2 # SQUID Web Proxy Cache http://www.squid-cache.org/
3 # ----------------------------------------------------------
5 # Squid is the result of efforts by numerous individuals from
6 # the Internet community; see the CONTRIBUTORS file for full
7 # details. Many organizations have provided support for Squid's
8 # development; see the SPONSORS file for full details. Squid is
9 # Copyrighted (C) 2000 by the Regents of the University of
10 # California; see the COPYRIGHT file for full details. Squid
11 # incorporates software developed and/or copyrighted by other
12 # sources; see the CREDITS file for full details.
14 # This program is free software; you can redistribute it and/or modify
15 # it under the terms of the GNU General Public License as published by
16 # the Free Software Foundation; either version 2 of the License, or
17 # (at your option) any later version.
19 # This program is distributed in the hope that it will be useful,
20 # but WITHOUT ANY WARRANTY; without even the implied warranty of
21 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
22 # GNU General Public License for more details.
24 # You should have received a copy of the GNU General Public License
25 # along with this program; if not, write to the Free Software
26 # Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111, USA.
31 ----------------------------
33 This is the default Squid configuration file. You may wish
34 to look at the Squid home page (http://www.squid-cache.org/)
35 for the FAQ and other documentation.
37 The default Squid config file shows what the defaults for
38 various options happen to be. If you don't need to change the
39 default, you shouldn't uncomment the line. Doing so may cause
40 run-time problems. In some cases "none" refers to no default
41 setting at all, while in other cases it refers to a valid
42 option - the comments for that keyword indicate if this is the
48 Configuration options can be included using the "include" directive.
49 Include takes a list of files to include. Quoting and wildcards is
54 include /path/to/included/file/squid.acl.config
56 Includes can be nested up to a hard-coded depth of 16 levels.
57 This arbitrary restriction is to prevent recursive include references
58 from causing Squid entering an infinite loop whilst trying to load
62 Conditional configuration
64 If-statements can be used to make configuration directives
68 ... regular configuration directives ...
70 ... regular configuration directives ...]
73 The else part is optional. The keywords "if", "else", and "endif"
74 must be typed on their own lines, as if they were regular
75 configuration directives.
77 These individual conditions types are supported:
80 Always evaluates to true.
82 Always evaluates to false.
84 Equality comparison of two integer numbers.
89 The following SMP-related preprocessor macros can be used.
91 ${process_name} expands to the current Squid process "name"
92 (e.g., squid1, squid2, or cache1).
94 ${process_number} expands to the current Squid process
95 identifier, which is an integer number (e.g., 1, 2, 3) unique
96 across all Squid processes.
99 # Options Removed in 3.2
100 NAME: ignore_expect_100
103 Remove this line. The HTTP/1.1 feature is now fully supported by default.
109 Remove this line. Configure FTP page display using the CSS controls in errorpages.css instead.
112 NAME: url_rewrite_concurrency
115 Remove this line. Set the 'concurrency=' option of url_rewrite_children instead.
118 # Options Removed in 3.1
122 Remove this line. DNS is no longer tested on startup.
125 NAME: extension_methods
128 Remove this line. All valid methods for HTTP are accepted by default.
131 # 2.7 Options Removed/Replaced in 3.1
139 Remove this line. HTTP/1.1 is supported by default.
142 NAME: upgrade_http0.9
145 Remove this line. ICY/1.0 streaming protocol is supported by default.
148 NAME: zph_local zph_mode zph_option zph_parent zph_sibling
151 Alter these entries. Use the qos_flows directive instead.
154 # Options Removed in 3.0
158 Since squid-3.0 replace with request_header_access or reply_header_access
159 depending on whether you wish to match client requests or server replies.
162 NAME: httpd_accel_no_pmtu_disc
165 Since squid-3.0 use the 'disable-pmtu-discovery' flag on http_port instead.
169 OPTIONS FOR AUTHENTICATION
170 -----------------------------------------------------------------------------
179 This is used to define parameters for the various authentication
180 schemes supported by Squid.
182 format: auth_param scheme parameter [setting]
184 The order in which authentication schemes are presented to the client is
185 dependent on the order the scheme first appears in config file. IE
186 has a bug (it's not RFC 2617 compliant) in that it will use the basic
187 scheme if basic is the first entry presented, even if more secure
188 schemes are presented. For now use the order in the recommended
189 settings section below. If other browsers have difficulties (don't
190 recognize the schemes offered even if you are using basic) either
191 put basic first, or disable the other schemes (by commenting out their
194 Once an authentication scheme is fully configured, it can only be
195 shutdown by shutting squid down and restarting. Changes can be made on
196 the fly and activated with a reconfigure. I.E. You can change to a
197 different helper, but not unconfigure the helper completely.
199 Please note that while this directive defines how Squid processes
200 authentication it does not automatically activate authentication.
201 To use authentication you must in addition make use of ACLs based
202 on login name in http_access (proxy_auth, proxy_auth_regex or
203 external with %LOGIN used in the format tag). The browser will be
204 challenged for authentication on the first such acl encountered
205 in http_access processing and will also be re-challenged for new
206 login credentials if the request is being denied by a proxy_auth
209 WARNING: authentication can't be used in a transparently intercepting
210 proxy as the client then thinks it is talking to an origin server and
211 not the proxy. This is a limitation of bending the TCP/IP protocol to
212 transparently intercepting port 80, not a limitation in Squid.
213 Ports flagged 'transparent', 'intercept', or 'tproxy' have
214 authentication disabled.
216 === Parameters for the basic scheme follow. ===
219 Specify the command for the external authenticator. Such a program
220 reads a line containing "username password" and replies "OK" or
221 "ERR" in an endless loop. "ERR" responses may optionally be followed
222 by a error description available as %m in the returned error page.
223 If you use an authenticator, make sure you have 1 acl of type
226 By default, the basic authentication scheme is not used unless a
227 program is specified.
229 If you want to use the traditional NCSA proxy authentication, set
230 this line to something like
232 auth_param basic program @DEFAULT_PREFIX@/libexec/ncsa_auth @DEFAULT_PREFIX@/etc/passwd
235 HTTP uses iso-latin-1 as characterset, while some authentication
236 backends such as LDAP expects UTF-8. If this is set to on Squid will
237 translate the HTTP iso-latin-1 charset to UTF-8 before sending the
238 username & password to the helper.
240 "children" numberofchildren [startup=N] [idle=N] [concurrency=N]
241 The maximum number of authenticator processes to spawn. If you start too few
242 Squid will have to wait for them to process a backlog of credential
243 verifications, slowing it down. When password verifications are
244 done via a (slow) network you are likely to need lots of
245 authenticator processes.
247 The startup= and idle= options permit some skew in the exact amount
248 run. A minimum of startup=N will begin during startup and reconfigure.
249 Squid will start more in groups of up to idle=N in an attempt to meet
250 traffic needs and to keep idle=N free above those traffic needs up to
253 The concurrency= option sets the number of concurrent requests the
254 helper can process. The default of 0 is used for helpers who only
255 supports one request at a time. Setting this to a number greater than
256 0 changes the protocol used to include a channel number first on the
257 request/response line, allowing multiple requests to be sent to the
258 same helper in parallell without wating for the response.
259 Must not be set unless it's known the helper supports this.
261 auth_param basic children 20 startup=0 idle=1
264 Specifies the realm name which is to be reported to the
265 client for the basic proxy authentication scheme (part of
266 the text the user will see when prompted their username and
267 password). There is no default.
268 auth_param basic realm Squid proxy-caching web server
270 "credentialsttl" timetolive
271 Specifies how long squid assumes an externally validated
272 username:password pair is valid for - in other words how
273 often the helper program is called for that user. Set this
274 low to force revalidation with short lived passwords. Note
275 setting this high does not impact your susceptibility
276 to replay attacks unless you are using an one-time password
277 system (such as SecureID). If you are using such a system,
278 you will be vulnerable to replay attacks unless you also
279 use the max_user_ip ACL in an http_access rule.
281 "casesensitive" on|off
282 Specifies if usernames are case sensitive. Most user databases are
283 case insensitive allowing the same username to be spelled using both
284 lower and upper case letters, but some are case sensitive. This
285 makes a big difference for user_max_ip ACL processing and similar.
286 auth_param basic casesensitive off
288 === Parameters for the digest scheme follow ===
291 Specify the command for the external authenticator. Such
292 a program reads a line containing "username":"realm" and
293 replies with the appropriate H(A1) value hex encoded or
294 ERR if the user (or his H(A1) hash) does not exists.
295 See rfc 2616 for the definition of H(A1).
296 "ERR" responses may optionally be followed by a error description
297 available as %m in the returned error page.
299 By default, the digest authentication scheme is not used unless a
300 program is specified.
302 If you want to use a digest authenticator, set this line to
305 auth_param digest program @DEFAULT_PREFIX@/bin/digest_pw_auth @DEFAULT_PREFIX@/etc/digpass
308 HTTP uses iso-latin-1 as characterset, while some authentication
309 backends such as LDAP expects UTF-8. If this is set to on Squid will
310 translate the HTTP iso-latin-1 charset to UTF-8 before sending the
311 username & password to the helper.
313 "children" numberofchildren [startup=N] [idle=N] [concurrency=N]
314 The maximum number of authenticator processes to spawn (default 5).
315 If you start too few Squid will have to wait for them to
316 process a backlog of H(A1) calculations, slowing it down.
317 When the H(A1) calculations are done via a (slow) network
318 you are likely to need lots of authenticator processes.
320 The startup= and idle= options permit some skew in the exact amount
321 run. A minimum of startup=N will begin during startup and reconfigure.
322 Squid will start more in groups of up to idle=N in an attempt to meet
323 traffic needs and to keep idle=N free above those traffic needs up to
326 The concurrency= option sets the number of concurrent requests the
327 helper can process. The default of 0 is used for helpers who only
328 supports one request at a time. Setting this to a number greater than
329 0 changes the protocol used to include a channel number first on the
330 request/response line, allowing multiple requests to be sent to the
331 same helper in parallell without wating for the response.
332 Must not be set unless it's known the helper supports this.
334 auth_param digest children 20 startup=0 idle=1
337 Specifies the realm name which is to be reported to the
338 client for the digest proxy authentication scheme (part of
339 the text the user will see when prompted their username and
340 password). There is no default.
341 auth_param digest realm Squid proxy-caching web server
343 "nonce_garbage_interval" timeinterval
344 Specifies the interval that nonces that have been issued
345 to client_agent's are checked for validity.
347 "nonce_max_duration" timeinterval
348 Specifies the maximum length of time a given nonce will be
351 "nonce_max_count" number
352 Specifies the maximum number of times a given nonce can be
355 "nonce_strictness" on|off
356 Determines if squid requires strict increment-by-1 behavior
357 for nonce counts, or just incrementing (off - for use when
358 useragents generate nonce counts that occasionally miss 1
359 (ie, 1,2,4,6)). Default off.
361 "check_nonce_count" on|off
362 This directive if set to off can disable the nonce count check
363 completely to work around buggy digest qop implementations in
364 certain mainstream browser versions. Default on to check the
365 nonce count to protect from authentication replay attacks.
367 "post_workaround" on|off
368 This is a workaround to certain buggy browsers who sends
369 an incorrect request digest in POST requests when reusing
370 the same nonce as acquired earlier on a GET request.
372 === NTLM scheme options follow ===
375 Specify the command for the external NTLM authenticator.
376 Such a program reads exchanged NTLMSSP packets with
377 the browser via Squid until authentication is completed.
378 If you use an NTLM authenticator, make sure you have 1 acl
379 of type proxy_auth. By default, the NTLM authenticator_program
382 auth_param ntlm program @DEFAULT_PREFIX@/bin/ntlm_auth
384 "children" numberofchildren [startup=N] [idle=N]
385 The maximum number of authenticator processes to spawn (default 5).
386 If you start too few Squid will have to wait for them to
387 process a backlog of credential verifications, slowing it
388 down. When credential verifications are done via a (slow)
389 network you are likely to need lots of authenticator
392 The startup= and idle= options permit some skew in the exact amount
393 run. A minimum of startup=N will begin during startup and reconfigure.
394 Squid will start more in groups of up to idle=N in an attempt to meet
395 traffic needs and to keep idle=N free above those traffic needs up to
398 auth_param ntlm children 20 startup=0 idle=1
401 If you experience problems with PUT/POST requests when using the
402 Negotiate authentication scheme then you can try setting this to
403 off. This will cause Squid to forcibly close the connection on
404 the initial requests where the browser asks which schemes are
405 supported by the proxy.
407 auth_param ntlm keep_alive on
409 === Options for configuring the NEGOTIATE auth-scheme follow ===
412 Specify the command for the external Negotiate authenticator.
413 This protocol is used in Microsoft Active-Directory enabled setups with
414 the Microsoft Internet Explorer or Mozilla Firefox browsers.
415 Its main purpose is to exchange credentials with the Squid proxy
416 using the Kerberos mechanisms.
417 If you use a Negotiate authenticator, make sure you have at least
418 one acl of type proxy_auth active. By default, the negotiate
419 authenticator_program is not used.
420 The only supported program for this role is the ntlm_auth
421 program distributed as part of Samba, version 4 or later.
423 auth_param negotiate program @DEFAULT_PREFIX@/bin/ntlm_auth --helper-protocol=gss-spnego
425 "children" numberofchildren [startup=N] [idle=N]
426 The maximum number of authenticator processes to spawn (default 5).
427 If you start too few Squid will have to wait for them to
428 process a backlog of credential verifications, slowing it
429 down. When crendential verifications are done via a (slow)
430 network you are likely to need lots of authenticator
433 The startup= and idle= options permit some skew in the exact amount
434 run. A minimum of startup=N will begin during startup and reconfigure.
435 Squid will start more in groups of up to idle=N in an attempt to meet
436 traffic needs and to keep idle=N free above those traffic needs up to
439 auth_param negotiate children 20 startup=0 idle=1
442 If you experience problems with PUT/POST requests when using the
443 Negotiate authentication scheme then you can try setting this to
444 off. This will cause Squid to forcibly close the connection on
445 the initial requests where the browser asks which schemes are
446 supported by the proxy.
448 auth_param negotiate keep_alive on
453 #Recommended minimum configuration per scheme:
454 #auth_param negotiate program <uncomment and complete this line to activate>
455 #auth_param negotiate children 20 startup=0 idle=1
456 #auth_param negotiate keep_alive on
458 #auth_param ntlm program <uncomment and complete this line to activate>
459 #auth_param ntlm children 20 startup=0 idle=1
460 #auth_param ntlm keep_alive on
462 #auth_param digest program <uncomment and complete this line>
463 #auth_param digest children 20 startup=0 idle=1
464 #auth_param digest realm Squid proxy-caching web server
465 #auth_param digest nonce_garbage_interval 5 minutes
466 #auth_param digest nonce_max_duration 30 minutes
467 #auth_param digest nonce_max_count 50
469 #auth_param basic program <uncomment and complete this line>
470 #auth_param basic children 5 startup=5 idle=1
471 #auth_param basic realm Squid proxy-caching web server
472 #auth_param basic credentialsttl 2 hours
475 NAME: authenticate_cache_garbage_interval
478 LOC: Config.authenticateGCInterval
480 The time period between garbage collection across the username cache.
481 This is a tradeoff between memory utilization (long intervals - say
482 2 days) and CPU (short intervals - say 1 minute). Only change if you
486 NAME: authenticate_ttl
489 LOC: Config.authenticateTTL
491 The time a user & their credentials stay in the logged in
492 user cache since their last request. When the garbage
493 interval passes, all user credentials that have passed their
494 TTL are removed from memory.
497 NAME: authenticate_ip_ttl
499 LOC: Config.authenticateIpTTL
502 If you use proxy authentication and the 'max_user_ip' ACL,
503 this directive controls how long Squid remembers the IP
504 addresses associated with each user. Use a small value
505 (e.g., 60 seconds) if your users might change addresses
506 quickly, as is the case with dialups. You might be safe
507 using a larger value (e.g., 2 hours) in a corporate LAN
508 environment with relatively static address assignments.
513 -----------------------------------------------------------------------------
516 NAME: external_acl_type
517 TYPE: externalAclHelper
518 LOC: Config.externalAclHelperList
521 This option defines external acl classes using a helper program
522 to look up the status
524 external_acl_type name [options] FORMAT.. /path/to/helper [helper arguments..]
528 ttl=n TTL in seconds for cached results (defaults to 3600
531 TTL for cached negative lookups (default same
534 Maximum number of acl helper processes spawned to service
535 external acl lookups of this type. (default 20)
537 Minimum number of acl helper processes to spawn during
538 startup and reconfigure to service external acl lookups
539 of this type. (default 0)
541 Number of acl helper processes to keep ahead of traffic
542 loads. Squid will spawn this many at once whenever load
543 rises above the capabilities of existing processes.
544 Up to the value of children-max. (default 1)
545 concurrency=n concurrency level per process. Only used with helpers
546 capable of processing more than one query at a time.
547 cache=n limit the result cache size, default is unbounded.
548 grace=n Percentage remaining of TTL where a refresh of a
549 cached entry should be initiated without needing to
550 wait for a new reply. (default is for no grace period)
551 protocol=2.5 Compatibility mode for Squid-2.5 external acl helpers
552 ipv4 / ipv6 IP-mode used to communicate to this helper.
553 For compatability with older configurations and helpers
554 the default is currently 'ipv4'.
556 FORMAT specifications
558 %LOGIN Authenticated user login name
559 %EXT_USER Username from previous external acl
560 %EXT_LOG Log details from previous external acl
561 %EXT_TAG Tag from previous external acl
562 %IDENT Ident user name
564 %SRCPORT Client source port
567 %PROTO Requested protocol
569 %PATH Requested URL path
570 %METHOD Request method
571 %MYADDR Squid interface address
572 %MYPORT Squid http_port number
573 %PATH Requested URL-path (including query-string if any)
574 %USER_CERT SSL User certificate in PEM format
575 %USER_CERTCHAIN SSL User certificate chain in PEM format
576 %USER_CERT_xx SSL User certificate subject attribute xx
577 %USER_CA_xx SSL User certificate issuer attribute xx
579 %>{Header} HTTP request header "Header"
581 HTTP request header "Hdr" list member "member"
583 HTTP request header list member using ; as
584 list separator. ; can be any non-alphanumeric
587 %<{Header} HTTP reply header "Header"
589 HTTP reply header "Hdr" list member "member"
591 HTTP reply header list member using ; as
592 list separator. ; can be any non-alphanumeric
595 In addition to the above, any string specified in the referencing
596 acl will also be included in the helper request line, after the
597 specified formats (see the "acl external" directive)
599 The helper receives lines per the above format specification,
600 and returns lines starting with OK or ERR indicating the validity
601 of the request and optionally followed by additional keywords with
604 General result syntax:
606 OK/ERR keyword=value ...
610 user= The users name (login)
611 password= The users password (for login= cache_peer option)
612 message= Message describing the reason. Available as %o
614 tag= Apply a tag to a request (for both ERR and OK results)
615 Only sets a tag, does not alter existing tags.
616 log= String to be logged in access.log. Available as
617 %ea in logformat specifications
619 If protocol=3.0 (the default) then URL escaping is used to protect
620 each value in both requests and responses.
622 If using protocol=2.5 then all values need to be enclosed in quotes
623 if they may contain whitespace, or the whitespace escaped using \.
624 And quotes or \ characters within the keyword value must be \ escaped.
626 When using the concurrency= option the protocol is changed by
627 introducing a query channel tag infront of the request/response.
628 The query channel tag is a number between 0 and concurrency-1.
635 DEFAULT: manager url_regex -i ^cache_object:// +i ^https?://[^/]+/squid-internal-mgr/
636 DEFAULT: localhost src 127.0.0.1/32 ::1
637 DEFAULT: to_localhost dst 127.0.0.0/8 0.0.0.0/32 ::1
638 DEFAULT_DOC: ACLs all, manager, localhost, and to_localhost are predefined.
640 Defining an Access List
642 Every access list definition must begin with an aclname and acltype,
643 followed by either type-specific arguments or a quoted filename that
646 acl aclname acltype argument ...
647 acl aclname acltype "file" ...
649 When using "file", the file should contain one item per line.
651 By default, regular expressions are CASE-SENSITIVE.
652 To make them case-insensitive, use the -i option. To return case-sensitive
653 use the +i option between patterns, or make a new ACL line without -i.
655 Some acl types require suspending the current request in order
656 to access some external data source.
657 Those which do are marked with the tag [slow], those which
658 don't are marked as [fast].
659 See http://wiki.squid-cache.org/SquidFaq/SquidAcl
660 for further information
662 ***** ACL TYPES AVAILABLE *****
664 acl aclname src ip-address/mask ... # clients IP address [fast]
665 acl aclname src addr1-addr2/mask ... # range of addresses [fast]
666 acl aclname dst ip-address/mask ... # URL host's IP address [slow]
667 acl aclname localip ip-address/mask ... # IP address the client connected to [fast]
669 acl aclname arp mac-address ... (xx:xx:xx:xx:xx:xx notation)
670 # The arp ACL requires the special configure option --enable-arp-acl.
671 # Furthermore, the ARP ACL code is not portable to all operating systems.
672 # It works on Linux, Solaris, Windows, FreeBSD, and some
673 # other *BSD variants.
676 # NOTE: Squid can only determine the MAC address for clients that are on
677 # the same subnet. If the client is on a different subnet,
678 # then Squid cannot find out its MAC address.
680 acl aclname srcdomain .foo.com ...
