ||
8878 | ipv6=on|off | |
8879 | Only has effect on split-stack systems. The default on those systems | |
8880 | is to use IPv4-only connections. When set to 'on' this option will | |
8881 | make Squid use IPv6-only connections to contact this ICAP service. | |
8882 | ||
8883 | on-overload=block|bypass|wait|force | |
8884 | If the service Max-Connections limit has been reached, do | |
8885 | one of the following for each new ICAP transaction: | |
8886 | * block: send an HTTP error response to the client | |
8887 | * bypass: ignore the "over-connected" ICAP service | |
8888 | * wait: wait (in a FIFO queue) for an ICAP connection slot | |
8889 | * force: proceed, ignoring the Max-Connections limit | |
8890 | ||
8891 | In SMP mode with N workers, each worker assumes the service | |
8892 | connection limit is Max-Connections/N, even though not all | |
8893 | workers may use a given service. | |
8894 | ||
8895 | The default value is "bypass" if service is bypassable, | |
8896 | otherwise it is set to "wait". | |
8897 | ||
8898 | ||
8899 | max-conn=number | |
8900 | Use the given number as the Max-Connections limit, regardless | |
8901 | of the Max-Connections value given by the service, if any. | |
8902 | ||
8903 | connection-encryption=on|off | |
8904 | Determines the ICAP service effect on the connections_encrypted | |
8905 | ACL. | |
8906 | ||
8907 | The default is "on" for Secure ICAP services (i.e., those | |
8908 | with the icaps:// service URIs scheme) and "off" for plain ICAP | |
8909 | services. | |
8910 | ||
8911 | Does not affect ICAP connections (e.g., does not turn Secure | |
8912 | ICAP on or off). | |
8913 | ||
8914 | ==== ICAPS / TLS OPTIONS ==== | |
8915 | ||
8916 | These options are used for Secure ICAP (icaps://....) services only. | |
8917 | ||
8918 | tls-cert=/path/to/ssl/certificate | |
8919 | A client SSL certificate to use when connecting to | |
8920 | this icap server. | |
8921 | ||
8922 | tls-key=/path/to/ssl/key | |
8923 | The private TLS/SSL key corresponding to sslcert above. | |
8924 | If 'tls-key' is not specified 'tls-cert' is assumed to | |
8925 | reference a combined PEM format file containing both the | |
8926 | certificate and the key. | |
8927 | ||
8928 | tls-cipher=... The list of valid TLS/SSL ciphers to use when connecting | |
8929 | to this icap server. | |
8930 | ||
8931 | tls-min-version=1.N | |
8932 | The minimum TLS protocol version to permit. To control | |
8933 | SSLv3 use the tls-options= parameter. | |
8934 | Supported Values: 1.0 (default), 1.1, 1.2 | |
8935 | ||
8936 | tls-options=... Specify various OpenSSL library options: | |
8937 | ||
8938 | NO_SSLv3 Disallow the use of SSLv3 | |
8939 | ||
8940 | SINGLE_DH_USE | |
8941 | Always create a new key when using | |
8942 | temporary/ephemeral DH key exchanges | |
8943 | ||
8944 | ALL Enable various bug workarounds | |
8945 | suggested as "harmless" by OpenSSL | |
8946 | Be warned that this reduces SSL/TLS | |
8947 | strength to some attacks. | |
8948 | ||
8949 | See the OpenSSL SSL_CTX_set_options documentation for a | |
8950 | more complete list. Options relevant only to SSLv2 are | |
8951 | not supported. | |
8952 | ||
8953 | tls-cafile= PEM file containing CA certificates to use when verifying | |
8954 | the icap server certificate. | |
8955 | Use to specify intermediate CA certificate(s) if not sent | |
8956 | by the server. Or the full CA chain for the server when | |
8957 | using the tls-default-ca=off flag. | |
8958 | May be repeated to load multiple files. | |
8959 | ||
8960 | tls-capath=... A directory containing additional CA certificates to | |
8961 | use when verifying the icap server certificate. | |
8962 | Requires OpenSSL or LibreSSL. | |
8963 | ||
8964 | tls-crlfile=... A certificate revocation list file to use when | |
8965 | verifying the icap server certificate. | |
8966 | ||
8967 | tls-flags=... Specify various flags modifying the Squid TLS implementation: | |
8968 | ||
8969 | DONT_VERIFY_PEER | |
8970 | Accept certificates even if they fail to | |
8971 | verify. | |
8972 | DONT_VERIFY_DOMAIN | |
8973 | Don't verify the icap server certificate | |
8974 | matches the server name | |
8975 | ||
8976 | tls-default-ca[=off] | |
8977 | Whether to use the system Trusted CAs. Default is ON. | |
8978 | ||
8979 | tls-domain= The icap server name as advertised in it's certificate. | |
8980 | Used for verifying the correctness of the received icap | |
8981 | server certificate. If not specified the icap server | |
8982 | hostname extracted from ICAP URI will be used. | |
8983 | ||
8984 | Older icap_service format without optional named parameters is | |
8985 | deprecated but supported for backward compatibility. | |
8986 | ||
8987 | Example: | |
8988 | icap_service svcBlocker reqmod_precache icap://icap1.mydomain.net:1344/reqmod bypass=0 | |
8989 | icap_service svcLogger reqmod_precache icaps://icap2.mydomain.net:11344/reqmod routing=on | |
8990 | DOC_END | |
8991 | ||
8992 | NAME: icap_class | |
8993 | TYPE: icap_class_type | |
8994 | IFDEF: ICAP_CLIENT | |
8995 | LOC: none | |
8996 | DEFAULT: none | |
8997 | DOC_START | |
8998 | This deprecated option was documented to define an ICAP service | |
8999 | chain, even though it actually defined a set of similar, redundant | |
9000 | services, and the chains were not supported. | |
9001 | ||
9002 | To define a set of redundant services, please use the | |
9003 | adaptation_service_set directive. For service chains, use | |
9004 | adaptation_service_chain. | |
9005 | DOC_END | |
9006 | ||
9007 | NAME: icap_access | |
9008 | TYPE: icap_access_type | |
9009 | IFDEF: ICAP_CLIENT | |
9010 | LOC: none | |
9011 | DEFAULT: none | |
9012 | DOC_START | |
9013 | This option is deprecated. Please use adaptation_access, which | |
9014 | has the same ICAP functionality, but comes with better | |
9015 | documentation, and eCAP support. | |
9016 | DOC_END | |
9017 | ||
9018 | COMMENT_START | |
9019 | eCAP OPTIONS | |
9020 | ----------------------------------------------------------------------------- | |
9021 | COMMENT_END | |
9022 | ||
9023 | NAME: ecap_enable | |
9024 | TYPE: onoff | |
9025 | IFDEF: USE_ECAP | |
9026 | COMMENT: on|off | |
9027 | LOC: Adaptation::Ecap::TheConfig.onoff | |
9028 | DEFAULT: off | |
9029 | DOC_START | |
9030 | Controls whether eCAP support is enabled. | |
9031 | DOC_END | |
9032 | ||
9033 | NAME: ecap_service | |
9034 | TYPE: ecap_service_type | |