681 # reverse lookup, from client IP [slow]
682 acl aclname dstdomain .foo.com ...
683 # Destination server from URL [fast]
684 acl aclname srcdom_regex [-i] \.foo\.com ...
685 # regex matching client name [slow]
686 acl aclname dstdom_regex [-i] \.foo\.com ...
687 # regex matching server [fast]
689 # For dstdomain and dstdom_regex a reverse lookup is tried if a IP
690 # based URL is used and no match is found. The name "none" is used
691 # if the reverse lookup fails.
693 acl aclname src_as number ...
694 acl aclname dst_as number ...
696 # Except for access control, AS numbers can be used for
697 # routing of requests to specific caches. Here's an
698 # example for routing all requests for AS#1241 and only
699 # those to mycache.mydomain.net:
700 # acl asexample dst_as 1241
701 # cache_peer_access mycache.mydomain.net allow asexample
702 # cache_peer_access mycache_mydomain.net deny all
704 acl aclname peername myPeer ...
706 # match against a named cache_peer entry
707 # set unique name= on cache_peer lines for reliable use.
709 acl aclname time [day-abbrevs] [h1:m1-h2:m2]
719 # h1:m1 must be less than h2:m2
721 acl aclname url_regex [-i] ^http:// ...
722 # regex matching on whole URL [fast]
723 acl aclname urlpath_regex [-i] \.gif$ ...
724 # regex matching on URL path [fast]
726 acl aclname port 80 70 21 0-1024... # destination TCP port [fast]
728 acl aclname localport 3128 ... # TCP port the client connected to [fast]
729 # NP: for interception mode this is usually '80'
731 acl aclname myportname 3128 ... # http(s)_port name [fast]
733 acl aclname proto HTTP FTP ... # request protocol [fast]
735 acl aclname method GET POST ... # HTTP request method [fast]
737 acl aclname http_status 200 301 500- 400-403 ...
738 # status code in reply [fast]
740 acl aclname browser [-i] regexp ...
741 # pattern match on User-Agent header (see also req_header below) [fast]
743 acl aclname referer_regex [-i] regexp ...
744 # pattern match on Referer header [fast]
745 # Referer is highly unreliable, so use with care
747 acl aclname ident username ...
748 acl aclname ident_regex [-i] pattern ...
749 # string match on ident output [slow]
750 # use REQUIRED to accept any non-null ident.
752 acl aclname proxy_auth [-i] username ...
753 acl aclname proxy_auth_regex [-i] pattern ...
754 # perform http authentication challenge to the client and match against
755 # supplied credentials [slow]
757 # takes a list of allowed usernames.
758 # use REQUIRED to accept any valid username.
760 # Will use proxy authentication in forward-proxy scenarios, and plain
761 # http authenticaiton in reverse-proxy scenarios
763 # NOTE: when a Proxy-Authentication header is sent but it is not
764 # needed during ACL checking the username is NOT logged
767 # NOTE: proxy_auth requires a EXTERNAL authentication program
768 # to check username/password combinations (see
769 # auth_param directive).
771 # NOTE: proxy_auth can't be used in a transparent/intercepting proxy
772 # as the browser needs to be configured for using a proxy in order
773 # to respond to proxy authentication.
775 acl aclname snmp_community string ...
776 # A community string to limit access to your SNMP Agent [fast]
779 # acl snmppublic snmp_community public
781 acl aclname maxconn number
782 # This will be matched when the client's IP address has
783 # more than <number> TCP connections established. [fast]
784 # NOTE: This only measures direct TCP links so X-Forwarded-For
785 # indirect clients are not counted.
787 acl aclname max_user_ip [-s] number
788 # This will be matched when the user attempts to log in from more
789 # than <number> different ip addresses. The authenticate_ip_ttl
790 # parameter controls the timeout on the ip entries. [fast]
791 # If -s is specified the limit is strict, denying browsing
792 # from any further IP addresses until the ttl has expired. Without
793 # -s Squid will just annoy the user by "randomly" denying requests.
794 # (the counter is reset each time the limit is reached and a
796 # NOTE: in acceleration mode or where there is mesh of child proxies,
797 # clients may appear to come from multiple addresses if they are
798 # going through proxy farms, so a limit of 1 may cause user problems.
800 acl aclname random probability
801 # Pseudo-randomly match requests. Based on the probability given.
802 # Probability may be written as a decimal (0.333), fraction (1/3)
803 # or ratio of matches:non-matches (3:5).
805 acl aclname req_mime_type [-i] mime-type ...
806 # regex match against the mime type of the request generated
807 # by the client. Can be used to detect file upload or some
808 # types HTTP tunneling requests [fast]
809 # NOTE: This does NOT match the reply. You cannot use this
810 # to match the returned file type.
812 acl aclname req_header header-name [-i] any\.regex\.here
813 # regex match against any of the known request headers. May be
814 # thought of as a superset of "browser", "referer" and "mime-type"
817 acl aclname rep_mime_type [-i] mime-type ...
818 # regex match against the mime type of the reply received by
819 # squid. Can be used to detect file download or some
820 # types HTTP tunneling requests. [fast]
821 # NOTE: This has no effect in http_access rules. It only has
822 # effect in rules that affect the reply data stream such as
825 acl aclname rep_header header-name [-i] any\.regex\.here
826 # regex match against any of the known reply headers. May be
827 # thought of as a superset of "browser", "referer" and "mime-type"
830 acl aclname external class_name [arguments...]
831 # external ACL lookup via a helper class defined by the
832 # external_acl_type directive [slow]
834 acl aclname user_cert attribute values...
835 # match against attributes in a user SSL certificate
836 # attribute is one of DN/C/O/CN/L/ST [fast]
838 acl aclname ca_cert attribute values...
839 # match against attributes a users issuing CA SSL certificate
840 # attribute is one of DN/C/O/CN/L/ST [fast]
842 acl aclname ext_user username ...
843 acl aclname ext_user_regex [-i] pattern ...
844 # string match on username returned by external acl helper [slow]
845 # use REQUIRED to accept any non-null user name.
847 acl aclname tag tagvalue ...
848 # string match on tag returned by external acl helper [slow]
850 acl aclname hier_code codename ...
851 # string match against squid hierarchy code(s); [fast]
852 # e.g., DIRECT, PARENT_HIT, NONE, etc.
854 # NOTE: This has no effect in http_access rules. It only has
855 # effect in rules that affect the reply data stream such as
859 acl macaddress arp 09:00:2b:23:45:67
860 acl myexample dst_as 1241
861 acl password proxy_auth REQUIRED
862 acl fileupload req_mime_type -i ^multipart/form-data$
863 acl javascript rep_mime_type -i ^application/x-javascript$
867 # Recommended minimum configuration:
870 # Example rule allowing access from your local networks.
871 # Adapt to list your (internal) IP networks from where browsing
873 acl localnet src 10.0.0.0/8 # RFC1918 possible internal network
874 acl localnet src 172.16.0.0/12 # RFC1918 possible internal network
875 acl localnet src 192.168.0.0/16 # RFC1918 possible internal network
876 acl localnet src fc00::/7 # RFC 4193 local private network range
877 acl localnet src fe80::/10 # RFC 4291 link-local (directly plugged) machines
879 acl SSL_ports port 443
880 acl Safe_ports port 80 # http
881 acl Safe_ports port 21 # ftp
882 acl Safe_ports port 443 # https
883 acl Safe_ports port 70 # gopher
884 acl Safe_ports port 210 # wais
885 acl Safe_ports port 1025-65535 # unregistered ports
886 acl Safe_ports port 280 # http-mgmt
887 acl Safe_ports port 488 # gss-http
888 acl Safe_ports port 591 # filemaker
889 acl Safe_ports port 777 # multiling http
890 acl CONNECT method CONNECT
894 NAME: follow_x_forwarded_for
896 IFDEF: FOLLOW_X_FORWARDED_FOR
897 LOC: Config.accessList.followXFF
898 DEFAULT_IF_NONE: deny all
900 Allowing or Denying the X-Forwarded-For header to be followed to
901 find the original source of a request.
903 Requests may pass through a chain of several other proxies
904 before reaching us. The X-Forwarded-For header will contain a
905 comma-separated list of the IP addresses in the chain, with the
906 rightmost address being the most recent.
908 If a request reaches us from a source that is allowed by this
909 configuration item, then we consult the X-Forwarded-For header
910 to see where that host received the request from. If the
911 X-Forwarded-For header contains multiple addresses, we continue
912 backtracking until we reach an address for which we are not allowed
913 to follow the X-Forwarded-For header, or until we reach the first
914 address in the list. For the purpose of ACL used in the
915 follow_x_forwarded_for directive the src ACL type always matches
916 the address we are testing and srcdomain matches its rDNS.
918 The end result of this process is an IP address that we will
919 refer to as the indirect client address. This address may
920 be treated as the client address for access control, ICAP, delay
921 pools and logging, depending on the acl_uses_indirect_client,
922 icap_uses_indirect_client, delay_pool_uses_indirect_client,
923 log_uses_indirect_client and tproxy_uses_indirect_client options.
925 This clause only supports fast acl types.
926 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
928 SECURITY CONSIDERATIONS:
930 Any host for which we follow the X-Forwarded-For header
931 can place incorrect information in the header, and Squid
932 will use the incorrect information as if it were the
933 source address of the request. This may enable remote
934 hosts to bypass any access control restrictions that are
935 based on the client's source addresses.
939 acl localhost src 127.0.0.1
940 acl my_other_proxy srcdomain .proxy.example.com
941 follow_x_forwarded_for allow localhost
942 follow_x_forwarded_for allow my_other_proxy
945 NAME: acl_uses_indirect_client
948 IFDEF: FOLLOW_X_FORWARDED_FOR
950 LOC: Config.onoff.acl_uses_indirect_client
952 Controls whether the indirect client address
953 (see follow_x_forwarded_for) is used instead of the
954 direct client address in acl matching.
956 NOTE: maxconn ACL considers direct TCP links and indirect
957 clients will always have zero. So no match.
960 NAME: delay_pool_uses_indirect_client
963 IFDEF: FOLLOW_X_FORWARDED_FOR&&USE_DELAY_POOLS
965 LOC: Config.onoff.delay_pool_uses_indirect_client
967 Controls whether the indirect client address
968 (see follow_x_forwarded_for) is used instead of the
969 direct client address in delay pools.
972 NAME: log_uses_indirect_client
975 IFDEF: FOLLOW_X_FORWARDED_FOR
977 LOC: Config.onoff.log_uses_indirect_client
979 Controls whether the indirect client address
980 (see follow_x_forwarded_for) is used instead of the
981 direct client address in the access log.
984 NAME: tproxy_uses_indirect_client
987 IFDEF: FOLLOW_X_FORWARDED_FOR&&LINUX_NETFILTER
989 LOC: Config.onoff.tproxy_uses_indirect_client
991 Controls whether the indirect client address
992 (see follow_x_forwarded_for) is used instead of the
993 direct client address when spoofing the outgoing client.
995 This has no effect on requests arriving in non-tproxy
998 SECURITY WARNING: Usage of this option is dangerous
999 and should not be used trivially. Correct configuration
1000 of folow_x_forewarded_for with a limited set of trusted
1001 sources is required to prevent abuse of your proxy.
1006 LOC: Config.accessList.http
1007 DEFAULT_IF_NONE: deny all
1009 Allowing or Denying access based on defined access lists
1011 Access to the HTTP port:
1012 http_access allow|deny [!]aclname ...
1014 NOTE on default values:
1016 If there are no "access" lines present, the default is to deny
1019 If none of the "access" lines cause a match, the default is the
1020 opposite of the last line in the list. If the last line was
1021 deny, the default is allow. Conversely, if the last line
1022 is allow, the default will be deny. For these reasons, it is a
1023 good idea to have an "deny all" entry at the end of your access
1024 lists to avoid potential confusion.
1026 This clause supports both fast and slow acl types.
1027 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1032 # Recommended minimum Access Permission configuration:
1034 # Only allow cachemgr access from localhost
1035 http_access allow localhost manager
1036 http_access deny manager
1038 # Deny requests to certain unsafe ports
1039 http_access deny !Safe_ports
1041 # Deny CONNECT to other than secure SSL ports
1042 http_access deny CONNECT !SSL_ports
1044 # We strongly recommend the following be uncommented to protect innocent
1045 # web applications running on the proxy server who think the only
1046 # one who can access services on "localhost" is a local user
1047 #http_access deny to_localhost
1050 # INSERT YOUR OWN RULE(S) HERE TO ALLOW ACCESS FROM YOUR CLIENTS
1053 # Example rule allowing access from your local networks.
1054 # Adapt localnet in the ACL section to list your (internal) IP networks
1055 # from where browsing should be allowed
1056 http_access allow localnet
1057 http_access allow localhost
1059 # And finally deny all other access to this proxy
1060 http_access deny all
1064 NAME: adapted_http_access http_access2
1066 LOC: Config.accessList.adapted_http
1069 Allowing or Denying access based on defined access lists
1071 Essentially identical to http_access, but runs after redirectors
1072 and ICAP/eCAP adaptation. Allowing access control based on their
1075 If not set then only http_access is used.
1078 NAME: http_reply_access
1080 LOC: Config.accessList.reply
1083 Allow replies to client requests. This is complementary to http_access.
1085 http_reply_access allow|deny [!] aclname ...
1087 NOTE: if there are no access lines present, the default is to allow
1090 If none of the access lines cause a match the opposite of the
1091 last line will apply. Thus it is good practice to end the rules
1092 with an "allow all" or "deny all" entry.
1094 This clause supports both fast and slow acl types.
1095 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1100 LOC: Config.accessList.icp
1101 DEFAULT_IF_NONE: deny all
1103 Allowing or Denying access to the ICP port based on defined
1106 icp_access allow|deny [!]aclname ...
1108 See http_access for details
1110 This clause only supports fast acl types.
1111 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1113 # Allow ICP queries from local networks only
1114 #icp_access allow localnet
1115 #icp_access deny all
1121 LOC: Config.accessList.htcp
1122 DEFAULT_IF_NONE: deny all
1124 Allowing or Denying access to the HTCP port based on defined
1127 htcp_access allow|deny [!]aclname ...
1129 See http_access for details
1131 NOTE: The default if no htcp_access lines are present is to
1132 deny all traffic. This default may cause problems with peers
1133 using the htcp option.
1135 This clause only supports fast acl types.
1136 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1138 # Allow HTCP queries from local networks only
1139 #htcp_access allow localnet
1140 #htcp_access deny all
1143 NAME: htcp_clr_access
1146 LOC: Config.accessList.htcp_clr
1147 DEFAULT_IF_NONE: deny all
1149 Allowing or Denying access to purge content using HTCP based
1150 on defined access lists
1152 htcp_clr_access allow|deny [!]aclname ...
1154 See http_access for details
1156 This clause only supports fast acl types.
1157 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1159 # Allow HTCP CLR requests from trusted peers
1160 acl htcp_clr_peer src 172.16.1.2
1161 htcp_clr_access allow htcp_clr_peer
1166 LOC: Config.accessList.miss
1169 Determins whether network access is permitted when satisfying a request.
1172 to force your neighbors to use you as a sibling instead of
1175 acl localclients src 172.16.0.0/16
1176 miss_access allow localclients
1177 miss_access deny !localclients
1179 This means only your local clients are allowed to fetch relayed/MISS
1180 replies from the network and all other clients can only fetch cached
1184 The default for this setting allows all clients who passed the
1185 http_access rules to relay via this proxy.
1187 This clause only supports fast acl types.
1188 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1191 NAME: ident_lookup_access
1194 DEFAULT_IF_NONE: deny all
1195 LOC: Ident::TheConfig.identLookup
1197 A list of ACL elements which, if matched, cause an ident
1198 (RFC 931) lookup to be performed for this request. For
1199 example, you might choose to always perform ident lookups
1200 for your main multi-user Unix boxes, but not for your Macs
1201 and PCs. By default, ident lookups are not performed for
1204 To enable ident lookups for specific client addresses, you
1205 can follow this example:
1207 acl ident_aware_hosts src 198.168.1.0/24
1208 ident_lookup_access allow ident_aware_hosts
1209 ident_lookup_access deny all
1211 Only src type ACL checks are fully supported. A srcdomain
1212 ACL might work at times, but it will not always provide
1215 This clause only supports fast acl types.
1216 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1219 NAME: reply_body_max_size
1220 COMMENT: size [acl acl...]
1223 LOC: Config.ReplyBodySize
1225 This option specifies the maximum size of a reply body. It can be
1226 used to prevent users from downloading very large files, such as
1227 MP3's and movies. When the reply headers are received, the
1228 reply_body_max_size lines are processed, and the first line where
1229 all (if any) listed ACLs are true is used as the maximum body size
1232 This size is checked twice. First when we get the reply headers,
1233 we check the content-length value. If the content length value exists
1234 and is larger than the allowed size, the request is denied and the
1235 user receives an error message that says "the request or reply
1236 is too large." If there is no content-length, and the reply
1237 size exceeds this limit, the client's connection is just closed
1238 and they will receive a partial reply.
1240 WARNING: downstream caches probably can not detect a partial reply
1241 if there is no content-length header, so they will cache
1242 partial responses and give them out as hits. You should NOT
1243 use this option if you have downstream caches.
1245 WARNING: A maximum size smaller than the size of squid's error messages
1246 will cause an infinite loop and crash squid. Ensure that the smallest
1247 non-zero value you use is greater that the maximum header size plus
1248 the size of your largest error page.
1250 If you set this parameter none (the default), there will be
1253 Configuration Format is:
1254 reply_body_max_size SIZE UNITS [acl ...]
1256 reply_body_max_size 10 MB
1262 -----------------------------------------------------------------------------
1265 NAME: http_port ascii_port
1266 TYPE: http_port_list
1268 LOC: Config.Sockaddr.http
1270 Usage: port [mode] [options]
1271 hostname:port [mode] [options]
1272 1.2.3.4:port [mode] [options]
1274 The socket addresses where Squid will listen for HTTP client
1275 requests. You may specify multiple socket addresses.
1276 There are three forms: port alone, hostname with port, and
1277 IP address with port. If you specify a hostname or IP
1278 address, Squid binds the socket to that specific
1279 address. Most likely, you do not need to bind to a specific
1280 address, so you can use the port number alone.
1282 If you are running Squid in accelerator mode, you
1283 probably want to listen on port 80 also, or instead.
1285 The -a command line option may be used to specify additional
1286 port(s) where Squid listens for proxy request. Such ports will
1287 be plain proxy ports with no options.
1289 You may specify multiple socket addresses on multiple lines.
1293 intercept Support for IP-Layer interception of
1294 outgoing requests without browser settings.
1295 NP: disables authentication and IPv6 on the port.
1297 tproxy Support Linux TPROXY for spoofing outgoing
1298 connections using the client IP address.
1299 NP: disables authentication and maybe IPv6 on the port.
1301 accel Accelerator / reverse proxy mode
1303 ssl-bump Intercept each CONNECT request matching ssl_bump ACL,
1304 establish secure connection with the client and with
1305 the server, decrypt HTTP messages as they pass through
1306 Squid, and treat them as unencrypted HTTP messages,
1307 becoming the man-in-the-middle.
1309 The ssl_bump option is required to fully enable
1310 the SslBump feature.
1312 Omitting the mode flag causes default forward proxy mode to be used.
1315 Accelerator Mode Options:
1317 defaultsite=domainname
1318 What to use for the Host: header if it is not present
1319 in a request. Determines what site (not origin server)
1320 accelerators should consider the default.
1322 no-vhost Disable using HTTP/1.1 Host header for virtual domain support.
1324 protocol= Protocol to reconstruct accelerated requests with.
1325 Defaults to http for http_port and https for
1328 vport Virtual host port support. Using the http_port number
1329 instead of the port passed on Host: headers.
1331 vport=NN Virtual host port support. Using the specified port
1332 number instead of the port passed on Host: headers.
1335 Act as if this Squid is the origin server.
1336 This currently means generate new Date: and Expires:
1337 headers on HIT instead of adding Age:.
1339 ignore-cc Ignore request Cache-Control headers.
1341 WARNING: This option violates HTTP specifications if
1342 used in non-accelerator setups.
1344 allow-direct Allow direct forwarding in accelerator mode. Normally
1345 accelerated requests are denied direct forwarding as if
1346 never_direct was used.
1348 WARNING: this option opens accelerator mode to security
1349 vulnerabilities usually only affecting in interception
1350 mode. Make sure to protect forwarding with suitable
1351 http_access rules when using this.
1354 SSL Bump Mode Options:
1356 cert= Path to SSL certificate (PEM format).
1358 key= Path to SSL private key file (PEM format)
1359 if not specified, the certificate file is
1360 assumed to be a combined certificate and
1363 version= The version of SSL/TLS supported
1364 1 automatic (default)
1369 cipher= Colon separated list of supported ciphers.
1371 options= Various SSL engine options. The most important
1373 NO_SSLv2 Disallow the use of SSLv2
1374 NO_SSLv3 Disallow the use of SSLv3
1375 NO_TLSv1 Disallow the use of TLSv1
1376 SINGLE_DH_USE Always create a new key when using
1377 temporary/ephemeral DH key exchanges
1378 See src/ssl_support.c or OpenSSL SSL_CTX_set_options
1379 documentation for a complete list of options.
1381 clientca= File containing the list of CAs to use when
1382 requesting a client certificate.
1384 cafile= File containing additional CA certificates to
1385 use when verifying client certificates. If unset
1386 clientca will be used.