9035 | IFDEF: USE_ECAP | |
9036 | LOC: Adaptation::Ecap::TheConfig | |
9037 | DEFAULT: none | |
9038 | DOC_START | |
9039 | Defines a single eCAP service | |
9040 | ||
9041 | ecap_service id vectoring_point uri [option ...] | |
9042 | ||
9043 | id: ID | |
9044 | an opaque identifier or name which is used to direct traffic to | |
9045 | this specific service. Must be unique among all adaptation | |
9046 | services in squid.conf. | |
9047 | ||
9048 | vectoring_point: reqmod_precache|reqmod_postcache|respmod_precache|respmod_postcache | |
9049 | This specifies at which point of transaction processing the | |
9050 | eCAP service should be activated. *_postcache vectoring points | |
9051 | are not yet supported. | |
9052 | ||
9053 | uri: ecap://vendor/service_name?custom&cgi=style¶meters=optional | |
9054 | Squid uses the eCAP service URI to match this configuration | |
9055 | line with one of the dynamically loaded services. Each loaded | |
9056 | eCAP service must have a unique URI. Obtain the right URI from | |
9057 | the service provider. | |
9058 | ||
9059 | To activate a service, use the adaptation_access directive. To group | |
9060 | services, use adaptation_service_chain and adaptation_service_set. | |
9061 | ||
9062 | Service options are separated by white space. eCAP services support | |
9063 | the following name=value options: | |
9064 | ||
9065 | bypass=on|off|1|0 | |
9066 | If set to 'on' or '1', the eCAP service is treated as optional. | |
9067 | If the service cannot be reached or malfunctions, Squid will try | |
9068 | to ignore any errors and process the message as if the service | |
9069 | was not enabled. No all eCAP errors can be bypassed. | |
9070 | If set to 'off' or '0', the eCAP service is treated as essential | |
9071 | and all eCAP errors will result in an error page returned to the | |
9072 | HTTP client. | |
9073 | ||
9074 | Bypass is off by default: services are treated as essential. | |
9075 | ||
9076 | routing=on|off|1|0 | |
9077 | If set to 'on' or '1', the eCAP service is allowed to | |
9078 | dynamically change the current message adaptation plan by | |
9079 | returning a chain of services to be used next. | |
9080 | ||
9081 | Dynamic adaptation plan may cross or cover multiple supported | |
9082 | vectoring points in their natural processing order. | |
9083 | ||
9084 | Routing is not allowed by default. | |
9085 | ||
9086 | connection-encryption=on|off | |
9087 | Determines the eCAP service effect on the connections_encrypted | |
9088 | ACL. | |
9089 | ||
9090 | Defaults to "on", which does not taint the master transaction | |
9091 | w.r.t. that ACL. | |
9092 | ||
9093 | Does not affect eCAP API calls. | |
9094 | ||
9095 | Older ecap_service format without optional named parameters is | |
9096 | deprecated but supported for backward compatibility. | |
9097 | ||
9098 | ||
9099 | Example: | |
9100 | ecap_service s1 reqmod_precache ecap://filters.R.us/leakDetector?on_error=block bypass=off | |
9101 | ecap_service s2 respmod_precache ecap://filters.R.us/virusFilter config=/etc/vf.cfg bypass=on | |
9102 | DOC_END | |
9103 | ||
9104 | NAME: loadable_modules | |
9105 | TYPE: wordlist | |
9106 | IFDEF: USE_LOADABLE_MODULES | |
9107 | LOC: Config.loadable_module_names | |
9108 | DEFAULT: none | |
9109 | DOC_START | |
9110 | Instructs Squid to load the specified dynamic module(s) or activate | |
9111 | preloaded module(s). | |
9112 | Example: | |
9113 | loadable_modules @DEFAULT_PREFIX@/lib/MinimalAdapter.so | |
9114 | DOC_END | |
9115 | ||
9116 | COMMENT_START | |
9117 | MESSAGE ADAPTATION OPTIONS | |
9118 | ----------------------------------------------------------------------------- | |
9119 | COMMENT_END | |
9120 | ||
9121 | NAME: adaptation_service_set | |
9122 | TYPE: adaptation_service_set_type | |
9123 | IFDEF: USE_ADAPTATION | |
9124 | LOC: none | |
9125 | DEFAULT: none | |
9126 | DOC_START | |
9127 | ||
9128 | Configures an ordered set of similar, redundant services. This is | |
9129 | useful when hot standby or backup adaptation servers are available. | |
9130 | ||
9131 | adaptation_service_set set_name service_name1 service_name2 ... | |
9132 | ||
9133 | The named services are used in the set declaration order. The first | |
9134 | applicable adaptation service from the set is used first. The next | |
9135 | applicable service is tried if and only if the transaction with the | |
9136 | previous service fails and the message waiting to be adapted is still | |
9137 | intact. | |
9138 | ||
9139 | When adaptation starts, broken services are ignored as if they were | |
9140 | not a part of the set. A broken service is a down optional service. | |
9141 | ||
9142 | The services in a set must be attached to the same vectoring point | |
9143 | (e.g., pre-cache) and use the same adaptation method (e.g., REQMOD). | |
9144 | ||
9145 | If all services in a set are optional then adaptation failures are | |
9146 | bypassable. If all services in the set are essential, then a | |
9147 | transaction failure with one service may still be retried using | |
9148 | another service from the set, but when all services fail, the master | |
9149 | transaction fails as well. | |
9150 | ||
9151 | A set may contain a mix of optional and essential services, but that | |
9152 | is likely to lead to surprising results because broken services become | |
9153 | ignored (see above), making previously bypassable failures fatal. | |
9154 | Technically, it is the bypassability of the last failed service that | |
9155 | matters. | |
9156 | ||
9157 | See also: adaptation_access adaptation_service_chain | |
9158 | ||
9159 | Example: | |
9160 | adaptation_service_set svcBlocker urlFilterPrimary urlFilterBackup | |
9161 | adaptation service_set svcLogger loggerLocal loggerRemote | |
9162 | DOC_END | |
9163 | ||
9164 | NAME: adaptation_service_chain | |
9165 | TYPE: adaptation_service_chain_type | |
9166 | IFDEF: USE_ADAPTATION | |
9167 | LOC: none | |
9168 | DEFAULT: none | |
9169 | DOC_START | |
9170 | ||
9171 | Configures a list of complementary services that will be applied | |
9172 | one-by-one, forming an adaptation chain or pipeline. This is useful | |
9173 | when Squid must perform different adaptations on the same message. | |
9174 | ||
9175 | adaptation_service_chain chain_name service_name1 svc_name2 ... | |
9176 | ||
9177 | The named services are used in the chain declaration order. The first | |
9178 | applicable adaptation service from the chain is used first. The next | |
9179 | applicable service is applied to the successful adaptation results of | |
9180 | the previous service in the chain. | |
9181 | ||
9182 | When adaptation starts, broken services are ignored as if they were | |
9183 | not a part of the chain. A broken service is a down optional service. | |
9184 | ||
9185 | Request satisfaction terminates the adaptation chain because Squid | |
9186 | does not currently allow declaration of RESPMOD services at the | |
9187 | "reqmod_precache" vectoring point (see icap_service or ecap_service). | |
9188 | ||
9189 | The services in a chain must be attached to the same vectoring point | |
9190 | (e.g., pre-cache) and use the same adaptation method (e.g., REQMOD). | |
9191 | ||
9192 | A chain may contain a mix of optional and essential services. If an | |
9193 | essential adaptation fails (or the failure cannot be bypassed for | |
9194 | other reasons), the master transaction fails. Otherwise, the failure | |
9195 | is bypassed as if the failed adaptation service was not in the chain. | |
9196 | ||
9197 | See also: adaptation_access adaptation_service_set | |
9198 | ||
9199 | Example: | |
9200 | adaptation_service_chain svcRequest requestLogger urlFilter leakDetector | |
9201 | DOC_END | |
9202 | ||
9203 | NAME: adaptation_access | |
9204 | TYPE: adaptation_access_type | |
9205 | IFDEF: USE_ADAPTATION | |
9206 | LOC: none | |
9207 | DEFAULT: none | |
9208 | DEFAULT_DOC: Allow, unless rules exist in squid.conf. | |
9209 | DOC_START | |
9210 | Sends an HTTP transaction to an ICAP or eCAP adaptation service. | |
9211 | ||
9212 | adaptation_access service_name allow|deny [!]aclname... | |
9213 | adaptation_access set_name allow|deny [!]aclname... | |
9214 | ||
9215 | At each supported vectoring point, the adaptation_access | |
9216 | statements are processed in the order they appear in this | |
9217 | configuration file. Statements pointing to the following services | |
9218 | are ignored (i.e., skipped without checking their ACL): | |
9219 | ||
9220 | - services serving different vectoring points | |
9221 | - "broken-but-bypassable" services | |
9222 | - "up" services configured to ignore such transactions | |
9223 | (e.g., based on the ICAP Transfer-Ignore header). | |
9224 | ||
9225 | When a set_name is used, all services in the set are checked | |
9226 | using the same rules, to find the first applicable one. See | |
9227 | adaptation_service_set for details. | |
9228 | ||
9229 | If an access list is checked and there is a match, the | |
9230 | processing stops: For an "allow" rule, the corresponding | |
9231 | adaptation service is used for the transaction. For a "deny" | |
9232 | rule, no adaptation service is activated. | |
9233 | ||
9234 | It is currently not possible to apply more than one adaptation | |
9235 | service at the same vectoring point to the same HTTP transaction. | |
9236 | ||
9237 | See also: icap_service and ecap_service | |
9238 | ||
9239 | Example: | |
9240 | adaptation_access service_1 allow all | |
9241 | DOC_END | |
9242 | ||
9243 | NAME: adaptation_service_iteration_limit | |
9244 | TYPE: int | |
9245 | IFDEF: USE_ADAPTATION | |
9246 | LOC: Adaptation::Config::service_iteration_limit | |
9247 | DEFAULT: 16 | |
9248 | DOC_START | |
9249 | Limits the number of iterations allowed when applying adaptation | |
9250 | services to a message. If your longest adaptation set or chain | |
9251 | may have more than 16 services, increase the limit beyond its | |
9252 | default value of 16. If detecting infinite iteration loops sooner | |
9253 | is critical, make the iteration limit match the actual number | |
9254 | of services in your longest adaptation set or chain. | |
9255 | ||
9256 | Infinite adaptation loops are most likely with routing services. | |
9257 | ||
9258 | See also: icap_service routing=1 | |
9259 | DOC_END | |
9260 | ||
9261 | NAME: adaptation_masterx_shared_names | |
9262 | TYPE: string | |
9263 | IFDEF: USE_ADAPTATION | |
9264 | LOC: Adaptation::Config::masterx_shared_name | |
9265 | DEFAULT: none | |
9266 | DOC_START | |
9267 | For each master transaction (i.e., the HTTP request and response | |
9268 | sequence, including all related ICAP and eCAP exchanges), Squid | |
9269 | maintains a table of metadata. The table entries are (name, value) | |
9270 | pairs shared among eCAP and ICAP exchanges. The table is destroyed | |
9271 | with the master transaction. | |
9272 | ||
9273 | This option specifies the table entry names that Squid must accept | |
9274 | from and forward to the adaptation transactions. | |
9275 | ||
9276 | An ICAP REQMOD or RESPMOD transaction may set an entry in the | |
9277 | shared table by returning an ICAP header field with a name | |
9278 | specified in adaptation_masterx_shared_names. | |
9279 | ||
9280 | An eCAP REQMOD or RESPMOD transaction may set an entry in the | |
9281 | shared table by implementing the libecap::visitEachOption() API | |
9282 | to provide an option with a name specified in | |
9283 | adaptation_masterx_shared_names. | |
9284 | ||
9285 | Squid will store and forward the set entry to subsequent adaptation | |
9286 | transactions within the same master transaction scope. | |
9287 | ||
9288 | Only one shared entry name is supported at this time. | |
9289 | ||
9290 | Example: | |
9291 | # share authentication information among ICAP services | |
9292 | adaptation_masterx_shared_names X-Subscriber-ID | |
9293 | DOC_END | |
9294 | ||
9295 | NAME: adaptation_meta | |
9296 | TYPE: note | |
9297 | IFDEF: USE_ADAPTATION | |
9298 | LOC: Adaptation::Config::metaHeaders | |
9299 | DEFAULT: none | |
9300 | DOC_START | |
9301 | This option allows Squid administrator to add custom ICAP request | |
9302 | headers or eCAP options to Squid ICAP requests or eCAP transactions. | |
9303 | Use it to pass custom authentication tokens and other | |
9304 | transaction-state related meta information to an ICAP/eCAP service. | |
9305 | ||
9306 | The addition of a meta header is ACL-driven: | |
9307 | adaptation_meta name value [!]aclname ... | |
9308 | ||
9309 | Processing for a given header name stops after the first ACL list match. | |
9310 | Thus, it is impossible to add two headers with the same name. If no ACL | |
9311 | lists match for a given header name, no such header is added. For | |
9312 | example: | |
9313 | ||
9314 | # do not debug transactions except for those that need debugging | |