1388 capath= Directory containing additional CA certificates
1389 and CRL lists to use when verifying client certificates.
1391 crlfile= File of additional CRL lists to use when verifying
1392 the client certificate, in addition to CRLs stored in
1393 the capath. Implies VERIFY_CRL flag below.
1395 dhparams= File containing DH parameters for temporary/ephemeral
1398 sslflags= Various flags modifying the use of SSL:
1400 Don't request client certificates
1401 immediately, but wait until acl processing
1402 requires a certificate (not yet implemented).
1404 Don't use the default CA lists built in
1407 Don't allow for session reuse. Each connection
1408 will result in a new SSL session.
1410 Verify CRL lists when accepting client
1413 Verify CRL lists for all certificates in the
1414 client certificate chain.
1416 sslcontext= SSL session ID context identifier.
1418 generate-host-certificates[=<on|off>]
1419 Dynamically create SSL server certificates for the
1420 destination hosts of bumped CONNECT requests.When
1421 enabled, the cert and key options are used to sign
1422 generated certificates. Otherwise generated
1423 certificate will be selfsigned.
1424 If there is CA certificate life time of generated
1425 certificate equals lifetime of CA certificate. If
1426 generated certificate is selfsigned lifetime is three
1428 This option is enabled by default when SslBump is used.
1429 See the sslBump option above for more information.
1431 dynamic_cert_mem_cache_size=SIZE
1432 Approximate total RAM size spent on cached generated
1433 certificates. If set to zero, caching is disabled. The
1434 default value is 4MB. An average XXX-bit certificate
1435 consumes about XXX bytes of RAM.
1439 connection-auth[=on|off]
1440 use connection-auth=off to tell Squid to prevent
1441 forwarding Microsoft connection oriented authentication
1442 (NTLM, Negotiate and Kerberos)
1444 disable-pmtu-discovery=
1445 Control Path-MTU discovery usage:
1446 off lets OS decide on what to do (default).
1447 transparent disable PMTU discovery when transparent
1449 always disable always PMTU discovery.
1451 In many setups of transparently intercepting proxies
1452 Path-MTU discovery can not work on traffic towards the
1453 clients. This is the case when the intercepting device
1454 does not fully track connections and fails to forward
1455 ICMP must fragment messages to the cache server. If you
1456 have such setup and experience that certain clients
1457 sporadically hang or never complete requests set
1458 disable-pmtu-discovery option to 'transparent'.
1460 name= Specifies a internal name for the port. Defaults to
1461 the port specification (port or addr:port)
1463 tcpkeepalive[=idle,interval,timeout]
1464 Enable TCP keepalive probes of idle connections.
1465 In seconds; idle is the initial time before TCP starts
1466 probing the connection, interval how often to probe, and
1467 timeout the time before giving up.
1469 If you run Squid on a dual-homed machine with an internal
1470 and an external interface we recommend you to specify the
1471 internal address:port in http_port. This way Squid will only be
1472 visible on the internal address.
1476 # Squid normally listens to port 3128
1477 http_port @DEFAULT_HTTP_PORT@
1483 TYPE: https_port_list
1485 LOC: Config.Sockaddr.https
1487 Usage: [ip:]port cert=certificate.pem [key=key.pem] [mode] [options...]
1489 The socket address where Squid will listen for HTTPS client
1492 This is really only useful for situations where you are running
1493 squid in accelerator mode and you want to do the SSL work at the
1496 You may specify multiple socket addresses on multiple lines,
1497 each with their own SSL certificate and/or options.
1501 accel Accelerator / reverse proxy mode
1503 Omitting the mode flag causes default forward proxy mode to be used.
1506 See http_port for a list of generic options
1511 cert= Path to SSL certificate (PEM format).
1513 key= Path to SSL private key file (PEM format)
1514 if not specified, the certificate file is
1515 assumed to be a combined certificate and
1518 version= The version of SSL/TLS supported
1519 1 automatic (default)
1524 cipher= Colon separated list of supported ciphers.
1526 options= Various SSL engine options. The most important
1528 NO_SSLv2 Disallow the use of SSLv2
1529 NO_SSLv3 Disallow the use of SSLv3
1530 NO_TLSv1 Disallow the use of TLSv1
1531 SINGLE_DH_USE Always create a new key when using
1532 temporary/ephemeral DH key exchanges
1533 See src/ssl_support.c or OpenSSL SSL_CTX_set_options
1534 documentation for a complete list of options.
1536 clientca= File containing the list of CAs to use when
1537 requesting a client certificate.
1539 cafile= File containing additional CA certificates to
1540 use when verifying client certificates. If unset
1541 clientca will be used.
1543 capath= Directory containing additional CA certificates
1544 and CRL lists to use when verifying client certificates.
1546 crlfile= File of additional CRL lists to use when verifying
1547 the client certificate, in addition to CRLs stored in
1548 the capath. Implies VERIFY_CRL flag below.
1550 dhparams= File containing DH parameters for temporary/ephemeral
1553 sslflags= Various flags modifying the use of SSL:
1555 Don't request client certificates
1556 immediately, but wait until acl processing
1557 requires a certificate (not yet implemented).
1559 Don't use the default CA lists built in
1562 Don't allow for session reuse. Each connection
1563 will result in a new SSL session.
1565 Verify CRL lists when accepting client
1568 Verify CRL lists for all certificates in the
1569 client certificate chain.
1571 sslcontext= SSL session ID context identifier.
1575 NAME: tcp_outgoing_tos tcp_outgoing_ds tcp_outgoing_dscp
1578 LOC: Ip::Qos::TheConfig.tosToServer
1580 Allows you to select a TOS/Diffserv value for packets outgoing
1581 on the server side, based on an ACL.
1583 tcp_outgoing_tos ds-field [!]aclname ...
1585 Example where normal_service_net uses the TOS value 0x00
1586 and good_service_net uses 0x20
1588 acl normal_service_net src 10.0.0.0/24
1589 acl good_service_net src 10.0.1.0/24
1590 tcp_outgoing_tos 0x00 normal_service_net
1591 tcp_outgoing_tos 0x20 good_service_net
1593 TOS/DSCP values really only have local significance - so you should
1594 know what you're specifying. For more information, see RFC2474,
1595 RFC2475, and RFC3260.
1597 The TOS/DSCP byte must be exactly that - a octet value 0 - 255, or
1598 "default" to use whatever default your host has. Note that in
1599 practice often only multiples of 4 is usable as the two rightmost bits
1600 have been redefined for use by ECN (RFC 3168 section 23.1).
1602 Processing proceeds in the order specified, and stops at first fully
1606 NAME: clientside_tos
1609 LOC: Ip::Qos::TheConfig.tosToClient
1611 Allows you to select a TOS/Diffserv value for packets being transmitted
1612 on the client-side, based on an ACL.
1614 clientside_tos ds-field [!]aclname ...
1616 Example where normal_service_net uses the TOS value 0x00
1617 and good_service_net uses 0x20
1619 acl normal_service_net src 10.0.0.0/24
1620 acl good_service_net src 10.0.1.0/24
1621 clientside_tos 0x00 normal_service_net
1622 clientside_tos 0x20 good_service_net
1624 Note: This feature is incompatible with qos_flows. Any TOS values set here
1625 will be overwritten by TOS values in qos_flows.
1628 NAME: tcp_outgoing_mark
1630 IFDEF: SO_MARK&&USE_LIBCAP
1632 LOC: Ip::Qos::TheConfig.nfmarkToServer
1634 Allows you to apply a Netfilter mark value to outgoing packets
1635 on the server side, based on an ACL.
1637 tcp_outgoing_mark mark-value [!]aclname ...
1639 Example where normal_service_net uses the mark value 0x00
1640 and good_service_net uses 0x20
1642 acl normal_service_net src 10.0.0.0/24
1643 acl good_service_net src 10.0.1.0/24
1644 tcp_outgoing_mark 0x00 normal_service_net
1645 tcp_outgoing_mark 0x20 good_service_net
1648 NAME: clientside_mark
1650 IFDEF: SO_MARK&&USE_LIBCAP
1652 LOC: Ip::Qos::TheConfig.nfmarkToClient
1654 Allows you to apply a Netfilter mark value to packets being transmitted
1655 on the client-side, based on an ACL.
1657 clientside_mark mark-value [!]aclname ...
1659 Example where normal_service_net uses the mark value 0x00
1660 and good_service_net uses 0x20
1662 acl normal_service_net src 10.0.0.0/24
1663 acl good_service_net src 10.0.1.0/24
1664 clientside_mark 0x00 normal_service_net
1665 clientside_mark 0x20 good_service_net
1667 Note: This feature is incompatible with qos_flows. Any mark values set here
1668 will be overwritten by mark values in qos_flows.
1675 LOC: Ip::Qos::TheConfig
1677 Allows you to select a TOS/DSCP value to mark outgoing
1678 connections with, based on where the reply was sourced. For
1679 platforms using netfilter, allows you to set a netfilter mark
1680 value instead of, or in addition to, a TOS value.
1682 TOS values really only have local significance - so you should
1683 know what you're specifying. For more information, see RFC2474,
1684 RFC2475, and RFC3260.
1686 The TOS/DSCP byte must be exactly that - a octet value 0 - 255. Note that
1687 in practice often only multiples of 4 is usable as the two rightmost bits
1688 have been redefined for use by ECN (RFC 3168 section 23.1).
1690 Mark values can be any unsigned 32-bit integer value.
1692 This setting is configured by setting the following values:
1694 tos|mark Whether to set TOS or netfilter mark values
1696 local-hit=0xFF Value to mark local cache hits.
1698 sibling-hit=0xFF Value to mark hits from sibling peers.
1700 parent-hit=0xFF Value to mark hits from parent peers.
1702 miss=0xFF Value to mark cache misses. Takes precedence
1703 over the preserve-miss feature (see below).
1705 The TOS variant of the following features are only possible on Linux
1706 and require your kernel to be patched with the TOS preserving ZPH
1707 patch, available from http://zph.bratcheda.org
1708 No patch is needed to preserve the netfilter mark, which will work
1709 with all variants of netfilter.
1711 disable-preserve-miss
1712 This option disables the preservation of the TOS or netfilter
1713 mark. By default, the existing TOS or netfilter mark value of
1714 the response coming from the remote server will be retained
1715 and masked with miss-mark.
1716 NOTE: in the case of a netfilter mark, the mark must be set on
1717 the connection (using the CONNMARK target) not on the packet
1721 Allows you to mask certain bits in the TOS or mark value
1722 received from the remote server, before copying the value to
1723 the TOS sent towards clients.
1724 Default for tos: 0xFF (TOS from server is not changed).
1725 Default for mark: 0xFFFFFFFF (mark from server is not changed).
1727 All of these features require the --enable-zph-qos compilation flag
1728 (enabled by default). Netfilter marking also requires the
1729 libnetfilter_conntrack libraries (--with-netfilter-conntrack) and
1730 libcap 2.09+ (--with-libcap).
1734 NAME: tcp_outgoing_address
1737 LOC: Config.accessList.outgoing_address
1739 Allows you to map requests to different outgoing IP addresses
1740 based on the username or source address of the user making
1743 tcp_outgoing_address ipaddr [[!]aclname] ...
1746 Forwarding clients with dedicated IPs for certain subnets.
1748 acl normal_service_net src 10.0.0.0/24
1749 acl good_service_net src 10.0.2.0/24
1751 tcp_outgoing_address 2001:db8::c001 good_service_net
1752 tcp_outgoing_address 10.1.0.2 good_service_net
1754 tcp_outgoing_address 2001:db8::beef normal_service_net
1755 tcp_outgoing_address 10.1.0.1 normal_service_net
1757 tcp_outgoing_address 2001:db8::1
1758 tcp_outgoing_address 10.1.0.3
1760 Processing proceeds in the order specified, and stops at first fully
1763 Squid will add an implicit IP version test to each line.
1764 Requests going to IPv4 websites will use the outgoing 10.1.0.* addresses.
1765 Requests going to IPv6 websites will use the outgoing 2001:db8:* addresses.
1768 NOTE: The use of this directive using client dependent ACLs is
1769 incompatible with the use of server side persistent connections. To
1770 ensure correct results it is best to set server_persistent_connections
1771 to off when using this directive in such configurations.
1773 NOTE: The use of this directive to set a local IP on outgoing TCP links
1774 is incompatible with using TPROXY to set client IP out outbound TCP links.
1775 When needing to contact peers use the no-tproxy cache_peer option and the
1776 client_dst_passthru directive re-enable normal forwarding such as this.
1780 NAME: client_dst_passthru
1783 LOC: Config.onoff.client_dst_passthru
1785 With NAT or TPROXY intercepted traffic Squid may pass the request
1786 directly to the original client destination IP or seek a faster
1789 This option (on by default) prevents cache_peer and alternative DNS
1790 entries being used on intercepted traffic. Both of which lead to
1791 the security vulnerability outlined below.
1795 This directive should only be disabled if cache_peer are required.
1797 As described in CVE-2009-0801 when the Host: header alone is used
1798 to determine the destination of a request it becomes trivial for
1799 malicious scripts on remote websites to bypass browser same-origin
1800 security policy and sandboxing protections.
1802 The cause of this is that such applets are allowed to perform their
1803 own HTTP stack, in which case the same-origin policy of the browser
1804 sandbox only verifies that the applet tries to contact the same IP
1805 as from where it was loaded at the IP level. The Host: header may
1806 be different from the connected IP and approved origin.
1812 -----------------------------------------------------------------------------
1815 NAME: ssl_unclean_shutdown
1819 LOC: Config.SSL.unclean_shutdown
1821 Some browsers (especially MSIE) bugs out on SSL shutdown
1828 LOC: Config.SSL.ssl_engine
1831 The OpenSSL engine to use. You will need to set this if you
1832 would like to use hardware SSL acceleration for example.
1835 NAME: sslproxy_client_certificate
1838 LOC: Config.ssl_client.cert
1841 Client SSL Certificate to use when proxying https:// URLs
1844 NAME: sslproxy_client_key
1847 LOC: Config.ssl_client.key
1850 Client SSL Key to use when proxying https:// URLs
1853 NAME: sslproxy_version
1856 LOC: Config.ssl_client.version
1859 SSL version level to use when proxying https:// URLs
1862 NAME: sslproxy_options
1865 LOC: Config.ssl_client.options
1868 SSL engine options to use when proxying https:// URLs
1870 The most important being:
1872 NO_SSLv2 Disallow the use of SSLv2
1873 NO_SSLv3 Disallow the use of SSLv3
1874 NO_TLSv1 Disallow the use of TLSv1
1876 Always create a new key when using
1877 temporary/ephemeral DH key exchanges
1879 These options vary depending on your SSL engine.
1880 See the OpenSSL SSL_CTX_set_options documentation for a
1881 complete list of possible options.
1884 NAME: sslproxy_cipher
1887 LOC: Config.ssl_client.cipher
1890 SSL cipher list to use when proxying https:// URLs
1892 Colon separated list of supported ciphers.
1895 NAME: sslproxy_cafile
1898 LOC: Config.ssl_client.cafile
1901 file containing CA certificates to use when verifying server
1902 certificates while proxying https:// URLs
1905 NAME: sslproxy_capath
1908 LOC: Config.ssl_client.capath
1911 directory containing CA certificates to use when verifying
1912 server certificates while proxying https:// URLs
1918 LOC: Config.accessList.ssl_bump
1921 This ACL controls which CONNECT requests to an http_port
1922 marked with an sslBump flag are actually "bumped". Please
1923 see the sslBump flag of an http_port option for more details
1924 about decoding proxied SSL connections.
1926 By default, no requests are bumped.
1928 See also: http_port ssl-bump
1930 This clause supports both fast and slow acl types.
1931 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1934 # Example: Bump all requests except those originating from localhost and
1935 # those going to webax.com or example.com sites.
1937 acl localhost src 127.0.0.1/32
1938 acl broken_sites dstdomain .webax.com
1939 acl broken_sites dstdomain .example.com
1940 ssl_bump deny localhost
1941 ssl_bump deny broken_sites
1945 NAME: sslproxy_flags
1948 LOC: Config.ssl_client.flags
1951 Various flags modifying the use of SSL while proxying https:// URLs:
1952 DONT_VERIFY_PEER Accept certificates that fail verification.
1953 For refined control, see sslproxy_cert_error.
1954 NO_DEFAULT_CA Don't use the default CA list built in
1958 NAME: sslproxy_cert_error
1961 LOC: Config.ssl_client.cert_error
1964 Use this ACL to bypass server certificate validation errors.
1966 For example, the following lines will bypass all validation errors
1967 when talking to servers located at 172.16.0.0/16. All other
1968 validation errors will result in ERR_SECURE_CONNECT_FAIL error.
1970 acl BrokenServersAtTrustedIP dst 172.16.0.0/16
1971 sslproxy_cert_error allow BrokenServersAtTrustedIP
1972 sslproxy_cert_error deny all
1974 This clause only supports fast acl types.
1975 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1976 Using slow acl types may result in server crashes
1978 Without this option, all server certificate validation errors
1979 terminate the transaction. Bypassing validation errors is dangerous
1980 because an error usually implies that the server cannot be trusted and
1981 the connection may be insecure.
1983 See also: sslproxy_flags and DONT_VERIFY_PEER.
1985 Default setting: sslproxy_cert_error deny all
1988 NAME: sslpassword_program
1991 LOC: Config.Program.ssl_password
1994 Specify a program used for entering SSL key passphrases
1995 when using encrypted SSL certificate keys. If not specified
1996 keys must either be unencrypted, or Squid started with the -N
1997 option to allow it to query interactively for the passphrase.
1999 The key file name is given as argument to the program allowing
2000 selection of the right password if you have multiple encrypted
2005 OPTIONS RELATING TO EXTERNAL SSL_CRTD
2006 -----------------------------------------------------------------------------
2009 NAME: sslcrtd_program
2012 DEFAULT: @DEFAULT_SSL_CRTD@ -s @DEFAULT_SSL_DB_DIR@ -M 4MB
2013 LOC: Ssl::TheConfig.ssl_crtd
2015 Specify the location and options of the executable for ssl_crtd process.
2016 @DEFAULT_SSL_CRTD@ program requires -s and -M parameters
2017 For more information use:
2018 @DEFAULT_SSL_CRTD@ -h
2021 NAME: sslcrtd_children
2022 TYPE: HelperChildConfig
2024 DEFAULT: 32 startup=5 idle=1
2025 LOC: Ssl::TheConfig.ssl_crtdChildren
2027 The maximum number of processes spawn to service ssl server.
2028 The maximum this may be safely set to is 32.
2030 The startup= and idle= options allow some measure of skew in your
2035 Sets the minimum number of processes to spawn when Squid
2036 starts or reconfigures. When set to zero the first request will
2037 cause spawning of the first child process to handle it.
2039 Starting too few children temporary slows Squid under load while it
2040 tries to spawn enough additional processes to cope with traffic.
2044 Sets a minimum of how many processes Squid is to try and keep available
2045 at all times. When traffic begins to rise above what the existing
2046 processes can handle this many more will be spawned up to the maximum
2047 configured. A minimum setting of 1 is required.
2049 You must have at least one ssl_crtd process.
2053 OPTIONS WHICH AFFECT THE NEIGHBOR SELECTION ALGORITHM
2054 -----------------------------------------------------------------------------
2062 To specify other caches in a hierarchy, use the format:
2064 cache_peer hostname type http-port icp-port [options]
2069 # hostname type port port options
2070 # -------------------- -------- ----- ----- -----------
2071 cache_peer parent.foo.net parent 3128 3130 default
2072 cache_peer sib1.foo.net sibling 3128 3130 proxy-only
2073 cache_peer sib2.foo.net sibling 3128 3130 proxy-only
2074 cache_peer example.com parent 80 0 no-query default
2075 cache_peer cdn.example.com sibling 3128 0
2077 type: either 'parent', 'sibling', or 'multicast'.
2079 proxy-port: The port number where the peer accept HTTP requests.
2080 For other Squid proxies this is usually 3128
2081 For web servers this is usually 80
2083 icp-port: Used for querying neighbor caches about objects.
2084 Set to 0 if the peer does not support ICP or HTCP.
2085 See ICP and HTCP options below for additional details.
2088 ==== ICP OPTIONS ====
2090 You MUST also set icp_port and icp_access explicitly when using these options.
2091 The defaults will prevent peer traffic using ICP.
2094 no-query Disable ICP queries to this neighbor.
2097 Indicates the named peer is a member of a multicast group.
2098 ICP queries will not be sent directly to the peer, but ICP
2099 replies will be accepted from it.
2101 closest-only Indicates that, for ICP_OP_MISS replies, we'll only forward
2102 CLOSEST_PARENT_MISSes and never FIRST_PARENT_MISSes.
2105 To only send ICP queries to this neighbor infrequently.
2106 This is used to keep the neighbor round trip time updated
2107 and is usually used in conjunction with weighted-round-robin.
2110 ==== HTCP OPTIONS ====
2112 You MUST also set htcp_port and htcp_access explicitly when using these options.
2113 The defaults will prevent peer traffic using HTCP.
2116 htcp Send HTCP, instead of ICP, queries to the neighbor.
2117 You probably also want to set the "icp-port" to 4827
2118 instead of 3130. This directive accepts a comma separated
2119 list of options described below.
2121 htcp=oldsquid Send HTCP to old Squid versions (2.5 or earlier).
2123 htcp=no-clr Send HTCP to the neighbor but without
2124 sending any CLR requests. This cannot be used with
2127 htcp=only-clr Send HTCP to the neighbor but ONLY CLR requests.