9315 | adaptation_meta X-Debug 1 needs_debugging | |
9316 | ||
9317 | # log all transactions except for those that must remain secret | |
9318 | adaptation_meta X-Log 1 !keep_secret | |
9319 | ||
9320 | # mark transactions from users in the "G 1" group | |
9321 | adaptation_meta X-Authenticated-Groups "G 1" authed_as_G1 | |
9322 | ||
9323 | The "value" parameter may be a regular squid.conf token or a "double | |
9324 | quoted string". Within the quoted string, use backslash (\) to escape | |
9325 | any character, which is currently only useful for escaping backslashes | |
9326 | and double quotes. For example, | |
9327 | "this string has one backslash (\\) and two \"quotes\"" | |
9328 | ||
9329 | Used adaptation_meta header values may be logged via %note | |
9330 | logformat code. If multiple adaptation_meta headers with the same name | |
9331 | are used during master transaction lifetime, the header values are | |
9332 | logged in the order they were used and duplicate values are ignored | |
9333 | (only the first repeated value will be logged). | |
9334 | DOC_END | |
9335 | ||
9336 | NAME: icap_retry | |
9337 | TYPE: acl_access | |
9338 | IFDEF: ICAP_CLIENT | |
9339 | LOC: Adaptation::Icap::TheConfig.repeat | |
9340 | DEFAULT_IF_NONE: deny all | |
9341 | DOC_START | |
9342 | This ACL determines which retriable ICAP transactions are | |
9343 | retried. Transactions that received a complete ICAP response | |
9344 | and did not have to consume or produce HTTP bodies to receive | |
9345 | that response are usually retriable. | |
9346 | ||
9347 | icap_retry allow|deny [!]aclname ... | |
9348 | ||
9349 | Squid automatically retries some ICAP I/O timeouts and errors | |
9350 | due to persistent connection race conditions. | |
9351 | ||
9352 | See also: icap_retry_limit | |
9353 | DOC_END | |
9354 | ||
9355 | NAME: icap_retry_limit | |
9356 | TYPE: int | |
9357 | IFDEF: ICAP_CLIENT | |
9358 | LOC: Adaptation::Icap::TheConfig.repeat_limit | |
9359 | DEFAULT: 0 | |
9360 | DEFAULT_DOC: No retries are allowed. | |
9361 | DOC_START | |
9362 | Limits the number of retries allowed. | |
9363 | ||
9364 | Communication errors due to persistent connection race | |
9365 | conditions are unavoidable, automatically retried, and do not | |
9366 | count against this limit. | |
9367 | ||
9368 | See also: icap_retry | |
9369 | DOC_END | |
9370 | ||
9371 | ||
9372 | COMMENT_START | |
9373 | DNS OPTIONS | |
9374 | ----------------------------------------------------------------------------- | |
9375 | COMMENT_END | |
9376 | ||
9377 | NAME: check_hostnames | |
9378 | TYPE: onoff | |
9379 | DEFAULT: off | |
9380 | LOC: Config.onoff.check_hostnames | |
9381 | DOC_START | |
9382 | For security and stability reasons Squid can check | |
9383 | hostnames for Internet standard RFC compliance. If you want | |
9384 | Squid to perform these checks turn this directive on. | |
9385 | DOC_END | |
9386 | ||
9387 | NAME: allow_underscore | |
9388 | TYPE: onoff | |
9389 | DEFAULT: on | |
9390 | LOC: Config.onoff.allow_underscore | |
9391 | DOC_START | |
9392 | Underscore characters is not strictly allowed in Internet hostnames | |
9393 | but nevertheless used by many sites. Set this to off if you want | |
9394 | Squid to be strict about the standard. | |
9395 | This check is performed only when check_hostnames is set to on. | |
9396 | DOC_END | |
9397 | ||
9398 | NAME: dns_retransmit_interval | |
9399 | TYPE: time_msec | |
9400 | DEFAULT: 5 seconds | |
9401 | LOC: Config.Timeout.idns_retransmit | |
9402 | DOC_START | |
9403 | Initial retransmit interval for DNS queries. The interval is | |
9404 | doubled each time all configured DNS servers have been tried. | |
9405 | DOC_END | |
9406 | ||
9407 | NAME: dns_timeout | |
9408 | TYPE: time_msec | |
9409 | DEFAULT: 30 seconds | |
9410 | LOC: Config.Timeout.idns_query | |
9411 | DOC_START | |
9412 | DNS Query timeout. If no response is received to a DNS query | |
9413 | within this time all DNS servers for the queried domain | |
9414 | are assumed to be unavailable. | |
9415 | DOC_END | |
9416 | ||
9417 | NAME: dns_packet_max | |
9418 | TYPE: b_ssize_t | |
9419 | DEFAULT_DOC: EDNS disabled | |
9420 | DEFAULT: none | |
9421 | LOC: Config.dns.packet_max | |
9422 | DOC_START | |
9423 | Maximum number of bytes packet size to advertise via EDNS. | |
9424 | Set to "none" to disable EDNS large packet support. | |
9425 | ||
9426 | For legacy reasons DNS UDP replies will default to 512 bytes which | |
9427 | is too small for many responses. EDNS provides a means for Squid to | |
9428 | negotiate receiving larger responses back immediately without having | |
9429 | to failover with repeat requests. Responses larger than this limit | |
9430 | will retain the old behaviour of failover to TCP DNS. | |
9431 | ||
9432 | Squid has no real fixed limit internally, but allowing packet sizes | |
9433 | over 1500 bytes requires network jumbogram support and is usually not | |
9434 | necessary. | |
9435 | ||
9436 | WARNING: The RFC also indicates that some older resolvers will reply | |
9437 | with failure of the whole request if the extension is added. Some | |
9438 | resolvers have already been identified which will reply with mangled | |
9439 | EDNS response on occasion. Usually in response to many-KB jumbogram | |
9440 | sizes being advertised by Squid. | |
9441 | Squid will currently treat these both as an unable-to-resolve domain | |
9442 | even if it would be resolvable without EDNS. | |
9443 | DOC_END | |
9444 | ||
9445 | NAME: dns_defnames | |
9446 | COMMENT: on|off | |
9447 | TYPE: onoff | |
9448 | DEFAULT: off | |
9449 | DEFAULT_DOC: Search for single-label domain names is disabled. | |
9450 | LOC: Config.onoff.res_defnames | |
9451 | DOC_START | |
9452 | Normally the RES_DEFNAMES resolver option is disabled | |
9453 | (see res_init(3)). This prevents caches in a hierarchy | |
9454 | from interpreting single-component hostnames locally. To allow | |
9455 | Squid to handle single-component names, enable this option. | |
9456 | DOC_END | |
9457 | ||
9458 | NAME: dns_multicast_local | |
9459 | COMMENT: on|off | |
9460 | TYPE: onoff | |
9461 | DEFAULT: off | |
9462 | DEFAULT_DOC: Search for .local and .arpa names is disabled. | |
9463 | LOC: Config.onoff.dns_mdns | |
9464 | DOC_START | |
9465 | When set to on, Squid sends multicast DNS lookups on the local | |
9466 | network for domains ending in .local and .arpa. | |
9467 | This enables local servers and devices to be contacted in an | |
9468 | ad-hoc or zero-configuration network environment. | |
9469 | DOC_END | |
9470 | ||
9471 | NAME: dns_nameservers | |
9472 | TYPE: SBufList | |
9473 | DEFAULT: none | |
9474 | DEFAULT_DOC: Use operating system definitions | |
9475 | LOC: Config.dns.nameservers | |
9476 | DOC_START | |
9477 | Use this if you want to specify a list of DNS name servers | |
9478 | (IP addresses) to use instead of those given in your | |
9479 | /etc/resolv.conf file. | |
9480 | ||
9481 | On Windows platforms, if no value is specified here or in | |
9482 | the /etc/resolv.conf file, the list of DNS name servers are | |
9483 | taken from the Windows registry, both static and dynamic DHCP | |
9484 | configurations are supported. | |
9485 | ||
9486 | Example: dns_nameservers 10.0.0.1 192.172.0.4 | |
9487 | DOC_END | |
9488 | ||
9489 | NAME: hosts_file | |
9490 | TYPE: string | |
9491 | DEFAULT: @DEFAULT_HOSTS@ | |
9492 | LOC: Config.etcHostsPath | |
9493 | DOC_START | |
9494 | Location of the host-local IP name-address associations | |
9495 | database. Most Operating Systems have such a file on different | |
9496 | default locations: | |
9497 | - Un*X & Linux: /etc/hosts | |
9498 | - Windows NT/2000: %SystemRoot%\system32\drivers\etc\hosts | |
9499 | (%SystemRoot% value install default is c:\winnt) | |
9500 | - Windows XP/2003: %SystemRoot%\system32\drivers\etc\hosts | |
9501 | (%SystemRoot% value install default is c:\windows) | |
9502 | - Windows 9x/Me: %windir%\hosts | |
9503 | (%windir% value is usually c:\windows) | |
9504 | - Cygwin: /etc/hosts | |
9505 | ||
9506 | The file contains newline-separated definitions, in the | |
9507 | form ip_address_in_dotted_form name [name ...] names are | |
9508 | whitespace-separated. Lines beginning with an hash (#) | |
9509 | character are comments. | |
9510 | ||
9511 | The file is checked at startup and upon configuration. | |
9512 | If set to 'none', it won't be checked. | |
9513 | If append_domain is used, that domain will be added to | |
9514 | domain-local (i.e. not containing any dot character) host | |
9515 | definitions. | |
9516 | DOC_END | |
9517 | ||
9518 | NAME: append_domain | |
9519 | TYPE: string | |
9520 | LOC: Config.appendDomain | |
9521 | DEFAULT: none | |
9522 | DEFAULT_DOC: Use operating system definitions | |
9523 | DOC_START | |
9524 | Appends local domain name to hostnames without any dots in | |
9525 | them. append_domain must begin with a period. | |
9526 | ||
9527 | Be warned there are now Internet names with no dots in | |
9528 | them using only top-domain names, so setting this may | |
9529 | cause some Internet sites to become unavailable. | |
9530 | ||
9531 | Example: | |
9532 | append_domain .yourdomain.com | |
9533 | DOC_END | |
9534 | ||
9535 | NAME: ignore_unknown_nameservers | |
9536 | TYPE: onoff | |
9537 | LOC: Config.onoff.ignore_unknown_nameservers | |
9538 | DEFAULT: on | |
9539 | DOC_START | |
9540 | By default Squid checks that DNS responses are received | |
9541 | from the same IP addresses they are sent to. If they | |
9542 | don't match, Squid ignores the response and writes a warning | |
9543 | message to cache.log. You can allow responses from unknown | |
9544 | nameservers by setting this option to 'off'. | |
9545 | DOC_END | |
9546 | ||
9547 | NAME: dns_v4_first | |
9548 | TYPE: onoff | |
9549 | DEFAULT: off | |
9550 | LOC: Config.dns.v4_first | |
9551 | DOC_START | |
9552 | With the IPv6 Internet being as fast or faster than IPv4 Internet | |
9553 | for most networks Squid prefers to contact websites over IPv6. | |
9554 | ||
9555 | This option reverses the order of preference to make Squid contact | |
9556 | dual-stack websites over IPv4 first. Squid will still perform both | |
9557 | IPv6 and IPv4 DNS lookups before connecting. | |
9558 | ||
9559 | WARNING: | |
9560 | This option will restrict the situations under which IPv6 | |
9561 | connectivity is used (and tested). Hiding network problems | |
9562 | which would otherwise be detected and warned about. | |
9563 | DOC_END | |
9564 | ||
9565 | NAME: ipcache_size | |
9566 | COMMENT: (number of entries) | |
9567 | TYPE: int | |
9568 | DEFAULT: 1024 | |
9569 | LOC: Config.ipcache.size | |
9570 | DOC_START | |
9571 | Maximum number of DNS IP cache entries. | |
9572 | DOC_END | |
9573 | ||
9574 | NAME: ipcache_low | |
9575 | COMMENT: (percent) | |
9576 | TYPE: int | |
9577 | DEFAULT: 90 | |
9578 | LOC: Config.ipcache.low | |
9579 | DOC_NONE | |
9580 | ||
9581 | NAME: ipcache_high | |
9582 | COMMENT: (percent) | |
9583 | TYPE: int | |
9584 | DEFAULT: 95 | |
9585 | LOC: Config.ipcache.high | |
9586 | DOC_START | |
9587 | The size, low-, and high-water marks for the IP cache. | |
9588 | DOC_END | |
9589 | ||
9590 | NAME: fqdncache_size | |
9591 | COMMENT: (number of entries) | |
9592 | TYPE: int | |
9593 | DEFAULT: 1024 | |
9594 | LOC: Config.fqdncache.size | |
9595 | DOC_START | |
9596 | Maximum number of FQDN cache entries. | |
9597 | DOC_END | |
9598 | ||
9599 | COMMENT_START | |
9600 | MISCELLANEOUS | |
9601 | ----------------------------------------------------------------------------- | |
9602 | COMMENT_END | |
9603 | ||
9604 | NAME: configuration_includes_quoted_values | |
9605 | COMMENT: on|off | |
9606 | TYPE: configuration_includes_quoted_values | |
9607 | DEFAULT: off | |
9608 | LOC: ConfigParser::RecognizeQuotedValues | |
9609 | DOC_START | |
9610 | If set, Squid will recognize each "quoted string" after a configuration | |
9611 | directive as a single parameter. The quotes are stripped before the | |
9612 | parameter value is interpreted or used. | |
9613 | See "Values with spaces, quotes, and other special characters" | |
9614 | section for more details. | |
9615 | DOC_END | |
9616 | ||
9617 | NAME: memory_pools | |
9618 | COMMENT: on|off | |
9619 | TYPE: onoff | |
9620 | DEFAULT: on | |
9621 | LOC: Config.onoff.mem_pools | |
9622 | DOC_START | |
9623 | If set, Squid will keep pools of allocated (but unused) memory | |
9624 | available for future use. If memory is a premium on your | |
9625 | system and you believe your malloc library outperforms Squid | |
9626 | routines, disable this. | |
9627 | DOC_END | |
9628 | ||
9629 | NAME: memory_pools_limit | |
9630 | COMMENT: (bytes) | |
9631 | TYPE: b_int64_t | |
9632 | DEFAULT: 5 MB | |
9633 | LOC: Config.MemPools.limit | |
9634 | DOC_START | |