2128 This cannot be used with no-clr.
2131 Send HTCP to the neighbor including CLRs but only when
2132 they do not result from PURGE requests.
2135 Forward any HTCP CLR requests this proxy receives to the peer.
2138 ==== PEER SELECTION METHODS ====
2140 The default peer selection method is ICP, with the first responding peer
2141 being used as source. These options can be used for better load balancing.
2144 default This is a parent cache which can be used as a "last-resort"
2145 if a peer cannot be located by any of the peer-selection methods.
2146 If specified more than once, only the first is used.
2148 round-robin Load-Balance parents which should be used in a round-robin
2149 fashion in the absence of any ICP queries.
2150 weight=N can be used to add bias.
2152 weighted-round-robin
2153 Load-Balance parents which should be used in a round-robin
2154 fashion with the frequency of each parent being based on the
2155 round trip time. Closer parents are used more often.
2156 Usually used for background-ping parents.
2157 weight=N can be used to add bias.
2159 carp Load-Balance parents which should be used as a CARP array.
2160 The requests will be distributed among the parents based on the
2161 CARP load balancing hash function based on their weight.
2163 userhash Load-balance parents based on the client proxy_auth or ident username.
2165 sourcehash Load-balance parents based on the client source IP.
2168 To be used only for cache peers of type "multicast".
2169 ALL members of this multicast group have "sibling"
2170 relationship with it, not "parent". This is to a mulicast
2171 group when the requested object would be fetched only from
2172 a "parent" cache, anyway. It's useful, e.g., when
2173 configuring a pool of redundant Squid proxies, being
2174 members of the same multicast group.
2177 ==== PEER SELECTION OPTIONS ====
2179 weight=N use to affect the selection of a peer during any weighted
2180 peer-selection mechanisms.
2181 The weight must be an integer; default is 1,
2182 larger weights are favored more.
2183 This option does not affect parent selection if a peering
2184 protocol is not in use.
2186 basetime=N Specify a base amount to be subtracted from round trip
2188 It is subtracted before division by weight in calculating
2189 which parent to fectch from. If the rtt is less than the
2190 base time the rtt is set to a minimal value.
2192 ttl=N Specify a TTL to use when sending multicast ICP queries
2194 Only useful when sending to a multicast group.
2195 Because we don't accept ICP replies from random
2196 hosts, you must configure other group members as
2197 peers with the 'multicast-responder' option.
2199 no-delay To prevent access to this neighbor from influencing the
2202 digest-url=URL Tell Squid to fetch the cache digest (if digests are
2203 enabled) for this host from the specified URL rather
2204 than the Squid default location.
2207 ==== CARP OPTIONS ====
2209 carp-key=key-specification
2210 use a different key than the full URL to hash against the peer.
2211 the key-specification is a comma-separated list of the keywords
2212 scheme, host, port, path, params
2213 Order is not important.
2215 ==== ACCELERATOR / REVERSE-PROXY OPTIONS ====
2217 originserver Causes this parent to be contacted as an origin server.
2218 Meant to be used in accelerator setups when the peer
2222 Set the Host header of requests forwarded to this peer.
2223 Useful in accelerator setups where the server (peer)
2224 expects a certain domain name but clients may request
2225 others. ie example.com or www.example.com
2227 no-digest Disable request of cache digests.
2230 Disables requesting ICMP RTT database (NetDB).
2233 ==== AUTHENTICATION OPTIONS ====
2236 If this is a personal/workgroup proxy and your parent
2237 requires proxy authentication.
2239 Note: The string can include URL escapes (i.e. %20 for
2240 spaces). This also means % must be written as %%.
2243 Send login details received from client to this peer.
2244 Both Proxy- and WWW-Authorization headers are passed
2245 without alteration to the peer.
2246 Authentication is not required by Squid for this to work.
2248 Note: This will pass any form of authentication but
2249 only Basic auth will work through a proxy unless the
2250 connection-auth options are also used.
2252 login=PASS Send login details received from client to this peer.
2253 Authentication is not required by this option.
2255 If there are no client-provided authentication headers
2256 to pass on, but username and password are available
2257 from an external ACL user= and password= result tags
2258 they may be sent instead.
2260 Note: To combine this with proxy_auth both proxies must
2261 share the same user database as HTTP only allows for
2262 a single login (one for proxy, one for origin server).
2263 Also be warned this will expose your users proxy
2264 password to the peer. USE WITH CAUTION
2267 Send the username to the upstream cache, but with a
2268 fixed password. This is meant to be used when the peer
2269 is in another administrative domain, but it is still
2270 needed to identify each user.
2271 The star can optionally be followed by some extra
2272 information which is added to the username. This can
2273 be used to identify this proxy to the peer, similar to
2274 the login=username:password option above.
2277 If this is a personal/workgroup proxy and your parent
2278 requires a secure proxy authentication.
2279 The first principal from the default keytab or defined by
2280 the environment variable KRB5_KTNAME will be used.
2282 WARNING: The connection may transmit requests from multiple
2283 clients. Negotiate often assumes end-to-end authentication
2284 and a single-client. Which is not strictly true here.
2286 login=NEGOTIATE:principal_name
2287 If this is a personal/workgroup proxy and your parent
2288 requires a secure proxy authentication.
2289 The principal principal_name from the default keytab or
2290 defined by the environment variable KRB5_KTNAME will be
2293 WARNING: The connection may transmit requests from multiple
2294 clients. Negotiate often assumes end-to-end authentication
2295 and a single-client. Which is not strictly true here.
2297 connection-auth=on|off
2298 Tell Squid that this peer does or not support Microsoft
2299 connection oriented authentication, and any such
2300 challenges received from there should be ignored.
2301 Default is auto to automatically determine the status
2305 ==== SSL / HTTPS / TLS OPTIONS ====
2307 ssl Encrypt connections to this peer with SSL/TLS.
2309 sslcert=/path/to/ssl/certificate
2310 A client SSL certificate to use when connecting to
2313 sslkey=/path/to/ssl/key
2314 The private SSL key corresponding to sslcert above.
2315 If 'sslkey' is not specified 'sslcert' is assumed to
2316 reference a combined file containing both the
2317 certificate and the key.
2320 The SSL version to use when connecting to this peer
2321 1 = automatic (default)
2326 sslcipher=... The list of valid SSL ciphers to use when connecting
2329 ssloptions=... Specify various SSL engine options:
2330 NO_SSLv2 Disallow the use of SSLv2
2331 NO_SSLv3 Disallow the use of SSLv3
2332 NO_TLSv1 Disallow the use of TLSv1
2333 See src/ssl_support.c or the OpenSSL documentation for
2334 a more complete list.
2336 sslcafile=... A file containing additional CA certificates to use
2337 when verifying the peer certificate.
2339 sslcapath=... A directory containing additional CA certificates to
2340 use when verifying the peer certificate.
2342 sslcrlfile=... A certificate revocation list file to use when
2343 verifying the peer certificate.
2345 sslflags=... Specify various flags modifying the SSL implementation:
2348 Accept certificates even if they fail to
2351 Don't use the default CA list built in
2354 Don't verify the peer certificate
2355 matches the server name
2357 ssldomain= The peer name as advertised in it's certificate.
2358 Used for verifying the correctness of the received peer
2359 certificate. If not specified the peer hostname will be
2363 Enable the "Front-End-Https: On" header needed when
2364 using Squid as a SSL frontend in front of Microsoft OWA.
2365 See MS KB document Q307347 for details on this header.
2366 If set to auto the header will only be added if the
2367 request is forwarded as a https:// URL.
2370 ==== GENERAL OPTIONS ====
2373 A peer-specific connect timeout.
2374 Also see the peer_connect_timeout directive.
2376 connect-fail-limit=N
2377 How many times connecting to a peer must fail before
2378 it is marked as down. Default is 10.
2380 allow-miss Disable Squid's use of only-if-cached when forwarding
2381 requests to siblings. This is primarily useful when
2382 icp_hit_stale is used by the sibling. To extensive use
2383 of this option may result in forwarding loops, and you
2384 should avoid having two-way peerings with this option.
2385 For example to deny peer usage on requests from peer
2386 by denying cache_peer_access if the source is a peer.
2388 max-conn=N Limit the amount of connections Squid may open to this
2391 name=xxx Unique name for the peer.
2392 Required if you have multiple peers on the same host
2393 but different ports.
2394 This name can be used in cache_peer_access and similar
2395 directives to dentify the peer.
2396 Can be used by outgoing access controls through the
2399 no-tproxy Do not use the client-spoof TPROXY support when forwarding
2400 requests to this peer. Use normal address selection instead.
2402 proxy-only objects fetched from the peer will not be stored locally.
2406 NAME: cache_peer_domain cache_host_domain
2411 Use to limit the domains for which a neighbor cache will be
2414 cache_peer_domain cache-host domain [domain ...]
2415 cache_peer_domain cache-host !domain
2417 For example, specifying
2419 cache_peer_domain parent.foo.net .edu
2421 has the effect such that UDP query packets are sent to
2422 'bigserver' only when the requested object exists on a
2423 server in the .edu domain. Prefixing the domainname
2424 with '!' means the cache will be queried for objects
2427 NOTE: * Any number of domains may be given for a cache-host,
2428 either on the same or separate lines.
2429 * When multiple domains are given for a particular
2430 cache-host, the first matched domain is applied.
2431 * Cache hosts with no domain restrictions are queried
2433 * There are no defaults.
2434 * There is also a 'cache_peer_access' tag in the ACL
2438 NAME: cache_peer_access
2443 Similar to 'cache_peer_domain' but provides more flexibility by
2446 cache_peer_access cache-host allow|deny [!]aclname ...
2448 The syntax is identical to 'http_access' and the other lists of
2449 ACL elements. See the comments for 'http_access' below, or
2450 the Squid FAQ (http://wiki.squid-cache.org/SquidFaq/SquidAcl).
2453 NAME: neighbor_type_domain
2454 TYPE: hostdomaintype
2458 usage: neighbor_type_domain neighbor parent|sibling domain domain ...
2460 Modifying the neighbor type for specific domains is now
2461 possible. You can treat some domains differently than the
2462 default neighbor type specified on the 'cache_peer' line.
2463 Normally it should only be necessary to list domains which
2464 should be treated differently because the default neighbor type
2465 applies for hostnames which do not match domains listed here.
2468 cache_peer cache.foo.org parent 3128 3130
2469 neighbor_type_domain cache.foo.org sibling .com .net
2470 neighbor_type_domain cache.foo.org sibling .au .de
2473 NAME: dead_peer_timeout
2477 LOC: Config.Timeout.deadPeer
2479 This controls how long Squid waits to declare a peer cache
2480 as "dead." If there are no ICP replies received in this
2481 amount of time, Squid will declare the peer dead and not
2482 expect to receive any further ICP replies. However, it
2483 continues to send ICP queries, and will mark the peer as
2484 alive upon receipt of the first subsequent ICP reply.
2486 This timeout also affects when Squid expects to receive ICP
2487 replies from peers. If more than 'dead_peer' seconds have
2488 passed since the last ICP reply was received, Squid will not
2489 expect to receive an ICP reply on the next query. Thus, if
2490 your time between requests is greater than this timeout, you
2491 will see a lot of requests sent DIRECT to origin servers
2492 instead of to your parents.
2495 NAME: forward_max_tries
2498 LOC: Config.forward_max_tries
2500 Controls how many different forward paths Squid will try
2501 before giving up. See also forward_timeout.
2503 NOTE: connect_retries (default: none) can make each of these
2504 possible forwarding paths be tried multiple times.
2507 NAME: hierarchy_stoplist
2510 LOC: Config.hierarchy_stoplist
2512 A list of words which, if found in a URL, cause the object to
2513 be handled directly by this cache. In other words, use this
2514 to not query neighbor caches for certain objects. You may
2515 list this option multiple times.
2518 hierarchy_stoplist cgi-bin ?
2520 Note: never_direct overrides this option.
2524 MEMORY CACHE OPTIONS
2525 -----------------------------------------------------------------------------
2532 LOC: Config.memMaxSize
2534 NOTE: THIS PARAMETER DOES NOT SPECIFY THE MAXIMUM PROCESS SIZE.
2535 IT ONLY PLACES A LIMIT ON HOW MUCH ADDITIONAL MEMORY SQUID WILL
2536 USE AS A MEMORY CACHE OF OBJECTS. SQUID USES MEMORY FOR OTHER
2537 THINGS AS WELL. SEE THE SQUID FAQ SECTION 8 FOR DETAILS.
2539 'cache_mem' specifies the ideal amount of memory to be used
2541 * In-Transit objects
2543 * Negative-Cached objects
2545 Data for these objects are stored in 4 KB blocks. This
2546 parameter specifies the ideal upper limit on the total size of
2547 4 KB blocks allocated. In-Transit objects take the highest
2550 In-transit objects have priority over the others. When
2551 additional space is needed for incoming data, negative-cached
2552 and hot objects will be released. In other words, the
2553 negative-cached and hot objects will fill up any unused space
2554 not needed for in-transit objects.
2556 If circumstances require, this limit will be exceeded.
2557 Specifically, if your incoming request rate requires more than
2558 'cache_mem' of memory to hold in-transit objects, Squid will
2559 exceed this limit to satisfy the new requests. When the load
2560 decreases, blocks will be freed until the high-water mark is
2561 reached. Thereafter, blocks will be used to store hot
2564 If shared memory caching is enabled, Squid does not use the shared
2565 cache space for in-transit objects, but they still consume as much
2566 local memory as they need. For more details about the shared memory
2567 cache, see memory_cache_shared.
2570 NAME: maximum_object_size_in_memory
2574 LOC: Config.Store.maxInMemObjSize
2576 Objects greater than this size will not be attempted to kept in
2577 the memory cache. This should be set high enough to keep objects
2578 accessed frequently in memory to improve performance whilst low
2579 enough to keep larger objects from hoarding cache_mem.
2582 NAME: memory_cache_shared
2585 LOC: Config.memShared
2587 DEFAULT_DOC: "on" where supported if doing memory caching with multiple SMP workers.
2589 Controls whether the memory cache is shared among SMP workers.
2591 The shared memory cache is meant to occupy cache_mem bytes and replace
2592 the non-shared memory cache, although some entities may still be
2593 cached locally by workers for now (e.g., internal and in-transit
2594 objects may be served from a local memory cache even if shared memory
2595 caching is enabled).
2597 By default, the memory cache is shared if and only if all of the
2598 following conditions are satisfied: Squid runs in SMP mode with
2599 multiple workers, cache_mem is positive, and Squid environment
2600 supports required IPC primitives (e.g., POSIX shared memory segments
2601 and GCC-style atomic operations).
2603 To avoid blocking locks, shared memory uses opportunistic algorithms
2604 that do not guarantee that every cachable entity that could have been
2605 shared among SMP workers will actually be shared.
2607 Currently, entities exceeding 32KB in size cannot be shared.
2610 NAME: memory_cache_mode
2615 Controls which objects to keep in the memory cache (cache_mem)
2617 always Keep most recently fetched objects in memory (default)
2619 disk Only disk cache hits are kept in memory, which means
2620 an object must first be cached on disk and then hit
2621 a second time before cached in memory.
2623 network Only objects fetched from network is kept in memory
2626 NAME: memory_replacement_policy
2628 LOC: Config.memPolicy
2631 The memory replacement policy parameter determines which
2632 objects are purged from memory when memory space is needed.
2634 See cache_replacement_policy for details.
2639 -----------------------------------------------------------------------------
2642 NAME: cache_replacement_policy
2644 LOC: Config.replPolicy
2647 The cache replacement policy parameter determines which
2648 objects are evicted (replaced) when disk space is needed.
2650 lru : Squid's original list based LRU policy
2651 heap GDSF : Greedy-Dual Size Frequency
2652 heap LFUDA: Least Frequently Used with Dynamic Aging
2653 heap LRU : LRU policy implemented using a heap
2655 Applies to any cache_dir lines listed below this.
2657 The LRU policies keeps recently referenced objects.
2659 The heap GDSF policy optimizes object hit rate by keeping smaller
2660 popular objects in cache so it has a better chance of getting a
2661 hit. It achieves a lower byte hit rate than LFUDA though since
2662 it evicts larger (possibly popular) objects.
2664 The heap LFUDA policy keeps popular objects in cache regardless of
2665 their size and thus optimizes byte hit rate at the expense of
2666 hit rate since one large, popular object will prevent many
2667 smaller, slightly less popular objects from being cached.
2669 Both policies utilize a dynamic aging mechanism that prevents
2670 cache pollution that can otherwise occur with frequency-based
2671 replacement policies.
2673 NOTE: if using the LFUDA replacement policy you should increase
2674 the value of maximum_object_size above its default of 4096 KB to
2675 to maximize the potential byte hit rate improvement of LFUDA.
2677 For more information about the GDSF and LFUDA cache replacement
2678 policies see http://www.hpl.hp.com/techreports/1999/HPL-1999-69.html
2679 and http://fog.hpl.external.hp.com/techreports/98/HPL-98-173.html.
2685 LOC: Config.cacheSwap
2689 cache_dir Type Directory-Name Fs-specific-data [options]
2691 You can specify multiple cache_dir lines to spread the
2692 cache among different disk partitions.
2694 Type specifies the kind of storage system to use. Only "ufs"
2695 is built by default. To enable any of the other storage systems
2696 see the --enable-storeio configure option.
2698 'Directory' is a top-level directory where cache swap
2699 files will be stored. If you want to use an entire disk
2700 for caching, this can be the mount-point directory.
2701 The directory must exist and be writable by the Squid
2702 process. Squid will NOT create this directory for you.
2704 In SMP configurations, cache_dir must not precede the workers option
2705 and should use configuration macros or conditionals to give each
2706 worker interested in disk caching a dedicated cache directory.
2710 "ufs" is the old well-known Squid storage format that has always
2713 cache_dir ufs Directory-Name Mbytes L1 L2 [options]
2715 'Mbytes' is the amount of disk space (MB) to use under this
2716 directory. The default is 100 MB. Change this to suit your
2717 configuration. Do NOT put the size of your disk drive here.
2718 Instead, if you want Squid to use the entire disk drive,
2719 subtract 20% and use that value.
2721 'L1' is the number of first-level subdirectories which
2722 will be created under the 'Directory'. The default is 16.
2724 'L2' is the number of second-level subdirectories which
2725 will be created under each first-level directory. The default
2728 The aufs store type:
2730 "aufs" uses the same storage format as "ufs", utilizing
2731 POSIX-threads to avoid blocking the main Squid process on
2732 disk-I/O. This was formerly known in Squid as async-io.
2734 cache_dir aufs Directory-Name Mbytes L1 L2 [options]
2736 see argument descriptions under ufs above
2738 The diskd store type:
2740 "diskd" uses the same storage format as "ufs", utilizing a
2741 separate process to avoid blocking the main Squid process on
2744 cache_dir diskd Directory-Name Mbytes L1 L2 [options] [Q1=n] [Q2=n]
2746 see argument descriptions under ufs above
2748 Q1 specifies the number of unacknowledged I/O requests when Squid
2749 stops opening new files. If this many messages are in the queues,
2750 Squid won't open new files. Default is 64
2752 Q2 specifies the number of unacknowledged messages when Squid
2753 starts blocking. If this many messages are in the queues,
2754 Squid blocks until it receives some replies. Default is 72
2756 When Q1 < Q2 (the default), the cache directory is optimized
2757 for lower response time at the expense of a decrease in hit
2758 ratio. If Q1 > Q2, the cache directory is optimized for
2759 higher hit ratio at the expense of an increase in response
2762 The rock store type:
2764 cache_dir rock Directory-Name Mbytes <max-size=bytes> [options]
2766 The Rock Store type is a database-style storage. All cached
2767 entries are stored in a "database" file, using fixed-size slots,
2768 one entry per slot. The database size is specified in MB. The
2769 slot size is specified in bytes using the max-size option. See
2770 below for more info on the max-size option.
2772 swap-timeout=msec: Squid will not start writing a miss to or
2773 reading a hit from disk if it estimates that the swap operation
2774 will take more than the specified number of milliseconds. By
2775 default and when set to zero, disables the disk I/O time limit
2776 enforcement. Ignored when using blocking I/O module because
2777 blocking synchronous I/O does not allow Squid to estimate the
2778 expected swap wait time.
2780 max-swap-rate=swaps/sec: Artificially limits disk access using
2781 the specified I/O rate limit. Swap in and swap out requests that
2782 would cause the average I/O rate to exceed the limit are
2783 delayed. This is necessary on file systems that buffer "too
2784 many" writes and then start blocking Squid and other processes
2785 while committing those writes to disk. Usually used together
2786 with swap-timeout to avoid excessive delays and queue overflows
2787 when disk demand exceeds available disk "bandwidth". By default
2788 and when set to zero, disables the disk I/O rate limit
2789 enforcement. Currently supported by IpcIo module only.
2792 The coss store type:
2794 NP: COSS filesystem in Squid-3 has been deemed too unstable for
2795 production use and has thus been removed from this release.
2796 We hope that it can be made usable again soon.
2798 block-size=n defines the "block size" for COSS cache_dir's.
2799 Squid uses file numbers as block numbers. Since file numbers
2800 are limited to 24 bits, the block size determines the maximum
2801 size of the COSS partition. The default is 512 bytes, which
2802 leads to a maximum cache_dir size of 512<<24, or 8 GB. Note
2803 you should not change the coss block size after Squid
2804 has written some objects to the cache_dir.