9635 | Used only with memory_pools on: | |
9636 | memory_pools_limit 50 MB | |
9637 | ||
9638 | If set to a non-zero value, Squid will keep at most the specified | |
9639 | limit of allocated (but unused) memory in memory pools. All free() | |
9640 | requests that exceed this limit will be handled by your malloc | |
9641 | library. Squid does not pre-allocate any memory, just safe-keeps | |
9642 | objects that otherwise would be free()d. Thus, it is safe to set | |
9643 | memory_pools_limit to a reasonably high value even if your | |
9644 | configuration will use less memory. | |
9645 | ||
9646 | If set to none, Squid will keep all memory it can. That is, there | |
9647 | will be no limit on the total amount of memory used for safe-keeping. | |
9648 | ||
9649 | To disable memory allocation optimization, do not set | |
9650 | memory_pools_limit to 0 or none. Set memory_pools to "off" instead. | |
9651 | ||
9652 | An overhead for maintaining memory pools is not taken into account | |
9653 | when the limit is checked. This overhead is close to four bytes per | |
9654 | object kept. However, pools may actually _save_ memory because of | |
9655 | reduced memory thrashing in your malloc library. | |
9656 | DOC_END | |
9657 | ||
9658 | NAME: forwarded_for | |
9659 | COMMENT: on|off|transparent|truncate|delete | |
9660 | TYPE: string | |
9661 | DEFAULT: on | |
9662 | LOC: opt_forwarded_for | |
9663 | DOC_START | |
9664 | If set to "on", Squid will append your client's IP address | |
9665 | in the HTTP requests it forwards. By default it looks like: | |
9666 | ||
9667 | X-Forwarded-For: 192.1.2.3 | |
9668 | ||
9669 | If set to "off", it will appear as | |
9670 | ||
9671 | X-Forwarded-For: unknown | |
9672 | ||
9673 | If set to "transparent", Squid will not alter the | |
9674 | X-Forwarded-For header in any way. | |
9675 | ||
9676 | If set to "delete", Squid will delete the entire | |
9677 | X-Forwarded-For header. | |
9678 | ||
9679 | If set to "truncate", Squid will remove all existing | |
9680 | X-Forwarded-For entries, and place the client IP as the sole entry. | |
9681 | DOC_END | |
9682 | ||
9683 | NAME: cachemgr_passwd | |
9684 | TYPE: cachemgrpasswd | |
9685 | DEFAULT: none | |
9686 | DEFAULT_DOC: No password. Actions which require password are denied. | |
9687 | LOC: Config.passwd_list | |
9688 | DOC_START | |
9689 | Specify passwords for cachemgr operations. | |
9690 | ||
9691 | Usage: cachemgr_passwd password action action ... | |
9692 | ||
9693 | Some valid actions are (see cache manager menu for a full list): | |
9694 | 5min | |
9695 | 60min | |
9696 | asndb | |
9697 | authenticator | |
9698 | cbdata | |
9699 | client_list | |
9700 | comm_incoming | |
9701 | config * | |
9702 | counters | |
9703 | delay | |
9704 | digest_stats | |
9705 | dns | |
9706 | events | |
9707 | filedescriptors | |
9708 | fqdncache | |
9709 | histograms | |
9710 | http_headers | |
9711 | info | |
9712 | io | |
9713 | ipcache | |
9714 | mem | |
9715 | menu | |
9716 | netdb | |
9717 | non_peers | |
9718 | objects | |
9719 | offline_toggle * | |
9720 | pconn | |
9721 | peer_select | |
9722 | reconfigure * | |
9723 | redirector | |
9724 | refresh | |
9725 | server_list | |
9726 | shutdown * | |
9727 | store_digest | |
9728 | storedir | |
9729 | utilization | |
9730 | via_headers | |
9731 | vm_objects | |
9732 | ||
9733 | * Indicates actions which will not be performed without a | |
9734 | valid password, others can be performed if not listed here. | |
9735 | ||
9736 | To disable an action, set the password to "disable". | |
9737 | To allow performing an action without a password, set the | |
9738 | password to "none". | |
9739 | ||
9740 | Use the keyword "all" to set the same password for all actions. | |
9741 | ||
9742 | Example: | |
9743 | cachemgr_passwd secret shutdown | |
9744 | cachemgr_passwd lesssssssecret info stats/objects | |
9745 | cachemgr_passwd disable all | |
9746 | DOC_END | |
9747 | ||
9748 | NAME: client_db | |
9749 | COMMENT: on|off | |
9750 | TYPE: onoff | |
9751 | DEFAULT: on | |
9752 | LOC: Config.onoff.client_db | |
9753 | DOC_START | |
9754 | If you want to disable collecting per-client statistics, | |
9755 | turn off client_db here. | |
9756 | DOC_END | |
9757 | ||
9758 | NAME: refresh_all_ims | |
9759 | COMMENT: on|off | |
9760 | TYPE: onoff | |
9761 | DEFAULT: off | |
9762 | LOC: Config.onoff.refresh_all_ims | |
9763 | DOC_START | |
9764 | When you enable this option, squid will always check | |
9765 | the origin server for an update when a client sends an | |
9766 | If-Modified-Since request. Many browsers use IMS | |
9767 | requests when the user requests a reload, and this | |
9768 | ensures those clients receive the latest version. | |
9769 | ||
9770 | By default (off), squid may return a Not Modified response | |
9771 | based on the age of the cached version. | |
9772 | DOC_END | |
9773 | ||
9774 | NAME: reload_into_ims | |
9775 | IFDEF: USE_HTTP_VIOLATIONS | |
9776 | COMMENT: on|off | |
9777 | TYPE: onoff | |
9778 | DEFAULT: off | |
9779 | LOC: Config.onoff.reload_into_ims | |
9780 | DOC_START | |
9781 | When you enable this option, client no-cache or ``reload'' | |
9782 | requests will be changed to If-Modified-Since requests. | |
9783 | Doing this VIOLATES the HTTP standard. Enabling this | |
9784 | feature could make you liable for problems which it | |
9785 | causes. | |
9786 | ||
9787 | see also refresh_pattern for a more selective approach. | |
9788 | DOC_END | |
9789 | ||
9790 | NAME: connect_retries | |
9791 | TYPE: int | |
9792 | LOC: Config.connect_retries | |
9793 | DEFAULT: 0 | |
9794 | DEFAULT_DOC: Do not retry failed connections. | |
9795 | DOC_START | |
9796 | This sets the maximum number of connection attempts made for each | |
9797 | TCP connection. The connect_retries attempts must all still | |
9798 | complete within the connection timeout period. | |
9799 | ||
9800 | The default is not to re-try if the first connection attempt fails. | |
9801 | The (not recommended) maximum is 10 tries. | |
9802 | ||
9803 | A warning message will be generated if it is set to a too-high | |
9804 | value and the configured value will be over-ridden. | |
9805 | ||
9806 | Note: These re-tries are in addition to forward_max_tries | |
9807 | which limit how many different addresses may be tried to find | |
9808 | a useful server. | |
9809 | DOC_END | |
9810 | ||
9811 | NAME: retry_on_error | |
9812 | TYPE: onoff | |
9813 | LOC: Config.retry.onerror | |
9814 | DEFAULT: off | |
9815 | DOC_START | |
9816 | If set to ON Squid will automatically retry requests when | |
9817 | receiving an error response with status 403 (Forbidden), | |
9818 | 500 (Internal Error), 501 or 503 (Service not available). | |
9819 | Status 502 and 504 (Gateway errors) are always retried. | |
9820 | ||
9821 | This is mainly useful if you are in a complex cache hierarchy to | |
9822 | work around access control errors. | |
9823 | ||
9824 | NOTE: This retry will attempt to find another working destination. | |
9825 | Which is different from the server which just failed. | |
9826 | DOC_END | |
9827 | ||
9828 | NAME: as_whois_server | |
9829 | TYPE: string | |
9830 | LOC: Config.as_whois_server | |
9831 | DEFAULT: whois.ra.net | |
9832 | DOC_START | |
9833 | WHOIS server to query for AS numbers. NOTE: AS numbers are | |
9834 | queried only when Squid starts up, not for every request. | |
9835 | DOC_END | |
9836 | ||
9837 | NAME: offline_mode | |
9838 | TYPE: onoff | |
9839 | LOC: Config.onoff.offline | |
9840 | DEFAULT: off | |
9841 | DOC_START | |
9842 | Enable this option and Squid will never try to validate cached | |
9843 | objects. | |
9844 | DOC_END | |
9845 | ||
9846 | NAME: uri_whitespace | |
9847 | TYPE: uri_whitespace | |
9848 | LOC: Config.uri_whitespace | |
9849 | DEFAULT: strip | |
9850 | DOC_START | |
9851 | What to do with requests that have whitespace characters in the | |
9852 | URI. Options: | |
9853 | ||
9854 | strip: The whitespace characters are stripped out of the URL. | |
9855 | This is the behavior recommended by RFC2396 and RFC3986 | |
9856 | for tolerant handling of generic URI. | |
9857 | NOTE: This is one difference between generic URI and HTTP URLs. | |
9858 | ||
9859 | deny: The request is denied. The user receives an "Invalid | |
9860 | Request" message. | |
9861 | This is the behaviour recommended by RFC2616 for safe | |
9862 | handling of HTTP request URL. | |
9863 | ||
9864 | allow: The request is allowed and the URI is not changed. The | |
9865 | whitespace characters remain in the URI. Note the | |
9866 | whitespace is passed to redirector processes if they | |
9867 | are in use. | |
9868 | Note this may be considered a violation of RFC2616 | |
9869 | request parsing where whitespace is prohibited in the | |
9870 | URL field. | |
9871 | ||
9872 | encode: The request is allowed and the whitespace characters are | |
9873 | encoded according to RFC1738. | |
9874 | ||
9875 | chop: The request is allowed and the URI is chopped at the | |
9876 | first whitespace. | |
9877 | ||
9878 | ||
9879 | NOTE the current Squid implementation of encode and chop violates | |
9880 | RFC2616 by not using a 301 redirect after altering the URL. | |
9881 | DOC_END | |
9882 | ||
9883 | NAME: chroot | |
9884 | TYPE: string | |
9885 | LOC: Config.chroot_dir | |
9886 | DEFAULT: none | |
9887 | DOC_START | |
9888 | Specifies a directory where Squid should do a chroot() while | |
9889 | initializing. This also causes Squid to fully drop root | |
9890 | privileges after initializing. This means, for example, if you | |
9891 | use a HTTP port less than 1024 and try to reconfigure, you may | |
9892 | get an error saying that Squid can not open the port. | |
9893 | DOC_END | |
9894 | ||
9895 | NAME: balance_on_multiple_ip | |
9896 | TYPE: onoff | |
9897 | LOC: Config.onoff.balance_on_multiple_ip | |
9898 | DEFAULT: off | |
9899 | DOC_START | |
9900 | Modern IP resolvers in squid sort lookup results by preferred access. | |
9901 | By default squid will use these IP in order and only rotates to | |
9902 | the next listed when the most preffered fails. | |
9903 | ||
9904 | Some load balancing servers based on round robin DNS have been | |
9905 | found not to preserve user session state across requests | |
9906 | to different IP addresses. | |
9907 | ||
9908 | Enabling this directive Squid rotates IP's per request. | |
9909 | DOC_END | |
9910 | ||
9911 | NAME: pipeline_prefetch | |
9912 | TYPE: pipelinePrefetch | |
9913 | LOC: Config.pipeline_max_prefetch | |
9914 | DEFAULT: 0 | |
9915 | DEFAULT_DOC: Do not pre-parse pipelined requests. | |
9916 | DOC_START | |
9917 | HTTP clients may send a pipeline of 1+N requests to Squid using a | |
9918 | single connection, without waiting for Squid to respond to the first | |
9919 | of those requests. This option limits the number of concurrent | |
9920 | requests Squid will try to handle in parallel. If set to N, Squid | |
9921 | will try to receive and process up to 1+N requests on the same | |
9922 | connection concurrently. | |
9923 | ||
9924 | Defaults to 0 (off) for bandwidth management and access logging | |
9925 | reasons. | |
9926 | ||
9927 | NOTE: pipelining requires persistent connections to clients. | |
9928 | ||
9929 | WARNING: pipelining breaks NTLM and Negotiate/Kerberos authentication. | |
9930 | DOC_END | |
9931 | ||
9932 | NAME: high_response_time_warning | |
9933 | TYPE: int | |
9934 | COMMENT: (msec) | |
9935 | LOC: Config.warnings.high_rptm | |
9936 | DEFAULT: 0 | |
9937 | DEFAULT_DOC: disabled. | |
9938 | DOC_START | |
9939 | If the one-minute median response time exceeds this value, | |
9940 | Squid prints a WARNING with debug level 0 to get the | |
9941 | administrators attention. The value is in milliseconds. | |
9942 | DOC_END | |
9943 | ||
9944 | NAME: high_page_fault_warning | |
9945 | TYPE: int | |
9946 | LOC: Config.warnings.high_pf | |
9947 | DEFAULT: 0 | |
9948 | DEFAULT_DOC: disabled. | |
9949 | DOC_START | |
9950 | If the one-minute average page fault rate exceeds this | |
9951 | value, Squid prints a WARNING with debug level 0 to get | |
9952 | the administrators attention. The value is in page faults | |
9953 | per second. | |
9954 | DOC_END | |
9955 | ||
9956 | NAME: high_memory_warning | |
9957 | TYPE: b_size_t | |
9958 | LOC: Config.warnings.high_memory | |
9959 | IFDEF: HAVE_MSTATS&&HAVE_GNUMALLOC_H | |
9960 | DEFAULT: 0 KB | |
9961 | DEFAULT_DOC: disabled. | |
9962 | DOC_START | |
9963 | If the memory usage (as determined by gnumalloc, if available and used) | |
9964 | exceeds this amount, Squid prints a WARNING with debug level 0 to get | |
9965 | the administrators attention. | |
9966 | DOC_END | |
9967 | # TODO: link high_memory_warning to mempools? | |
9968 | ||
9969 | NAME: sleep_after_fork | |
9970 | COMMENT: (microseconds) | |
9971 | TYPE: int | |
9972 | LOC: Config.sleep_after_fork | |