2806 The coss file store has changed from 2.5. Now it uses a file
2807 called 'stripe' in the directory names in the config - and
2808 this will be created by squid -z.
2812 no-store, no new objects should be stored to this cache_dir
2814 min-size=n, refers to the min object size in bytes this cache_dir
2815 will accept. It's used to restrict a cache_dir to only store
2816 large objects (e.g. aufs) while other storedirs are optimized
2817 for smaller objects (e.g. COSS). Defaults to 0.
2819 max-size=n, refers to the max object size in bytes this cache_dir
2820 supports. It is used to select the cache_dir to store the object.
2821 Note: To make optimal use of the max-size limits you should order
2822 the cache_dir lines with the smallest max-size value first and the
2823 ones with no max-size specification last.
2825 Note for coss, max-size must be less than COSS_MEMBUF_SZ,
2826 which can be changed with the --with-coss-membuf-size=N configure
2830 # Uncomment and adjust the following to add a disk cache directory.
2831 #cache_dir ufs @DEFAULT_SWAP_DIR@ 100 16 256
2835 NAME: store_dir_select_algorithm
2837 LOC: Config.store_dir_select_algorithm
2840 Set this to 'round-robin' as an alternative.
2843 NAME: max_open_disk_fds
2845 LOC: Config.max_open_disk_fds
2848 To avoid having disk as the I/O bottleneck Squid can optionally
2849 bypass the on-disk cache if more than this amount of disk file
2850 descriptors are open.
2852 A value of 0 indicates no limit.
2855 NAME: minimum_object_size
2859 LOC: Config.Store.minObjectSize
2861 Objects smaller than this size will NOT be saved on disk. The
2862 value is specified in kilobytes, and the default is 0 KB, which
2863 means there is no minimum.
2866 NAME: maximum_object_size
2870 LOC: Config.Store.maxObjectSize
2872 Objects larger than this size will NOT be saved on disk. The
2873 value is specified in kilobytes, and the default is 4MB. If
2874 you wish to get a high BYTES hit ratio, you should probably
2875 increase this (one 32 MB object hit counts for 3200 10KB
2876 hits). If you wish to increase speed more than your want to
2877 save bandwidth you should leave this low.
2879 NOTE: if using the LFUDA replacement policy you should increase
2880 this value to maximize the byte hit rate improvement of LFUDA!
2881 See replacement_policy below for a discussion of this policy.
2884 NAME: cache_swap_low
2885 COMMENT: (percent, 0-100)
2888 LOC: Config.Swap.lowWaterMark
2891 NAME: cache_swap_high
2892 COMMENT: (percent, 0-100)
2895 LOC: Config.Swap.highWaterMark
2898 The low- and high-water marks for cache object replacement.
2899 Replacement begins when the swap (disk) usage is above the
2900 low-water mark and attempts to maintain utilization near the
2901 low-water mark. As swap utilization gets close to high-water
2902 mark object eviction becomes more aggressive. If utilization is
2903 close to the low-water mark less replacement is done each time.
2905 Defaults are 90% and 95%. If you have a large cache, 5% could be
2906 hundreds of MB. If this is the case you may wish to set these
2907 numbers closer together.
2912 -----------------------------------------------------------------------------
2922 logformat <name> <format specification>
2924 Defines an access log format.
2926 The <format specification> is a string with embedded % format codes
2928 % format codes all follow the same basic structure where all but
2929 the formatcode is optional. Output strings are automatically escaped
2930 as required according to their context and the output format
2931 modifiers are usually not needed, but can be specified if an explicit
2932 output format is desired.
2934 % ["|[|'|#] [-] [[0]width] [{argument}] formatcode
2936 " output in quoted string format
2937 [ output in squid text log format as used by log_mime_hdrs
2938 # output in URL quoted format
2943 width minimum and/or maximum field width:
2944 [width_min][.width_max]
2945 When minimum starts with 0, the field is zero-padded.
2946 String values exceeding maximum width are truncated.
2948 {arg} argument such as header name etc
2952 % a literal % character
2953 sn Unique sequence number per log line entry
2954 err_code The ID of an error response served by Squid or
2955 a similar internal error identifier.
2956 err_detail Additional err_code-dependent error information.
2958 Connection related format codes:
2960 >a Client source IP address
2962 >p Client source port
2963 >eui Client source EUI (MAC address, EUI-48 or EUI-64 identifier)
2964 >la Local IP address the client connected to
2965 >lp Local port number the client connected to
2967 la Local listening IP address the client connection was connected to.
2968 lp Local listening port number the client connection was connected to.
2970 <a Server IP address of the last server or peer connection
2971 <A Server FQDN or peer name
2972 <p Server port number of the last server or peer connection
2973 <la Local IP address of the last server or peer connection
2974 <lp Local port number of the last server or peer connection
2976 Time related format codes:
2978 ts Seconds since epoch
2979 tu subsecond time (milliseconds)
2980 tl Local time. Optional strftime format argument
2981 default %d/%b/%Y:%H:%M:%S %z
2982 tg GMT time. Optional strftime format argument
2983 default %d/%b/%Y:%H:%M:%S %z
2984 tr Response time (milliseconds)
2985 dt Total time spent making DNS lookups (milliseconds)
2987 Access Control related format codes:
2989 et Tag returned by external acl
2990 ea Log string returned by external acl
2991 un User name (any available)
2992 ul User name from authentication
2993 ue User name from external acl helper
2994 ui User name from ident
2995 us User name from SSL
2997 HTTP related format codes:
2999 [http::]>h Original request header. Optional header name argument
3000 on the format header[:[separator]element]
3001 [http::]>ha The HTTP request headers after adaptation and redirection.
3002 Optional header name argument as for >h
3003 [http::]<h Reply header. Optional header name argument
3005 [http::]>Hs HTTP status code sent to the client
3006 [http::]<Hs HTTP status code received from the next hop
3007 [http::]<bs Number of HTTP-equivalent message body bytes
3008 received from the next hop, excluding chunked
3009 transfer encoding and control messages.
3010 Generated FTP/Gopher listings are treated as
3012 [http::]mt MIME content type
3013 [http::]rm Request method (GET/POST etc)
3014 [http::]>rm Request method from client
3015 [http::]<rm Request method sent to server or peer
3016 [http::]ru Request URL from client (historic, filtered for logging)
3017 [http::]>ru Request URL from client
3018 [http::]<ru Request URL sent to server or peer
3019 [http::]rp Request URL-Path excluding hostname
3020 [http::]>rp Request URL-Path excluding hostname from client
3021 [http::]<rp Request URL-Path excluding hostname sento to server or peer
3022 [http::]rv Request protocol version
3023 [http::]>rv Request protocol version from client
3024 [http::]<rv Request protocol version sent to server or peer
3025 [http::]<st Sent reply size including HTTP headers
3026 [http::]>st Received request size including HTTP headers. In the
3027 case of chunked requests the chunked encoding metadata
3029 [http::]>sh Received HTTP request headers size
3030 [http::]<sh Sent HTTP reply headers size
3031 [http::]st Request+Reply size including HTTP headers
3032 [http::]<sH Reply high offset sent
3033 [http::]<sS Upstream object size
3034 [http::]<pt Peer response time in milliseconds. The timer starts
3035 when the last request byte is sent to the next hop
3036 and stops when the last response byte is received.
3037 [http::]<tt Total server-side time in milliseconds. The timer
3038 starts with the first connect request (or write I/O)
3039 sent to the first selected peer. The timer stops
3040 with the last I/O with the last peer.
3042 Squid handling related format codes:
3044 Ss Squid request status (TCP_MISS etc)
3045 Sh Squid hierarchy status (DEFAULT_PARENT etc)
3047 If ICAP is enabled, the following code becomes available (as
3048 well as ICAP log codes documented with the icap_log option):
3050 icap::tt Total ICAP processing time for the HTTP
3051 transaction. The timer ticks when ICAP
3052 ACLs are checked and when ICAP
3053 transaction is in progress.
3055 If adaptation is enabled the following three codes become available:
3057 adapt::<last_h The header of the last ICAP response or
3058 meta-information from the last eCAP
3059 transaction related to the HTTP transaction.
3060 Like <h, accepts an optional header name
3063 adapt::sum_trs Summed adaptation transaction response
3064 times recorded as a comma-separated list in
3065 the order of transaction start time. Each time
3066 value is recorded as an integer number,
3067 representing response time of one or more
3068 adaptation (ICAP or eCAP) transaction in
3069 milliseconds. When a failed transaction is
3070 being retried or repeated, its time is not
3071 logged individually but added to the
3072 replacement (next) transaction. See also:
3075 adapt::all_trs All adaptation transaction response times.
3076 Same as adaptation_strs but response times of
3077 individual transactions are never added
3078 together. Instead, all transaction response
3079 times are recorded individually.
3081 You can prefix adapt::*_trs format codes with adaptation
3082 service name in curly braces to record response time(s) specific
3083 to that service. For example: %{my_service}adapt::sum_trs
3085 The default formats available (which do not need re-defining) are:
3087 logformat squid %ts.%03tu %6tr %>a %Ss/%03>Hs %<st %rm %ru %[un %Sh/%<a %mt
3088 logformat common %>a %[ui %[un [%tl] "%rm %ru HTTP/%rv" %>Hs %<st %Ss:%Sh
3089 logformat combined %>a %[ui %[un [%tl] "%rm %ru HTTP/%rv" %>Hs %<st "%{Referer}>h" "%{User-Agent}>h" %Ss:%Sh
3090 logformat referrer %ts.%03tu %>a %{Referer}>h %ru
3091 logformat useragent %>a [%tl] "%{User-Agent}>h"
3093 NOTE: When the log_mime_hdrs directive is set to ON.
3094 The squid, common and combined formats have a safely encoded copy
3095 of the mime headers appended to each line within a pair of brackets.
3097 NOTE: The common and combined formats are not quite true to the Apache definition.
3098 The logs from Squid contain an extra status and hierarchy code appended.
3102 NAME: access_log cache_access_log
3104 LOC: Config.Log.accesslogs
3105 DEFAULT_IF_NONE: daemon:@DEFAULT_ACCESS_LOG@ squid
3107 These files log client request activities. Has a line every HTTP or
3108 ICP request. The format is:
3109 access_log <module>:<place> [<logformat name> [acl acl ...]]
3110 access_log none [acl acl ...]]
3112 Will log to the specified module:place using the specified format (which
3113 must be defined in a logformat directive) those entries which match
3114 ALL the acl's specified (which must be defined in acl clauses).
3115 If no acl is specified, all requests will be logged to this destination.
3117 ===== Modules Currently available =====
3119 none Do not log any requests matching these ACL.
3120 Do not specify Place or logformat name.
3122 stdio Write each log line to disk immediately at the completion of
3124 Place: the filename and path to be written.
3126 daemon Very similar to stdio. But instead of writing to disk the log
3127 line is passed to a daemon helper for asychronous handling instead.
3128 Place: varies depending on the daemon.
3130 log_file_daemon Place: the file name and path to be written.
3132 syslog To log each request via syslog facility.
3133 Place: The syslog facility and priority level for these entries.
3134 Place Format: facility.priority
3136 where facility could be any of:
3137 authpriv, daemon, local0 ... local7 or user.
3139 And priority could be any of:
3140 err, warning, notice, info, debug.
3142 udp To send each log line as text data to a UDP receiver.
3143 Place: The destination host name or IP and port.
3144 Place Format: \\host:port
3146 tcp To send each log line as text data to a TCP receiver.
3147 Place: The destination host name or IP and port.
3148 Place Format: \\host:port
3151 access_log daemon:@DEFAULT_ACCESS_LOG@ squid
3157 LOC: Config.Log.icaplogs
3160 ICAP log files record ICAP transaction summaries, one line per
3163 The icap_log option format is:
3164 icap_log <filepath> [<logformat name> [acl acl ...]]
3165 icap_log none [acl acl ...]]
3167 Please see access_log option documentation for details. The two
3168 kinds of logs share the overall configuration approach and many
3171 ICAP processing of a single HTTP message or transaction may
3172 require multiple ICAP transactions. In such cases, multiple
3173 ICAP transaction log lines will correspond to a single access
3176 ICAP log uses logformat codes that make sense for an ICAP
3177 transaction. Header-related codes are applied to the HTTP header
3178 embedded in an ICAP server response, with the following caveats:
3179 For REQMOD, there is no HTTP response header unless the ICAP
3180 server performed request satisfaction. For RESPMOD, the HTTP
3181 request header is the header sent to the ICAP server. For
3182 OPTIONS, there are no HTTP headers.
3184 The following format codes are also available for ICAP logs:
3186 icap::<A ICAP server IP address. Similar to <A.
3188 icap::<service_name ICAP service name from the icap_service
3189 option in Squid configuration file.
3191 icap::ru ICAP Request-URI. Similar to ru.
3193 icap::rm ICAP request method (REQMOD, RESPMOD, or
3194 OPTIONS). Similar to existing rm.
3196 icap::>st Bytes sent to the ICAP server (TCP payload
3197 only; i.e., what Squid writes to the socket).
3199 icap::<st Bytes received from the ICAP server (TCP
3200 payload only; i.e., what Squid reads from
3203 icap::<bs Number of message body bytes received from the
3204 ICAP server. ICAP message body, if any, usually
3205 includes encapsulated HTTP message headers and
3206 possibly encapsulated HTTP message body. The
3207 HTTP body part is dechunked before its size is
3210 icap::tr Transaction response time (in
3211 milliseconds). The timer starts when
3212 the ICAP transaction is created and
3213 stops when the transaction is completed.
3216 icap::tio Transaction I/O time (in milliseconds). The
3217 timer starts when the first ICAP request
3218 byte is scheduled for sending. The timers
3219 stops when the last byte of the ICAP response
3222 icap::to Transaction outcome: ICAP_ERR* for all
3223 transaction errors, ICAP_OPT for OPTION
3224 transactions, ICAP_ECHO for 204
3225 responses, ICAP_MOD for message
3226 modification, and ICAP_SAT for request
3227 satisfaction. Similar to Ss.
3229 icap::Hs ICAP response status code. Similar to Hs.
3231 icap::>h ICAP request header(s). Similar to >h.
3233 icap::<h ICAP response header(s). Similar to <h.
3235 The default ICAP log format, which can be used without an explicit
3236 definition, is called icap_squid:
3238 logformat icap_squid %ts.%03tu %6icap::tr %>a %icap::to/%03icap::Hs %icap::<size %icap::rm %icap::ru% %un -/%icap::<A -
3240 See also: logformat, log_icap, and %adapt::<last_h
3243 NAME: logfile_daemon
3245 DEFAULT: @DEFAULT_LOGFILED@
3246 LOC: Log::TheConfig.logfile_daemon
3248 Specify the path to the logfile-writing daemon. This daemon is
3249 used to write the access and store logs, if configured.
3251 Squid sends a number of commands to the log daemon:
3252 L<data>\n - logfile data
3257 r<n>\n - set rotate count to <n>
3258 b<n>\n - 1 = buffer output, 0 = don't buffer output
3260 No responses is expected.
3265 LOC: Config.accessList.log
3267 COMMENT: allow|deny acl acl...
3269 This options allows you to control which requests gets logged
3270 to access.log (see access_log directive). Requests denied for
3271 logging will also not be accounted for in performance counters.
3273 This clause only supports fast acl types.
3274 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
3280 LOC: Config.accessList.icap
3283 This options allows you to control which requests get logged
3284 to icap.log. See the icap_log directive for ICAP log details.
3287 NAME: cache_store_log
3290 LOC: Config.Log.store
3292 Logs the activities of the storage manager. Shows which
3293 objects are ejected from the cache, and which objects are
3294 saved and for how long. To disable, enter "none" or remove the line.
3295 There are not really utilities to analyze this data, so you can safely
3299 cache_store_log @DEFAULT_STORE_LOG@
3302 NAME: cache_swap_state cache_swap_log
3304 LOC: Config.Log.swap
3307 Location for the cache "swap.state" file. This index file holds
3308 the metadata of objects saved on disk. It is used to rebuild
3309 the cache during startup. Normally this file resides in each
3310 'cache_dir' directory, but you may specify an alternate
3311 pathname here. Note you must give a full filename, not just
3312 a directory. Since this is the index for the whole object
3313 list you CANNOT periodically rotate it!
3315 If %s can be used in the file name it will be replaced with a
3316 a representation of the cache_dir name where each / is replaced
3317 with '.'. This is needed to allow adding/removing cache_dir
3318 lines when cache_swap_log is being used.
3320 If have more than one 'cache_dir', and %s is not used in the name
3321 these swap logs will have names such as:
3327 The numbered extension (which is added automatically)
3328 corresponds to the order of the 'cache_dir' lines in this
3329 configuration file. If you change the order of the 'cache_dir'
3330 lines in this file, these index files will NOT correspond to
3331 the correct 'cache_dir' entry (unless you manually rename
3332 them). We recommend you do NOT use this option. It is
3333 better to keep these index files in each 'cache_dir' directory.
3336 NAME: logfile_rotate
3339 LOC: Config.Log.rotateNumber
3341 Specifies the number of logfile rotations to make when you
3342 type 'squid -k rotate'. The default is 10, which will rotate
3343 with extensions 0 through 9. Setting logfile_rotate to 0 will
3344 disable the file name rotation, but the logfiles are still closed
3345 and re-opened. This will enable you to rename the logfiles
3346 yourself just before sending the rotate signal.
3348 Note, the 'squid -k rotate' command normally sends a USR1
3349 signal to the running squid process. In certain situations
3350 (e.g. on Linux with Async I/O), USR1 is used for other
3351 purposes, so -k rotate uses another signal. It is best to get
3352 in the habit of using 'squid -k rotate' instead of 'kill -USR1
3355 Note, from Squid-3.1 this option has no effect on the cache.log,
3356 that log can be rotated separately by using debug_options
3359 NAME: emulate_httpd_log
3362 Replace this with an access_log directive using the format 'common' or 'combined'.
3365 NAME: log_ip_on_direct
3368 Remove this option from your config. To log server or peer names use %<A in the log format.
3373 DEFAULT: @DEFAULT_MIME_TABLE@
3374 LOC: Config.mimeTablePathname
3376 Pathname to Squid's MIME table. You shouldn't need to change
3377 this, but the default file contains examples and formatting
3378 information if you do.
3384 LOC: Config.onoff.log_mime_hdrs
3387 The Cache can record both the request and the response MIME
3388 headers for each HTTP transaction. The headers are encoded
3389 safely and will appear as two bracketed fields at the end of
3390 the access log (for either the native or httpd-emulated log
3391 formats). To enable this logging set log_mime_hdrs to 'on'.
3397 Replace this with an access_log directive using the format 'useragent'.
3400 NAME: referer_log referrer_log
3403 Replace this with an access_log directive using the format 'referrer'.
3408 DEFAULT: @DEFAULT_PID_FILE@
3409 LOC: Config.pidFilename
3411 A filename to write the process-id to. To disable, enter "none".
3417 Remove this option from your config. To log FQDN use %>A in the log format.
3420 NAME: client_netmask
3422 LOC: Config.Addrs.client_netmask
3425 A netmask for client addresses in logfiles and cachemgr output.
3426 Change this to protect the privacy of your cache clients.
3427 A netmask of 255.255.255.0 will log all IP's in that range with
3428 the last digit set to '0'.
3434 Use a regular access.log with ACL limiting it to MISS events.
3437 NAME: strip_query_terms
3439 LOC: Config.onoff.strip_query_terms
3442 By default, Squid strips query terms from requested URLs before
3443 logging. This protects your user's privacy.
3450 LOC: Config.onoff.buffered_logs
3452 cache.log log file is written with stdio functions, and as such
3453 it can be buffered or unbuffered. By default it will be unbuffered.
3454 Buffering it can speed up the writing slightly (though you are
3455 unlikely to need to worry unless you run with tons of debugging
3456 enabled in which case performance will suffer badly anyway..).
3459 NAME: netdb_filename
3461 DEFAULT: stdio:@DEFAULT_NETDB_FILE@
3462 LOC: Config.netdbFilename
3465 A filename where Squid stores it's netdb state between restarts.
3466 To disable, enter "none".
3470 OPTIONS FOR TROUBLESHOOTING
3471 -----------------------------------------------------------------------------
3476 DEFAULT_IF_NONE: @DEFAULT_CACHE_LOG@
3477 LOC: Debug::cache_log
3479 Cache logging file. This is where general information about
3480 your cache's behavior goes. You can increase the amount of data
3481 logged to this file and how often its rotated with "debug_options"
3487 LOC: Debug::debugOptions
3489 Logging options are set as section,level where each source file
3490 is assigned a unique section. Lower levels result in less
3491 output, Full debugging (level 9) can result in a very large
3492 log file, so be careful.
3494 The magic word "ALL" sets debugging levels for all sections.
3495 We recommend normally running with "ALL,1".
3497 The rotate=N option can be used to keep more or less of these logs
3498 than would otherwise be kept by logfile_rotate.
3499 For most uses a single log should be enough to monitor current
3500 events affecting Squid.
3505 LOC: Config.coredump_dir
3506 DEFAULT_IF_NONE: none
3508 By default Squid leaves core files in the directory from where
3509 it was started. If you set 'coredump_dir' to a directory
3510 that exists, Squid will chdir() to that directory at startup
3511 and coredump files will be left there.