9973 | DEFAULT: 0 | |
9974 | DOC_START | |
9975 | When this is set to a non-zero value, the main Squid process | |
9976 | sleeps the specified number of microseconds after a fork() | |
9977 | system call. This sleep may help the situation where your | |
9978 | system reports fork() failures due to lack of (virtual) | |
9979 | memory. Note, however, if you have a lot of child | |
9980 | processes, these sleep delays will add up and your | |
9981 | Squid will not service requests for some amount of time | |
9982 | until all the child processes have been started. | |
9983 | On Windows value less then 1000 (1 milliseconds) are | |
9984 | rounded to 1000. | |
9985 | DOC_END | |
9986 | ||
9987 | NAME: windows_ipaddrchangemonitor | |
9988 | IFDEF: _SQUID_WINDOWS_ | |
9989 | COMMENT: on|off | |
9990 | TYPE: onoff | |
9991 | DEFAULT: on | |
9992 | LOC: Config.onoff.WIN32_IpAddrChangeMonitor | |
9993 | DOC_START | |
9994 | On Windows Squid by default will monitor IP address changes and will | |
9995 | reconfigure itself after any detected event. This is very useful for | |
9996 | proxies connected to internet with dial-up interfaces. | |
9997 | In some cases (a Proxy server acting as VPN gateway is one) it could be | |
9998 | desiderable to disable this behaviour setting this to 'off'. | |
9999 | Note: after changing this, Squid service must be restarted. | |
10000 | DOC_END | |
10001 | ||
10002 | NAME: eui_lookup | |
10003 | TYPE: onoff | |
10004 | IFDEF: USE_SQUID_EUI | |
10005 | DEFAULT: on | |
10006 | LOC: Eui::TheConfig.euiLookup | |
10007 | DOC_START | |
10008 | Whether to lookup the EUI or MAC address of a connected client. | |
10009 | DOC_END | |
10010 | ||
10011 | NAME: max_filedescriptors max_filedesc | |
10012 | TYPE: int | |
10013 | DEFAULT: 0 | |
10014 | DEFAULT_DOC: Use operating system limits set by ulimit. | |
10015 | LOC: Config.max_filedescriptors | |
10016 | DOC_START | |
10017 | Reduce the maximum number of filedescriptors supported below | |
10018 | the usual operating system defaults. | |
10019 | ||
10020 | Remove from squid.conf to inherit the current ulimit setting. | |
10021 | ||
10022 | Note: Changing this requires a restart of Squid. Also | |
10023 | not all I/O types supports large values (eg on Windows). | |
10024 | DOC_END | |
10025 | ||
10026 | NAME: force_request_body_continuation | |
10027 | TYPE: acl_access | |
10028 | LOC: Config.accessList.forceRequestBodyContinuation | |
10029 | DEFAULT: none | |
10030 | DEFAULT_DOC: Deny, unless rules exist in squid.conf. | |
10031 | DOC_START | |
10032 | This option controls how Squid handles data upload requests from HTTP | |
10033 | and FTP agents that require a "Please Continue" control message response | |
10034 | to actually send the request body to Squid. It is mostly useful in | |
10035 | adaptation environments. | |
10036 | ||
10037 | When Squid receives an HTTP request with an "Expect: 100-continue" | |
10038 | header or an FTP upload command (e.g., STOR), Squid normally sends the | |
10039 | request headers or FTP command information to an adaptation service (or | |
10040 | peer) and waits for a response. Most adaptation services (and some | |
10041 | broken peers) may not respond to Squid at that stage because they may | |
10042 | decide to wait for the HTTP request body or FTP data transfer. However, | |
10043 | that request body or data transfer may never come because Squid has not | |
10044 | responded with the HTTP 100 or FTP 150 (Please Continue) control message | |
10045 | to the request sender yet! | |
10046 | ||
10047 | An allow match tells Squid to respond with the HTTP 100 or FTP 150 | |
10048 | (Please Continue) control message on its own, before forwarding the | |
10049 | request to an adaptation service or peer. Such a response usually forces | |
10050 | the request sender to proceed with sending the body. A deny match tells | |
10051 | Squid to delay that control response until the origin server confirms | |
10052 | that the request body is needed. Delaying is the default behavior. | |
10053 | DOC_END | |
10054 | ||
10055 | NAME: server_pconn_for_nonretriable | |
10056 | TYPE: acl_access | |
10057 | DEFAULT: none | |
10058 | DEFAULT_DOC: Open new connections for forwarding requests Squid cannot retry safely. | |
10059 | LOC: Config.accessList.serverPconnForNonretriable | |
10060 | DOC_START | |
10061 | This option provides fine-grained control over persistent connection | |
10062 | reuse when forwarding HTTP requests that Squid cannot retry. It is useful | |
10063 | in environments where opening new connections is very expensive | |
10064 | (e.g., all connections are secured with TLS with complex client and server | |
10065 | certificate validation) and race conditions associated with persistent | |
10066 | connections are very rare and/or only cause minor problems. | |
10067 | ||
10068 | HTTP prohibits retrying unsafe and non-idempotent requests (e.g., POST). | |
10069 | Squid limitations also prohibit retrying all requests with bodies (e.g., PUT). | |
10070 | By default, when forwarding such "risky" requests, Squid opens a new | |
10071 | connection to the server or cache_peer, even if there is an idle persistent | |
10072 | connection available. When Squid is configured to risk sending a non-retriable | |
10073 | request on a previously used persistent connection, and the server closes | |
10074 | the connection before seeing that risky request, the user gets an error response | |
10075 | from Squid. In most cases, that error response will be HTTP 502 (Bad Gateway) | |
10076 | with ERR_ZERO_SIZE_OBJECT or ERR_WRITE_ERROR (peer connection reset) error detail. | |
10077 | ||
10078 | If an allow rule matches, Squid reuses an available idle persistent connection | |
10079 | (if any) for the request that Squid cannot retry. If a deny rule matches, then | |
10080 | Squid opens a new connection for the request that Squid cannot retry. | |
10081 | ||
10082 | This option does not affect requests that Squid can retry. They will reuse idle | |
10083 | persistent connections (if any). | |
10084 | ||
10085 | This clause only supports fast acl types. | |
10086 | See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details. | |
10087 | ||
10088 | Example: | |
10089 | acl SpeedIsWorthTheRisk method POST | |
10090 | server_pconn_for_nonretriable allow SpeedIsWorthTheRisk | |
10091 | DOC_END | |
10092 | ||
10093 | EOF |