3515 # Leave coredumps in the first cache dir
3516 coredump_dir @DEFAULT_SWAP_DIR@
3522 OPTIONS FOR FTP GATEWAYING
3523 -----------------------------------------------------------------------------
3529 LOC: Config.Ftp.anon_user
3531 If you want the anonymous login password to be more informative
3532 (and enable the use of picky ftp servers), set this to something
3533 reasonable for your domain, like wwwuser@somewhere.net
3535 The reason why this is domainless by default is the
3536 request can be made on the behalf of a user in any domain,
3537 depending on how the cache is used.
3538 Some ftp server also validate the email address is valid
3539 (for example perl.com).
3545 LOC: Config.Ftp.passive
3547 If your firewall does not allow Squid to use passive
3548 connections, turn off this option.
3550 Use of ftp_epsv_all option requires this to be ON.
3556 LOC: Config.Ftp.epsv_all
3558 FTP Protocol extensions permit the use of a special "EPSV ALL" command.
3560 NATs may be able to put the connection on a "fast path" through the
3561 translator, as the EPRT command will never be used and therefore,
3562 translation of the data portion of the segments will never be needed.
3564 When a client only expects to do two-way FTP transfers this may be
3566 If squid finds that it must do a three-way FTP transfer after issuing
3567 an EPSV ALL command, the FTP session will fail.
3569 If you have any doubts about this option do not use it.
3570 Squid will nicely attempt all other connection methods.
3572 Requires ftp_passive to be ON (default) for any effect.
3578 LOC: Config.Ftp.epsv
3580 FTP Protocol extensions permit the use of a special "EPSV" command.
3582 NATs may be able to put the connection on a "fast path" through the
3583 translator using EPSV, as the EPRT command will never be used
3584 and therefore, translation of the data portion of the segments
3585 will never be needed.
3587 Turning this OFF will prevent EPSV being attempted.
3588 WARNING: Doing so will convert Squid back to the old behavior with all
3589 the related problems with external NAT devices/layers.
3591 Requires ftp_passive to be ON (default) for any effect.
3597 LOC: Config.Ftp.eprt
3599 FTP Protocol extensions permit the use of a special "EPRT" command.
3601 This extension provides a protocol neutral alternative to the
3602 IPv4-only PORT command. When supported it enables active FTP data
3603 channels over IPv6 and efficient NAT handling.
3605 Turning this OFF will prevent EPRT being attempted and will skip
3606 straight to using PORT for IPv4 servers.
3608 Some devices are known to not handle this extension correctly and
3609 may result in crashes. Devices which suport EPRT enough to fail
3610 cleanly will result in Squid attempting PORT anyway. This directive
3611 should only be disabled when EPRT results in device failures.
3613 WARNING: Doing so will convert Squid back to the old behavior with all
3614 the related problems with external NAT devices/layers and IPv4-only FTP.
3617 NAME: ftp_sanitycheck
3620 LOC: Config.Ftp.sanitycheck
3622 For security and data integrity reasons Squid by default performs
3623 sanity checks of the addresses of FTP data connections ensure the
3624 data connection is to the requested server. If you need to allow
3625 FTP connections to servers using another IP address for the data
3626 connection turn this off.
3629 NAME: ftp_telnet_protocol
3632 LOC: Config.Ftp.telnet
3634 The FTP protocol is officially defined to use the telnet protocol
3635 as transport channel for the control connection. However, many
3636 implementations are broken and does not respect this aspect of
3639 If you have trouble accessing files with ASCII code 255 in the
3640 path or similar problems involving this ASCII code you can
3641 try setting this directive to off. If that helps, report to the
3642 operator of the FTP server in question that their FTP server
3643 is broken and does not follow the FTP standard.
3647 OPTIONS FOR EXTERNAL SUPPORT PROGRAMS
3648 -----------------------------------------------------------------------------
3653 DEFAULT: @DEFAULT_DISKD@
3654 LOC: Config.Program.diskd
3656 Specify the location of the diskd executable.
3657 Note this is only useful if you have compiled in
3658 diskd as one of the store io modules.
3661 NAME: unlinkd_program
3664 DEFAULT: @DEFAULT_UNLINKD@
3665 LOC: Config.Program.unlinkd
3667 Specify the location of the executable for file deletion process.
3670 NAME: pinger_program
3672 DEFAULT: @DEFAULT_PINGER@
3673 LOC: Config.pinger.program
3676 Specify the location of the executable for the pinger process.
3682 LOC: Config.pinger.enable
3685 Control whether the pinger is active at run-time.
3686 Enables turning ICMP pinger on and off with a simple
3687 squid -k reconfigure.
3692 OPTIONS FOR URL REWRITING
3693 -----------------------------------------------------------------------------
3696 NAME: url_rewrite_program redirect_program
3698 LOC: Config.Program.redirect
3701 Specify the location of the executable URL rewriter to use.
3702 Since they can perform almost any function there isn't one included.
3704 For each requested URL, the rewriter will receive on line with the format
3706 URL <SP> client_ip "/" fqdn <SP> user <SP> method [<SP> kvpairs]<NL>
3708 In the future, the rewriter interface will be extended with
3709 key=value pairs ("kvpairs" shown above). Rewriter programs
3710 should be prepared to receive and possibly ignore additional
3711 whitespace-separated tokens on each input line.
3713 And the rewriter may return a rewritten URL. The other components of
3714 the request line does not need to be returned (ignored if they are).
3716 The rewriter can also indicate that a client-side redirect should
3717 be performed to the new URL. This is done by prefixing the returned
3718 URL with "301:" (moved permanently) or 302: (moved temporarily), etc.
3720 By default, a URL rewriter is not used.
3723 NAME: url_rewrite_children redirect_children
3724 TYPE: HelperChildConfig
3725 DEFAULT: 20 startup=0 idle=1 concurrency=0
3726 LOC: Config.redirectChildren
3728 The maximum number of redirector processes to spawn. If you limit
3729 it too few Squid will have to wait for them to process a backlog of
3730 URLs, slowing it down. If you allow too many they will use RAM
3731 and other system resources noticably.
3733 The startup= and idle= options allow some measure of skew in your
3738 Sets a minimum of how many processes are to be spawned when Squid
3739 starts or reconfigures. When set to zero the first request will
3740 cause spawning of the first child process to handle it.
3742 Starting too few will cause an initial slowdown in traffic as Squid
3743 attempts to simultaneously spawn enough processes to cope.
3747 Sets a minimum of how many processes Squid is to try and keep available
3748 at all times. When traffic begins to rise above what the existing
3749 processes can handle this many more will be spawned up to the maximum
3750 configured. A minimum setting of 1 is required.
3754 The number of requests each redirector helper can handle in
3755 parallel. Defaults to 0 which indicates the redirector
3756 is a old-style single threaded redirector.
3758 When this directive is set to a value >= 1 then the protocol
3759 used to communicate with the helper is modified to include
3760 a request ID in front of the request/response. The request
3761 ID from the request must be echoed back with the response
3765 NAME: url_rewrite_host_header redirect_rewrites_host_header
3768 LOC: Config.onoff.redir_rewrites_host
3770 To preserve same-origin security policies in browsers and
3771 prevent Host: header forgery by redirectors Squid rewrites
3772 any Host: header in redirected requests.
3774 If you are running an accelerator this may not be a wanted
3775 effect of a redirector. This directive enables you disable
3776 Host: alteration in reverse-proxy traffic.
3778 WARNING: Entries are cached on the result of the URL rewriting
3779 process, so be careful if you have domain-virtual hosts.
3781 WARNING: Squid and other software verifies the URL and Host
3782 are matching, so be careful not to relay through other proxies
3783 or inspecting firewalls with this disabled.
3786 NAME: url_rewrite_access redirector_access
3789 LOC: Config.accessList.redirector
3791 If defined, this access list specifies which requests are
3792 sent to the redirector processes. By default all requests
3795 This clause supports both fast and slow acl types.
3796 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
3799 NAME: url_rewrite_bypass redirector_bypass
3801 LOC: Config.onoff.redirector_bypass
3804 When this is 'on', a request will not go through the
3805 redirector if all redirectors are busy. If this is 'off'
3806 and the redirector queue grows too large, Squid will exit
3807 with a FATAL error and ask you to increase the number of
3808 redirectors. You should only enable this if the redirectors
3809 are not critical to your caching system. If you use
3810 redirectors for access control, and you enable this option,
3811 users may have access to pages they should not
3812 be allowed to request.
3816 OPTIONS FOR TUNING THE CACHE
3817 -----------------------------------------------------------------------------
3820 NAME: cache no_cache
3823 LOC: Config.accessList.noCache
3825 A list of ACL elements which, if matched and denied, cause the request to
3826 not be satisfied from the cache and the reply to not be cached.
3827 In other words, use this to force certain objects to never be cached.
3829 You must use the words 'allow' or 'deny' to indicate whether items
3830 matching the ACL should be allowed or denied into the cache.
3832 Default is to allow all to be cached.
3834 This clause supports both fast and slow acl types.
3835 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
3841 LOC: Config.maxStale
3844 This option puts an upper limit on how stale content Squid
3845 will serve from the cache if cache validation fails.
3846 Can be overriden by the refresh_pattern max-stale option.
3849 NAME: refresh_pattern
3850 TYPE: refreshpattern
3854 usage: refresh_pattern [-i] regex min percent max [options]
3856 By default, regular expressions are CASE-SENSITIVE. To make
3857 them case-insensitive, use the -i option.
3859 'Min' is the time (in minutes) an object without an explicit
3860 expiry time should be considered fresh. The recommended
3861 value is 0, any higher values may cause dynamic applications
3862 to be erroneously cached unless the application designer
3863 has taken the appropriate actions.
3865 'Percent' is a percentage of the objects age (time since last
3866 modification age) an object without explicit expiry time
3867 will be considered fresh.
3869 'Max' is an upper limit on how long objects without an explicit
3870 expiry time will be considered fresh.
3872 options: override-expire
3878 ignore-must-revalidate
3885 override-expire enforces min age even if the server
3886 sent an explicit expiry time (e.g., with the
3887 Expires: header or Cache-Control: max-age). Doing this
3888 VIOLATES the HTTP standard. Enabling this feature
3889 could make you liable for problems which it causes.
3891 Note: override-expire does not enforce staleness - it only extends
3892 freshness / min. If the server returns a Expires time which
3893 is longer than your max time, Squid will still consider
3894 the object fresh for that period of time.
3896 override-lastmod enforces min age even on objects
3897 that were modified recently.
3899 reload-into-ims changes client no-cache or ``reload''
3900 to If-Modified-Since requests. Doing this VIOLATES the
3901 HTTP standard. Enabling this feature could make you
3902 liable for problems which it causes.
3904 ignore-reload ignores a client no-cache or ``reload''
3905 header. Doing this VIOLATES the HTTP standard. Enabling
3906 this feature could make you liable for problems which
3909 ignore-no-cache ignores any ``Pragma: no-cache'' and
3910 ``Cache-control: no-cache'' headers received from a server.
3911 The HTTP RFC never allows the use of this (Pragma) header
3912 from a server, only a client, though plenty of servers
3915 ignore-no-store ignores any ``Cache-control: no-store''
3916 headers received from a server. Doing this VIOLATES
3917 the HTTP standard. Enabling this feature could make you
3918 liable for problems which it causes.
3920 ignore-must-revalidate ignores any ``Cache-Control: must-revalidate``
3921 headers received from a server. Doing this VIOLATES
3922 the HTTP standard. Enabling this feature could make you
3923 liable for problems which it causes.
3925 ignore-private ignores any ``Cache-control: private''
3926 headers received from a server. Doing this VIOLATES
3927 the HTTP standard. Enabling this feature could make you
3928 liable for problems which it causes.
3930 ignore-auth caches responses to requests with authorization,
3931 as if the originserver had sent ``Cache-control: public''
3932 in the response header. Doing this VIOLATES the HTTP standard.
3933 Enabling this feature could make you liable for problems which
3936 refresh-ims causes squid to contact the origin server
3937 when a client issues an If-Modified-Since request. This
3938 ensures that the client will receive an updated version
3939 if one is available.
3941 store-stale stores responses even if they don't have explicit
3942 freshness or a validator (i.e., Last-Modified or an ETag)
3943 present, or if they're already stale. By default, Squid will
3944 not cache such responses because they usually can't be
3945 reused. Note that such responses will be stale by default.
3947 max-stale=NN provide a maximum staleness factor. Squid won't
3948 serve objects more stale than this even if it failed to
3949 validate the object. Default: use the max_stale global limit.
3951 Basically a cached object is:
3953 FRESH if expires < now, else STALE
3955 FRESH if lm-factor < percent, else STALE
3959 The refresh_pattern lines are checked in the order listed here.
3960 The first entry which matches is used. If none of the entries
3961 match the default will be used.
3963 Note, you must uncomment all the default lines if you want
3964 to change one. The default setting is only active if none is
3969 # Add any of your own refresh_pattern entries above these.
3970 refresh_pattern ^ftp: 1440 20% 10080
3971 refresh_pattern ^gopher: 1440 0% 1440
3972 refresh_pattern -i (/cgi-bin/|\?) 0 0% 0
3973 refresh_pattern . 0 20% 4320
3977 NAME: quick_abort_min
3981 LOC: Config.quickAbort.min
3984 NAME: quick_abort_max
3988 LOC: Config.quickAbort.max
3991 NAME: quick_abort_pct
3995 LOC: Config.quickAbort.pct
3997 The cache by default continues downloading aborted requests
3998 which are almost completed (less than 16 KB remaining). This
3999 may be undesirable on slow (e.g. SLIP) links and/or very busy
4000 caches. Impatient users may tie up file descriptors and
4001 bandwidth by repeatedly requesting and immediately aborting
4004 When the user aborts a request, Squid will check the
4005 quick_abort values to the amount of data transfered until
4008 If the transfer has less than 'quick_abort_min' KB remaining,
4009 it will finish the retrieval.
4011 If the transfer has more than 'quick_abort_max' KB remaining,
4012 it will abort the retrieval.
4014 If more than 'quick_abort_pct' of the transfer has completed,
4015 it will finish the retrieval.
4017 If you do not want any retrieval to continue after the client
4018 has aborted, set both 'quick_abort_min' and 'quick_abort_max'
4021 If you want retrievals to always continue if they are being
4022 cached set 'quick_abort_min' to '-1 KB'.
4025 NAME: read_ahead_gap
4026 COMMENT: buffer-size
4028 LOC: Config.readAheadGap
4031 The amount of data the cache will buffer ahead of what has been
4032 sent to the client when retrieving an object from another server.
4036 IFDEF: USE_HTTP_VIOLATIONS
4039 LOC: Config.negativeTtl
4042 Set the Default Time-to-Live (TTL) for failed requests.
4043 Certain types of failures (such as "connection refused" and
4044 "404 Not Found") are able to be negatively-cached for a short time.
4045 Modern web servers should provide Expires: header, however if they
4046 do not this can provide a minimum TTL.
4047 The default is not to cache errors with unknown expiry details.
4049 Note that this is different from negative caching of DNS lookups.
4051 WARNING: Doing this VIOLATES the HTTP standard. Enabling
4052 this feature could make you liable for problems which it
4056 NAME: positive_dns_ttl
4059 LOC: Config.positiveDnsTtl
4062 Upper limit on how long Squid will cache positive DNS responses.
4063 Default is 6 hours (360 minutes). This directive must be set
4064 larger than negative_dns_ttl.
4067 NAME: negative_dns_ttl
4070 LOC: Config.negativeDnsTtl
4073 Time-to-Live (TTL) for negative caching of failed DNS lookups.
4074 This also sets the lower cache limit on positive lookups.
4075 Minimum value is 1 second, and it is not recommendable to go
4076 much below 10 seconds.
4079 NAME: range_offset_limit
4080 COMMENT: size [acl acl...]
4082 LOC: Config.rangeOffsetLimit
4085 usage: (size) [units] [[!]aclname]
4087 Sets an upper limit on how far (number of bytes) into the file
4088 a Range request may be to cause Squid to prefetch the whole file.
4089 If beyond this limit, Squid forwards the Range request as it is and
4090 the result is NOT cached.
4092 This is to stop a far ahead range request (lets say start at 17MB)
4093 from making Squid fetch the whole object up to that point before
4094 sending anything to the client.
4096 Multiple range_offset_limit lines may be specified, and they will
4097 be searched from top to bottom on each request until a match is found.
4098 The first match found will be used. If no line matches a request, the
4099 default limit of 0 bytes will be used.
4101 'size' is the limit specified as a number of units.
4103 'units' specifies whether to use bytes, KB, MB, etc.
4104 If no units are specified bytes are assumed.
4106 A size of 0 causes Squid to never fetch more than the
4107 client requested. (default)
4109 A size of 'none' causes Squid to always fetch the object from the
4110 beginning so it may cache the result. (2.0 style)
4112 'aclname' is the name of a defined ACL.
4114 NP: Using 'none' as the byte value here will override any quick_abort settings
4115 that may otherwise apply to the range request. The range request will
4116 be fully fetched from start to finish regardless of the client
4117 actions. This affects bandwidth usage.
4120 NAME: minimum_expiry_time
4123 LOC: Config.minimum_expiry_time
4126 The minimum caching time according to (Expires - Date)
4127 Headers Squid honors if the object can't be revalidated
4128 defaults to 60 seconds. In reverse proxy environments it
4129 might be desirable to honor shorter object lifetimes. It
4130 is most likely better to make your server return a
4131 meaningful Last-Modified header however. In ESI environments
4132 where page fragments often have short lifetimes, this will
4133 often be best set to 0.
4136 NAME: store_avg_object_size
4140 LOC: Config.Store.avgObjectSize
4142 Average object size, used to estimate number of objects your
4143 cache can hold. The default is 13 KB.
4146 NAME: store_objects_per_bucket
4149 LOC: Config.Store.objectsPerBucket
4151 Target number of objects per bucket in the store hash table.
4152 Lowering this value increases the total number of buckets and
4153 also the storage maintenance rate. The default is 20.
4158 -----------------------------------------------------------------------------
4161 NAME: request_header_max_size
4165 LOC: Config.maxRequestHeaderSize
4167 This specifies the maximum size for HTTP headers in a request.
4168 Request headers are usually relatively small (about 512 bytes).
4169 Placing a limit on the request header size will catch certain
4170 bugs (for example with persistent connections) and possibly
4171 buffer-overflow or denial-of-service attacks.
4174 NAME: reply_header_max_size
4178 LOC: Config.maxReplyHeaderSize
4180 This specifies the maximum size for HTTP headers in a reply.
4181 Reply headers are usually relatively small (about 512 bytes).
4182 Placing a limit on the reply header size will catch certain
4183 bugs (for example with persistent connections) and possibly
4184 buffer-overflow or denial-of-service attacks.
4187 NAME: request_body_max_size
4191 LOC: Config.maxRequestBodySize
4193 This specifies the maximum size for an HTTP request body.
4194 In other words, the maximum size of a PUT/POST request.
4195 A user who attempts to send a request with a body larger
4196 than this limit receives an "Invalid Request" error message.
4197 If you set this parameter to a zero (the default), there will
4198 be no limit imposed.
4201 NAME: client_request_buffer_max_size
4205 LOC: Config.maxRequestBufferSize
4207 This specifies the maximum buffer size of a client request.
4208 It prevents squid eating too much memory when somebody uploads
4212 NAME: chunked_request_body_max_size
4216 LOC: Config.maxChunkedRequestBodySize
4218 A broken or confused HTTP/1.1 client may send a chunked HTTP
4219 request to Squid. Squid does not have full support for that
4220 feature yet. To cope with such requests, Squid buffers the
4221 entire request and then dechunks request body to create a
4222 plain HTTP/1.0 request with a known content length. The plain
4223 request is then used by the rest of Squid code as usual.
4225 The option value specifies the maximum size of the buffer used
4226 to hold the request before the conversion. If the chunked
4227 request size exceeds the specified limit, the conversion
4228 fails, and the client receives an "unsupported request" error,
4229 as if dechunking was disabled.
4231 Dechunking is enabled by default. To disable conversion of
4232 chunked requests, set the maximum to zero.
4234 Request dechunking feature and this option in particular are a
4235 temporary hack. When chunking requests and responses are fully
4236 supported, there will be no need to buffer a chunked request.
4240 IFDEF: USE_HTTP_VIOLATIONS
4243 LOC: Config.accessList.brokenPosts
4245 A list of ACL elements which, if matched, causes Squid to send
4246 an extra CRLF pair after the body of a PUT/POST request.
4248 Some HTTP servers has broken implementations of PUT/POST,
4249 and rely on an extra CRLF pair sent by some WWW clients.
4251 Quote from RFC2616 section 4.1 on this matter:
4253 Note: certain buggy HTTP/1.0 client implementations generate an
4254 extra CRLF's after a POST request. To restate what is explicitly
4255 forbidden by the BNF, an HTTP/1.1 client must not preface or follow
4256 a request with an extra CRLF.
4258 This clause only supports fast acl types.
4259 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
4262 acl buggy_server url_regex ^http://....
4263 broken_posts allow buggy_server
4266 NAME: adaptation_uses_indirect_client icap_uses_indirect_client
4269 IFDEF: FOLLOW_X_FORWARDED_FOR&&USE_ADAPTATION
4271 LOC: Adaptation::Config::use_indirect_client
4273 Controls whether the indirect client IP address (instead of the direct
4274 client IP address) is passed to adaptation services.
4276 See also: follow_x_forwarded_for adaptation_send_client_ip
4280 IFDEF: USE_HTTP_VIOLATIONS
4284 LOC: Config.onoff.via
4286 If set (default), Squid will include a Via header in requests and
4287 replies as required by RFC2616.
4293 LOC: Config.onoff.ie_refresh
4296 Microsoft Internet Explorer up until version 5.5 Service
4297 Pack 1 has an issue with transparent proxies, wherein it
4298 is impossible to force a refresh. Turning this on provides
4299 a partial fix to the problem, by causing all IMS-REFRESH
4300 requests from older IE versions to check the origin server
4301 for fresh content. This reduces hit ratio by some amount
4302 (~10% in my experience), but allows users to actually get
4303 fresh content when they want it. Note because Squid
4304 cannot tell if the user is using 5.5 or 5.5SP1, the behavior
4305 of 5.5 is unchanged from old versions of Squid (i.e. a
4306 forced refresh is impossible). Newer versions of IE will,
4307 hopefully, continue to have the new behavior and will be
4308 handled based on that assumption. This option defaults to
4309 the old Squid behavior, which is better for hit ratios but
4310 worse for clients using IE, if they need to be able to
4311 force fresh content.
4314 NAME: vary_ignore_expire
4317 LOC: Config.onoff.vary_ignore_expire
4320 Many HTTP servers supporting Vary gives such objects
4321 immediate expiry time with no cache-control header
4322 when requested by a HTTP/1.0 client. This option
4323 enables Squid to ignore such expiry times until
4324 HTTP/1.1 is fully implemented.
4326 WARNING: If turned on this may eventually cause some
4327 varying objects not intended for caching to get cached.
4330 NAME: request_entities
4332 LOC: Config.onoff.request_entities
4335 Squid defaults to deny GET and HEAD requests with request entities,
4336 as the meaning of such requests are undefined in the HTTP standard
4337 even if not explicitly forbidden.
4339 Set this directive to on if you have clients which insists
4340 on sending request entities in GET or HEAD requests. But be warned
4341 that there is server software (both proxies and web servers) which
4342 can fail to properly process this kind of request which may make you
4343 vulnerable to cache pollution attacks if enabled.
4346 NAME: request_header_access
4347 IFDEF: USE_HTTP_VIOLATIONS
4348 TYPE: http_header_access[]
4349 LOC: Config.request_header_access
4352 Usage: request_header_access header_name allow|deny [!]aclname ...
4354 WARNING: Doing this VIOLATES the HTTP standard. Enabling
4355 this feature could make you liable for problems which it
4358 This option replaces the old 'anonymize_headers' and the
4359 older 'http_anonymizer' option with something that is much
4360 more configurable. This new method creates a list of ACLs
4361 for each header, allowing you very fine-tuned header
4364 This option only applies to request headers, i.e., from the
4365 client to the server.
4367 You can only specify known headers for the header name.
4368 Other headers are reclassified as 'Other'. You can also
4369 refer to all the headers with 'All'.
4371 For example, to achieve the same behavior as the old
4372 'http_anonymizer standard' option, you should use:
4374 request_header_access From deny all
4375 request_header_access Referer deny all
4376 request_header_access Server deny all
4377 request_header_access User-Agent deny all
4378 request_header_access WWW-Authenticate deny all
4379 request_header_access Link deny all
4381 Or, to reproduce the old 'http_anonymizer paranoid' feature
4384 request_header_access Allow allow all
4385 request_header_access Authorization allow all
4386 request_header_access WWW-Authenticate allow all
4387 request_header_access Proxy-Authorization allow all
4388 request_header_access Proxy-Authenticate allow all
4389 request_header_access Cache-Control allow all
4390 request_header_access Content-Encoding allow all
4391 request_header_access Content-Length allow all
4392 request_header_access Content-Type allow all
4393 request_header_access Date allow all
4394 request_header_access Expires allow all
4395 request_header_access Host allow all
4396 request_header_access If-Modified-Since allow all
4397 request_header_access Last-Modified allow all
4398 request_header_access Location allow all
4399 request_header_access Pragma allow all
4400 request_header_access Accept allow all
4401 request_header_access Accept-Charset allow all
4402 request_header_access Accept-Encoding allow all
4403 request_header_access Accept-Language allow all
4404 request_header_access Content-Language allow all
4405 request_header_access Mime-Version allow all
4406 request_header_access Retry-After allow all
4407 request_header_access Title allow all
4408 request_header_access Connection allow all
4409 request_header_access All deny all
4411 although many of those are HTTP reply headers, and so should be
4412 controlled with the reply_header_access directive.
4414 By default, all headers are allowed (no anonymizing is
4418 NAME: reply_header_access
4419 IFDEF: USE_HTTP_VIOLATIONS
4420 TYPE: http_header_access[]
4421 LOC: Config.reply_header_access
4424 Usage: reply_header_access header_name allow|deny [!]aclname ...
4426 WARNING: Doing this VIOLATES the HTTP standard. Enabling
4427 this feature could make you liable for problems which it
4430 This option only applies to reply headers, i.e., from the
4431 server to the client.
4433 This is the same as request_header_access, but in the other
4436 This option replaces the old 'anonymize_headers' and the
4437 older 'http_anonymizer' option with something that is much
4438 more configurable. This new method creates a list of ACLs
4439 for each header, allowing you very fine-tuned header
4442 You can only specify known headers for the header name.
4443 Other headers are reclassified as 'Other'. You can also
4444 refer to all the headers with 'All'.
4446 For example, to achieve the same behavior as the old
4447 'http_anonymizer standard' option, you should use:
4449 reply_header_access From deny all
4450 reply_header_access Referer deny all
4451 reply_header_access Server deny all
4452 reply_header_access User-Agent deny all
4453 reply_header_access WWW-Authenticate deny all
4454 reply_header_access Link deny all
4456 Or, to reproduce the old 'http_anonymizer paranoid' feature
4459 reply_header_access Allow allow all
4460 reply_header_access Authorization allow all
4461 reply_header_access WWW-Authenticate allow all
4462 reply_header_access Proxy-Authorization allow all
4463 reply_header_access Proxy-Authenticate allow all
4464 reply_header_access Cache-Control allow all
4465 reply_header_access Content-Encoding allow all
4466 reply_header_access Content-Length allow all
4467 reply_header_access Content-Type allow all
4468 reply_header_access Date allow all
4469 reply_header_access Expires allow all
4470 reply_header_access Host allow all
4471 reply_header_access If-Modified-Since allow all
4472 reply_header_access Last-Modified allow all
4473 reply_header_access Location allow all
4474 reply_header_access Pragma allow all
4475 reply_header_access Accept allow all
4476 reply_header_access Accept-Charset allow all
4477 reply_header_access Accept-Encoding allow all
4478 reply_header_access Accept-Language allow all
4479 reply_header_access Content-Language allow all
4480 reply_header_access Mime-Version allow all
4481 reply_header_access Retry-After allow all
4482 reply_header_access Title allow all
4483 reply_header_access Connection allow all
4484 reply_header_access All deny all
4486 although the HTTP request headers won't be usefully controlled
4487 by this directive -- see request_header_access for details.
4489 By default, all headers are allowed (no anonymizing is
4493 NAME: request_header_replace header_replace
4494 IFDEF: USE_HTTP_VIOLATIONS
4495 TYPE: http_header_replace[]
4496 LOC: Config.request_header_access
4499 Usage: request_header_replace header_name message
4500 Example: request_header_replace User-Agent Nutscrape/1.0 (CP/M; 8-bit)
4502 This option allows you to change the contents of headers
4503 denied with request_header_access above, by replacing them
4504 with some fixed string. This replaces the old fake_user_agent
4507 This only applies to request headers, not reply headers.
4509 By default, headers are removed if denied.
4512 NAME: reply_header_replace
4513 IFDEF: USE_HTTP_VIOLATIONS
4514 TYPE: http_header_replace[]
4515 LOC: Config.reply_header_access
4518 Usage: reply_header_replace header_name message
4519 Example: reply_header_replace Server Foo/1.0
4521 This option allows you to change the contents of headers
4522 denied with reply_header_access above, by replacing them
4523 with some fixed string.
4525 This only applies to reply headers, not request headers.
4527 By default, headers are removed if denied.
4530 NAME: relaxed_header_parser
4531 COMMENT: on|off|warn
4533 LOC: Config.onoff.relaxed_header_parser
4536 In the default "on" setting Squid accepts certain forms
4537 of non-compliant HTTP messages where it is unambiguous
4538 what the sending application intended even if the message
4539 is not correctly formatted. The messages is then normalized
4540 to the correct form when forwarded by Squid.
4542 If set to "warn" then a warning will be emitted in cache.log
4543 each time such HTTP error is encountered.
4545 If set to "off" then such HTTP errors will cause the request
4546 or response to be rejected.
4551 -----------------------------------------------------------------------------
4554 NAME: forward_timeout
4557 LOC: Config.Timeout.forward
4560 This parameter specifies how long Squid should at most attempt in
4561 finding a forwarding path for the request before giving up.
4564 NAME: connect_timeout
4567 LOC: Config.Timeout.connect
4570 This parameter specifies how long to wait for the TCP connect to
4571 the requested server or peer to complete before Squid should
4572 attempt to find another path where to forward the request.
4575 NAME: peer_connect_timeout
4578 LOC: Config.Timeout.peer_connect
4581 This parameter specifies how long to wait for a pending TCP
4582 connection to a peer cache. The default is 30 seconds. You
4583 may also set different timeout values for individual neighbors
4584 with the 'connect-timeout' option on a 'cache_peer' line.
4590 LOC: Config.Timeout.read
4593 The read_timeout is applied on server-side connections. After
4594 each successful read(), the timeout will be extended by this
4595 amount. If no data is read again after this amount of time,
4596 the request is aborted and logged with ERR_READ_TIMEOUT. The
4597 default is 15 minutes.
4603 LOC: Config.Timeout.write
4606 This timeout is tracked for all connections that have data
4607 available for writing and are waiting for the socket to become
4608 ready. After each successful write, the timeout is extended by
4609 the configured amount. If Squid has data to write but the
4610 connection is not ready for the configured duration, the
4611 transaction associated with the connection is terminated. The
4612 default is 15 minutes.
4615 NAME: request_timeout
4617 LOC: Config.Timeout.request
4620 How long to wait for an HTTP request after initial
4621 connection establishment.
4624 NAME: persistent_request_timeout
4626 LOC: Config.Timeout.persistent_request
4629 How long to wait for the next HTTP request on a persistent
4630 connection after the previous request completes.
4633 NAME: client_lifetime
4636 LOC: Config.Timeout.lifetime
4639 The maximum amount of time a client (browser) is allowed to
4640 remain connected to the cache process. This protects the Cache
4641 from having a lot of sockets (and hence file descriptors) tied up
4642 in a CLOSE_WAIT state from remote clients that go away without
4643 properly shutting down (either because of a network failure or
4644 because of a poor client implementation). The default is one
4647 NOTE: The default value is intended to be much larger than any
4648 client would ever need to be connected to your cache. You
4649 should probably change client_lifetime only as a last resort.
4650 If you seem to have many client connections tying up
4651 filedescriptors, we recommend first tuning the read_timeout,
4652 request_timeout, persistent_request_timeout and quick_abort values.
4655 NAME: half_closed_clients
4657 LOC: Config.onoff.half_closed_clients
4660 Some clients may shutdown the sending side of their TCP
4661 connections, while leaving their receiving sides open. Sometimes,
4662 Squid can not tell the difference between a half-closed and a
4663 fully-closed TCP connection.
4665 By default, Squid will immediately close client connections when
4666 read(2) returns "no more data to read."
4668 Change this option to 'on' and Squid will keep open connections
4669 until a read(2) or write(2) on the socket returns an error.
4670 This may show some benefits for reverse proxies. But if not
4671 it is recommended to leave OFF.
4676 LOC: Config.Timeout.pconn
4679 Timeout for idle persistent connections to servers and other
4686 LOC: Ident::TheConfig.timeout
4689 Maximum time to wait for IDENT lookups to complete.
4691 If this is too high, and you enabled IDENT lookups from untrusted
4692 users, you might be susceptible to denial-of-service by having
4693 many ident requests going at once.
4696 NAME: shutdown_lifetime
4699 LOC: Config.shutdownLifetime
4702 When SIGTERM or SIGHUP is received, the cache is put into
4703 "shutdown pending" mode until all active sockets are closed.
4704 This value is the lifetime to set for all open descriptors
4705 during shutdown mode. Any active clients after this many
4706 seconds will receive a 'timeout' message.
4710 ADMINISTRATIVE PARAMETERS
4711 -----------------------------------------------------------------------------
4717 LOC: Config.adminEmail
4719 Email-address of local cache manager who will receive
4720 mail if the cache dies. The default is "webmaster."
4726 LOC: Config.EmailFrom
4728 From: email-address for mail sent when the cache dies.
4729 The default is to use 'appname@unique_hostname'.
4730 Default appname value is "squid", can be changed into
4731 src/globals.h before building squid.
4737 LOC: Config.EmailProgram
4739 Email program used to send mail if the cache dies.
4740 The default is "mail". The specified program must comply
4741 with the standard Unix mail syntax:
4742 mail-program recipient < mailfile
4744 Optional command line options can be specified.
4747 NAME: cache_effective_user
4749 DEFAULT: @DEFAULT_CACHE_EFFECTIVE_USER@
4750 LOC: Config.effectiveUser
4752 If you start Squid as root, it will change its effective/real
4753 UID/GID to the user specified below. The default is to change
4754 to UID of @DEFAULT_CACHE_EFFECTIVE_USER@.
4755 see also; cache_effective_group
4758 NAME: cache_effective_group
4761 LOC: Config.effectiveGroup
4763 Squid sets the GID to the effective user's default group ID
4764 (taken from the password file) and supplementary group list
4765 from the groups membership.
4767 If you want Squid to run with a specific GID regardless of
4768 the group memberships of the effective user then set this
4769 to the group (or GID) you want Squid to run as. When set
4770 all other group privileges of the effective user are ignored
4771 and only this GID is effective. If Squid is not started as
4772 root the user starting Squid MUST be member of the specified
4775 This option is not recommended by the Squid Team.
4776 Our preference is for administrators to configure a secure
4777 user account for squid with UID/GID matching system policies.
4780 NAME: httpd_suppress_version_string
4784 LOC: Config.onoff.httpd_suppress_version_string
4786 Suppress Squid version string info in HTTP headers and HTML error pages.
4789 NAME: visible_hostname
4791 LOC: Config.visibleHostname
4794 If you want to present a special hostname in error messages, etc,
4795 define this. Otherwise, the return value of gethostname()
4796 will be used. If you have multiple caches in a cluster and
4797 get errors about IP-forwarding you must set them to have individual
4798 names with this setting.
4801 NAME: unique_hostname
4803 LOC: Config.uniqueHostname
4806 If you want to have multiple machines with the same
4807 'visible_hostname' you must give each machine a different
4808 'unique_hostname' so forwarding loops can be detected.
4811 NAME: hostname_aliases
4813 LOC: Config.hostnameAliases
4816 A list of other DNS names your cache has.
4824 Minimum umask which should be enforced while the proxy
4825 is running, in addition to the umask set at startup.
4827 For a traditional octal representation of umasks, start
4832 OPTIONS FOR THE CACHE REGISTRATION SERVICE
4833 -----------------------------------------------------------------------------
4835 This section contains parameters for the (optional) cache
4836 announcement service. This service is provided to help
4837 cache administrators locate one another in order to join or
4838 create cache hierarchies.
4840 An 'announcement' message is sent (via UDP) to the registration
4841 service by Squid. By default, the announcement message is NOT
4842 SENT unless you enable it with 'announce_period' below.
4844 The announcement message includes your hostname, plus the
4845 following information from this configuration file:
4851 All current information is processed regularly and made
4852 available on the Web at http://www.ircache.net/Cache/Tracker/.
4855 NAME: announce_period
4857 LOC: Config.Announce.period
4860 This is how frequently to send cache announcements. The
4861 default is `0' which disables sending the announcement
4864 To enable announcing your cache, just set an announce period.
4867 announce_period 1 day
4872 DEFAULT: tracker.ircache.net
4873 LOC: Config.Announce.host
4879 LOC: Config.Announce.file
4885 LOC: Config.Announce.port
4887 announce_host and announce_port set the hostname and port
4888 number where the registration message will be sent.
4890 Hostname will default to 'tracker.ircache.net' and port will
4891 default default to 3131. If the 'filename' argument is given,
4892 the contents of that file will be included in the announce
4897 HTTPD-ACCELERATOR OPTIONS
4898 -----------------------------------------------------------------------------
4901 NAME: httpd_accel_surrogate_id
4904 LOC: Config.Accel.surrogate_id
4906 Surrogates (http://www.esi.org/architecture_spec_1.0.html)
4907 need an identification token to allow control targeting. Because
4908 a farm of surrogates may all perform the same tasks, they may share
4909 an identification token.
4911 The default ID is the visible_hostname
4914 NAME: http_accel_surrogate_remote
4918 LOC: Config.onoff.surrogate_is_remote
4920 Remote surrogates (such as those in a CDN) honour Surrogate-Control: no-store-remote.
4921 Set this to on to have squid behave as a remote surrogate.
4925 IFDEF: USE_SQUID_ESI
4926 COMMENT: libxml2|expat|custom
4928 LOC: ESIParser::Type
4931 ESI markup is not strictly XML compatible. The custom ESI parser
4932 will give higher performance, but cannot handle non ASCII character
4937 DELAY POOL PARAMETERS
4938 -----------------------------------------------------------------------------
4942 TYPE: delay_pool_count
4944 IFDEF: USE_DELAY_POOLS
4947 This represents the number of delay pools to be used. For example,
4948 if you have one class 2 delay pool and one class 3 delays pool, you
4949 have a total of 2 delay pools.
4953 TYPE: delay_pool_class
4955 IFDEF: USE_DELAY_POOLS
4958 This defines the class of each delay pool. There must be exactly one
4959 delay_class line for each delay pool. For example, to define two
4960 delay pools, one of class 2 and one of class 3, the settings above
4964 delay_pools 4 # 4 delay pools
4965 delay_class 1 2 # pool 1 is a class 2 pool
4966 delay_class 2 3 # pool 2 is a class 3 pool
4967 delay_class 3 4 # pool 3 is a class 4 pool
4968 delay_class 4 5 # pool 4 is a class 5 pool
4970 The delay pool classes are:
4972 class 1 Everything is limited by a single aggregate
4975 class 2 Everything is limited by a single aggregate
4976 bucket as well as an "individual" bucket chosen
4977 from bits 25 through 32 of the IPv4 address.
4979 class 3 Everything is limited by a single aggregate
4980 bucket as well as a "network" bucket chosen
4981 from bits 17 through 24 of the IP address and a
4982 "individual" bucket chosen from bits 17 through
4983 32 of the IPv4 address.
4985 class 4 Everything in a class 3 delay pool, with an
4986 additional limit on a per user basis. This
4987 only takes effect if the username is established
4988 in advance - by forcing authentication in your
4991 class 5 Requests are grouped according their tag (see
4992 external_acl's tag= reply).
4995 Each pool also requires a delay_parameters directive to configure the pool size
4996 and speed limits used whenever the pool is applied to a request. Along with
4997 a set of delay_access directives to determine when it is used.
4999 NOTE: If an IP address is a.b.c.d
5000 -> bits 25 through 32 are "d"
5001 -> bits 17 through 24 are "c"
5002 -> bits 17 through 32 are "c * 256 + d"
5004 NOTE-2: Due to the use of bitmasks in class 2,3,4 pools they only apply to
5005 IPv4 traffic. Class 1 and 5 pools may be used with IPv6 traffic.
5009 TYPE: delay_pool_access
5011 IFDEF: USE_DELAY_POOLS
5014 This is used to determine which delay pool a request falls into.
5016 delay_access is sorted per pool and the matching starts with pool 1,
5017 then pool 2, ..., and finally pool N. The first delay pool where the
5018 request is allowed is selected for the request. If it does not allow
5019 the request to any pool then the request is not delayed (default).
5021 For example, if you want some_big_clients in delay
5022 pool 1 and lotsa_little_clients in delay pool 2:
5025 delay_access 1 allow some_big_clients
5026 delay_access 1 deny all
5027 delay_access 2 allow lotsa_little_clients
5028 delay_access 2 deny all
5029 delay_access 3 allow authenticated_clients
5032 NAME: delay_parameters
5033 TYPE: delay_pool_rates
5035 IFDEF: USE_DELAY_POOLS
5038 This defines the parameters for a delay pool. Each delay pool has
5039 a number of "buckets" associated with it, as explained in the
5040 description of delay_class.
5042 For a class 1 delay pool, the syntax is:
5044 delay_parameters pool aggregate
5046 For a class 2 delay pool:
5048 delay_parameters pool aggregate individual
5050 For a class 3 delay pool:
5052 delay_parameters pool aggregate network individual
5054 For a class 4 delay pool:
5056 delay_parameters pool aggregate network individual user
5058 For a class 5 delay pool:
5060 delay_parameters pool tagrate
5062 The option variables are:
5064 pool a pool number - ie, a number between 1 and the
5065 number specified in delay_pools as used in
5068 aggregate the speed limit parameters for the aggregate bucket
5071 individual the speed limit parameters for the individual
5072 buckets (class 2, 3).
5074 network the speed limit parameters for the network buckets
5077 user the speed limit parameters for the user buckets
5080 tagrate the speed limit parameters for the tag buckets
5083 A pair of delay parameters is written restore/maximum, where restore is
5084 the number of bytes (not bits - modem and network speeds are usually
5085 quoted in bits) per second placed into the bucket, and maximum is the
5086 maximum number of bytes which can be in the bucket at any time.
5088 There must be one delay_parameters line for each delay pool.
5091 For example, if delay pool number 1 is a class 2 delay pool as in the
5092 above example, and is being used to strictly limit each host to 64Kbit/sec
5093 (plus overheads), with no overall limit, the line is:
5095 delay_parameters 1 -1/-1 8000/8000
5097 Note that 8 x 8000 KByte/sec -> 64Kbit/sec.
5099 Note that the figure -1 is used to represent "unlimited".
5102 And, if delay pool number 2 is a class 3 delay pool as in the above
5103 example, and you want to limit it to a total of 256Kbit/sec (strict limit)
5104 with each 8-bit network permitted 64Kbit/sec (strict limit) and each
5105 individual host permitted 4800bit/sec with a bucket maximum size of 64Kbits
5106 to permit a decent web page to be downloaded at a decent speed
5107 (if the network is not being limited due to overuse) but slow down
5108 large downloads more significantly:
5110 delay_parameters 2 32000/32000 8000/8000 600/8000
5112 Note that 8 x 32000 KByte/sec -> 256Kbit/sec.
5113 8 x 8000 KByte/sec -> 64Kbit/sec.
5114 8 x 600 Byte/sec -> 4800bit/sec.
5117 Finally, for a class 4 delay pool as in the example - each user will
5118 be limited to 128Kbits/sec no matter how many workstations they are logged into.:
5120 delay_parameters 4 32000/32000 8000/8000 600/64000 16000/16000
5123 NAME: delay_initial_bucket_level
5124 COMMENT: (percent, 0-100)
5127 IFDEF: USE_DELAY_POOLS
5128 LOC: Config.Delay.initial
5130 The initial bucket percentage is used to determine how much is put
5131 in each bucket when squid starts, is reconfigured, or first notices
5132 a host accessing it (in class 2 and class 3, individual hosts and
5133 networks only have buckets associated with them once they have been
5138 CLIENT DELAY POOL PARAMETERS
5139 -----------------------------------------------------------------------------
5142 NAME: client_delay_pools
5143 TYPE: client_delay_pool_count
5145 IFDEF: USE_DELAY_POOLS
5146 LOC: Config.ClientDelay
5148 This option specifies the number of client delay pools used. It must
5149 preceed other client_delay_* options.
5152 client_delay_pools 2
5155 NAME: client_delay_initial_bucket_level
5156 COMMENT: (percent, 0-no_limit)
5159 IFDEF: USE_DELAY_POOLS
5160 LOC: Config.ClientDelay.initial
5162 This option determines the initial bucket size as a percentage of
5163 max_bucket_size from client_delay_parameters. Buckets are created
5164 at the time of the "first" connection from the matching IP. Idle
5165 buckets are periodically deleted up.
5167 You can specify more than 100 percent but note that such "oversized"
5168 buckets are not refilled until their size goes down to max_bucket_size
5169 from client_delay_parameters.
5172 client_delay_initial_bucket_level 50
5175 NAME: client_delay_parameters
5176 TYPE: client_delay_pool_rates
5178 IFDEF: USE_DELAY_POOLS
5179 LOC: Config.ClientDelay
5182 This option configures client-side bandwidth limits using the
5185 client_delay_parameters pool speed_limit max_bucket_size
5187 pool is an integer ID used for client_delay_access matching.
5189 speed_limit is bytes added to the bucket per second.
5191 max_bucket_size is the maximum size of a bucket, enforced after any
5192 speed_limit additions.
5194 Please see the delay_parameters option for more information and
5198 client_delay_parameters 1 1024 2048
5199 client_delay_parameters 2 51200 16384
5202 NAME: client_delay_access
5203 TYPE: client_delay_pool_access
5205 IFDEF: USE_DELAY_POOLS
5206 LOC: Config.ClientDelay
5209 This option determines the client-side delay pool for the
5212 client_delay_access pool_ID allow|deny acl_name
5214 All client_delay_access options are checked in their pool ID
5215 order, starting with pool 1. The first checked pool with allowed
5216 request is selected for the request. If no ACL matches or there
5217 are no client_delay_access options, the request bandwidth is not
5220 The ACL-selected pool is then used to find the
5221 client_delay_parameters for the request. Client-side pools are
5222 not used to aggregate clients. Clients are always aggregated
5223 based on their source IP addresses (one bucket per source IP).
5225 Please see delay_access for more examples.
5228 client_delay_access 1 allow low_rate_network
5229 client_delay_access 2 allow vips_network
5233 WCCPv1 AND WCCPv2 CONFIGURATION OPTIONS
5234 -----------------------------------------------------------------------------
5239 LOC: Config.Wccp.router
5243 Use this option to define your WCCP ``home'' router for
5246 wccp_router supports a single WCCP(v1) router
5248 wccp2_router supports multiple WCCPv2 routers
5250 only one of the two may be used at the same time and defines
5251 which version of WCCP to use.
5255 TYPE: IpAddress_list
5256 LOC: Config.Wccp2.router
5260 Use this option to define your WCCP ``home'' router for
5263 wccp_router supports a single WCCP(v1) router
5265 wccp2_router supports multiple WCCPv2 routers
5267 only one of the two may be used at the same time and defines
5268 which version of WCCP to use.
5273 LOC: Config.Wccp.version
5277 This directive is only relevant if you need to set up WCCP(v1)
5278 to some very old and end-of-life Cisco routers. In all other
5279 setups it must be left unset or at the default setting.
5280 It defines an internal version in the WCCP(v1) protocol,
5281 with version 4 being the officially documented protocol.
5283 According to some users, Cisco IOS 11.2 and earlier only
5284 support WCCP version 3. If you're using that or an earlier
5285 version of IOS, you may need to change this value to 3, otherwise
5286 do not specify this parameter.
5289 NAME: wccp2_rebuild_wait
5291 LOC: Config.Wccp2.rebuildwait
5295 If this is enabled Squid will wait for the cache dir rebuild to finish
5296 before sending the first wccp2 HereIAm packet
5299 NAME: wccp2_forwarding_method
5301 LOC: Config.Wccp2.forwarding_method
5305 WCCP2 allows the setting of forwarding methods between the
5306 router/switch and the cache. Valid values are as follows:
5308 gre - GRE encapsulation (forward the packet in a GRE/WCCP tunnel)
5309 l2 - L2 redirect (forward the packet using Layer 2/MAC rewriting)
5311 Currently (as of IOS 12.4) cisco routers only support GRE.
5312 Cisco switches only support the L2 redirect assignment method.
5315 NAME: wccp2_return_method
5317 LOC: Config.Wccp2.return_method
5321 WCCP2 allows the setting of return methods between the
5322 router/switch and the cache for packets that the cache
5323 decides not to handle. Valid values are as follows:
5325 gre - GRE encapsulation (forward the packet in a GRE/WCCP tunnel)
5326 l2 - L2 redirect (forward the packet using Layer 2/MAC rewriting)
5328 Currently (as of IOS 12.4) cisco routers only support GRE.
5329 Cisco switches only support the L2 redirect assignment.
5331 If the "ip wccp redirect exclude in" command has been
5332 enabled on the cache interface, then it is still safe for
5333 the proxy server to use a l2 redirect method even if this
5334 option is set to GRE.
5337 NAME: wccp2_assignment_method
5339 LOC: Config.Wccp2.assignment_method
5343 WCCP2 allows the setting of methods to assign the WCCP hash
5344 Valid values are as follows:
5346 hash - Hash assignment
5347 mask - Mask assignment
5349 As a general rule, cisco routers support the hash assignment method
5350 and cisco switches support the mask assignment method.
5355 LOC: Config.Wccp2.info
5356 DEFAULT_IF_NONE: standard 0
5359 WCCP2 allows for multiple traffic services. There are two
5360 types: "standard" and "dynamic". The standard type defines
5361 one service id - http (id 0). The dynamic service ids can be from
5362 51 to 255 inclusive. In order to use a dynamic service id
5363 one must define the type of traffic to be redirected; this is done
5364 using the wccp2_service_info option.
5366 The "standard" type does not require a wccp2_service_info option,
5367 just specifying the service id will suffice.
5369 MD5 service authentication can be enabled by adding
5370 "password=<password>" to the end of this service declaration.
5374 wccp2_service standard 0 # for the 'web-cache' standard service
5375 wccp2_service dynamic 80 # a dynamic service type which will be
5376 # fleshed out with subsequent options.
5377 wccp2_service standard 0 password=foo
5380 NAME: wccp2_service_info
5381 TYPE: wccp2_service_info
5382 LOC: Config.Wccp2.info
5386 Dynamic WCCPv2 services require further information to define the
5387 traffic you wish to have diverted.
5391 wccp2_service_info <id> protocol=<protocol> flags=<flag>,<flag>..
5392 priority=<priority> ports=<port>,<port>..
5394 The relevant WCCPv2 flags:
5395 + src_ip_hash, dst_ip_hash
5396 + source_port_hash, dst_port_hash
5397 + src_ip_alt_hash, dst_ip_alt_hash
5398 + src_port_alt_hash, dst_port_alt_hash
5401 The port list can be one to eight entries.
5405 wccp2_service_info 80 protocol=tcp flags=src_ip_hash,ports_source
5406 priority=240 ports=80
5408 Note: the service id must have been defined by a previous
5409 'wccp2_service dynamic <id>' entry.
5414 LOC: Config.Wccp2.weight
5418 Each cache server gets assigned a set of the destination
5419 hash proportional to their weight.
5424 LOC: Config.Wccp.address
5431 LOC: Config.Wccp2.address
5435 Use this option if you require WCCP to use a specific
5438 The default behavior is to not bind to any specific address.
5442 PERSISTENT CONNECTION HANDLING
5443 -----------------------------------------------------------------------------
5445 Also see "pconn_timeout" in the TIMEOUTS section
5448 NAME: client_persistent_connections
5450 LOC: Config.onoff.client_pconns
5454 NAME: server_persistent_connections
5456 LOC: Config.onoff.server_pconns
5459 Persistent connection support for clients and servers. By
5460 default, Squid uses persistent connections (when allowed)
5461 with its clients and servers. You can use these options to
5462 disable persistent connections with clients and/or servers.
5465 NAME: persistent_connection_after_error
5467 LOC: Config.onoff.error_pconns
5470 With this directive the use of persistent connections after
5471 HTTP errors can be disabled. Useful if you have clients
5472 who fail to handle errors on persistent connections proper.
5475 NAME: detect_broken_pconn
5477 LOC: Config.onoff.detect_broken_server_pconns
5480 Some servers have been found to incorrectly signal the use
5481 of HTTP/1.0 persistent connections even on replies not
5482 compatible, causing significant delays. This server problem
5483 has mostly been seen on redirects.
5485 By enabling this directive Squid attempts to detect such
5486 broken replies and automatically assume the reply is finished
5487 after 10 seconds timeout.
5491 CACHE DIGEST OPTIONS
5492 -----------------------------------------------------------------------------
5495 NAME: digest_generation
5496 IFDEF: USE_CACHE_DIGESTS
5498 LOC: Config.onoff.digest_generation
5501 This controls whether the server will generate a Cache Digest
5502 of its contents. By default, Cache Digest generation is
5503 enabled if Squid is compiled with --enable-cache-digests defined.
5506 NAME: digest_bits_per_entry
5507 IFDEF: USE_CACHE_DIGESTS
5509 LOC: Config.digest.bits_per_entry
5512 This is the number of bits of the server's Cache Digest which
5513 will be associated with the Digest entry for a given HTTP
5514 Method and URL (public key) combination. The default is 5.
5517 NAME: digest_rebuild_period
5518 IFDEF: USE_CACHE_DIGESTS
5521 LOC: Config.digest.rebuild_period
5524 This is the wait time between Cache Digest rebuilds.
5527 NAME: digest_rewrite_period
5529 IFDEF: USE_CACHE_DIGESTS
5531 LOC: Config.digest.rewrite_period
5534 This is the wait time between Cache Digest writes to
5538 NAME: digest_swapout_chunk_size
5541 IFDEF: USE_CACHE_DIGESTS
5542 LOC: Config.digest.swapout_chunk_size
5545 This is the number of bytes of the Cache Digest to write to
5546 disk at a time. It defaults to 4096 bytes (4KB), the Squid
5550 NAME: digest_rebuild_chunk_percentage
5551 COMMENT: (percent, 0-100)
5552 IFDEF: USE_CACHE_DIGESTS
5554 LOC: Config.digest.rebuild_chunk_percentage
5557 This is the percentage of the Cache Digest to be scanned at a
5558 time. By default it is set to 10% of the Cache Digest.
5563 -----------------------------------------------------------------------------
5568 LOC: Config.Port.snmp
5572 The port number where Squid listens for SNMP requests. To enable
5573 SNMP support set this to a suitable port number. Port number
5574 3401 is often used for the Squid SNMP agent. By default it's
5575 set to "0" (disabled)
5583 LOC: Config.accessList.snmp
5584 DEFAULT_IF_NONE: deny all
5587 Allowing or denying access to the SNMP port.
5589 All access to the agent is denied by default.
5592 snmp_access allow|deny [!]aclname ...
5594 This clause only supports fast acl types.
5595 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
5597 snmp_access allow snmppublic localhost
5598 snmp_access deny all
5601 NAME: snmp_incoming_address
5603 LOC: Config.Addrs.snmp_incoming
5608 NAME: snmp_outgoing_address
5610 LOC: Config.Addrs.snmp_outgoing
5614 Just like 'udp_incoming_address', but for the SNMP port.
5616 snmp_incoming_address is used for the SNMP socket receiving
5617 messages from SNMP agents.
5618 snmp_outgoing_address is used for SNMP packets returned to SNMP
5621 The default snmp_incoming_address is to listen on all
5622 available network interfaces.
5624 If snmp_outgoing_address is not set it will use the same socket
5625 as snmp_incoming_address. Only change this if you want to have
5626 SNMP replies sent using another address than where this Squid
5627 listens for SNMP queries.
5629 NOTE, snmp_incoming_address and snmp_outgoing_address can not have
5630 the same value since they both use port 3401.
5635 -----------------------------------------------------------------------------
5638 NAME: icp_port udp_port
5641 LOC: Config.Port.icp
5643 The port number where Squid sends and receives ICP queries to
5644 and from neighbor caches. The standard UDP port for ICP is 3130.
5645 Default is disabled (0).
5648 icp_port @DEFAULT_ICP_PORT@
5655 LOC: Config.Port.htcp
5657 The port number where Squid sends and receives HTCP queries to
5658 and from neighbor caches. To turn it on you want to set it to
5659 4827. By default it is set to "0" (disabled).
5665 NAME: log_icp_queries
5669 LOC: Config.onoff.log_udp
5671 If set, ICP queries are logged to access.log. You may wish
5672 do disable this if your ICP load is VERY high to speed things
5673 up or to simplify log analysis.
5676 NAME: udp_incoming_address
5678 LOC:Config.Addrs.udp_incoming
5681 udp_incoming_address is used for UDP packets received from other
5684 The default behavior is to not bind to any specific address.
5686 Only change this if you want to have all UDP queries received on
5687 a specific interface/address.
5689 NOTE: udp_incoming_address is used by the ICP, HTCP, and DNS
5690 modules. Altering it will affect all of them in the same manner.
5692 see also; udp_outgoing_address
5694 NOTE, udp_incoming_address and udp_outgoing_address can not
5695 have the same value since they both use the same port.
5698 NAME: udp_outgoing_address
5700 LOC: Config.Addrs.udp_outgoing
5703 udp_outgoing_address is used for UDP packets sent out to other
5706 The default behavior is to not bind to any specific address.
5708 Instead it will use the same socket as udp_incoming_address.
5709 Only change this if you want to have UDP queries sent using another
5710 address than where this Squid listens for UDP queries from other
5713 NOTE: udp_outgoing_address is used by the ICP, HTCP, and DNS
5714 modules. Altering it will affect all of them in the same manner.
5716 see also; udp_incoming_address
5718 NOTE, udp_incoming_address and udp_outgoing_address can not
5719 have the same value since they both use the same port.
5726 LOC: Config.onoff.icp_hit_stale
5728 If you want to return ICP_HIT for stale cache objects, set this
5729 option to 'on'. If you have sibling relationships with caches
5730 in other administrative domains, this should be 'off'. If you only
5731 have sibling relationships with caches under your control,
5732 it is probably okay to set this to 'on'.
5733 If set to 'on', your siblings should use the option "allow-miss"
5734 on their cache_peer lines for connecting to you.
5737 NAME: minimum_direct_hops
5740 LOC: Config.minDirectHops
5742 If using the ICMP pinging stuff, do direct fetches for sites
5743 which are no more than this many hops away.
5746 NAME: minimum_direct_rtt
5749 LOC: Config.minDirectRtt
5751 If using the ICMP pinging stuff, do direct fetches for sites
5752 which are no more than this many rtt milliseconds away.
5758 LOC: Config.Netdb.low
5764 LOC: Config.Netdb.high
5766 The low and high water marks for the ICMP measurement
5767 database. These are counts, not percents. The defaults are
5768 900 and 1000. When the high water mark is reached, database
5769 entries will be deleted until the low mark is reached.
5772 NAME: netdb_ping_period
5774 LOC: Config.Netdb.period
5777 The minimum period for measuring a site. There will be at
5778 least this much delay between successive pings to the same
5779 network. The default is five minutes.
5786 LOC: Config.onoff.query_icmp
5788 If you want to ask your peers to include ICMP data in their ICP
5789 replies, enable this option.
5791 If your peer has configured Squid (during compilation) with
5792 '--enable-icmp' that peer will send ICMP pings to origin server
5793 sites of the URLs it receives. If you enable this option the
5794 ICP replies from that peer will include the ICMP data (if available).
5795 Then, when choosing a parent cache, Squid will choose the parent with
5796 the minimal RTT to the origin server. When this happens, the
5797 hierarchy field of the access.log will be
5798 "CLOSEST_PARENT_MISS". This option is off by default.
5801 NAME: test_reachability
5805 LOC: Config.onoff.test_reachability
5807 When this is 'on', ICP MISS replies will be ICP_MISS_NOFETCH
5808 instead of ICP_MISS if the target host is NOT in the ICMP
5809 database, or has a zero RTT.
5812 NAME: icp_query_timeout
5816 LOC: Config.Timeout.icp_query
5818 Normally Squid will automatically determine an optimal ICP
5819 query timeout value based on the round-trip-time of recent ICP
5820 queries. If you want to override the value determined by
5821 Squid, set this 'icp_query_timeout' to a non-zero value. This
5822 value is specified in MILLISECONDS, so, to use a 2-second
5823 timeout (the old default), you would write:
5825 icp_query_timeout 2000
5828 NAME: maximum_icp_query_timeout
5832 LOC: Config.Timeout.icp_query_max
5834 Normally the ICP query timeout is determined dynamically. But
5835 sometimes it can lead to very large values (say 5 seconds).
5836 Use this option to put an upper limit on the dynamic timeout
5837 value. Do NOT use this option to always use a fixed (instead
5838 of a dynamic) timeout value. To set a fixed timeout see the
5839 'icp_query_timeout' directive.
5842 NAME: minimum_icp_query_timeout
5846 LOC: Config.Timeout.icp_query_min
5848 Normally the ICP query timeout is determined dynamically. But
5849 sometimes it can lead to very small timeouts, even lower than
5850 the normal latency variance on your link due to traffic.
5851 Use this option to put an lower limit on the dynamic timeout
5852 value. Do NOT use this option to always use a fixed (instead
5853 of a dynamic) timeout value. To set a fixed timeout see the
5854 'icp_query_timeout' directive.
5857 NAME: background_ping_rate
5861 LOC: Config.backgroundPingRate
5863 Controls how often the ICP pings are sent to siblings that
5864 have background-ping set.
5868 MULTICAST ICP OPTIONS
5869 -----------------------------------------------------------------------------
5874 LOC: Config.mcast_group_list
5877 This tag specifies a list of multicast groups which your server
5878 should join to receive multicasted ICP queries.
5880 NOTE! Be very careful what you put here! Be sure you
5881 understand the difference between an ICP _query_ and an ICP
5882 _reply_. This option is to be set only if you want to RECEIVE
5883 multicast queries. Do NOT set this option to SEND multicast
5884 ICP (use cache_peer for that). ICP replies are always sent via
5885 unicast, so this option does not affect whether or not you will
5886 receive replies from multicast group members.
5888 You must be very careful to NOT use a multicast address which
5889 is already in use by another group of caches.
5891 If you are unsure about multicast, please read the Multicast
5892 chapter in the Squid FAQ (http://www.squid-cache.org/FAQ/).
5894 Usage: mcast_groups 239.128.16.128 224.0.1.20
5896 By default, Squid doesn't listen on any multicast groups.
5899 NAME: mcast_miss_addr
5900 IFDEF: MULTICAST_MISS_STREAM
5902 LOC: Config.mcast_miss.addr
5905 If you enable this option, every "cache miss" URL will
5906 be sent out on the specified multicast address.
5908 Do not enable this option unless you are are absolutely
5909 certain you understand what you are doing.
5912 NAME: mcast_miss_ttl
5913 IFDEF: MULTICAST_MISS_STREAM
5915 LOC: Config.mcast_miss.ttl
5918 This is the time-to-live value for packets multicasted
5919 when multicasting off cache miss URLs is enabled. By
5920 default this is set to 'site scope', i.e. 16.
5923 NAME: mcast_miss_port
5924 IFDEF: MULTICAST_MISS_STREAM
5926 LOC: Config.mcast_miss.port
5929 This is the port number to be used in conjunction with
5933 NAME: mcast_miss_encode_key
5934 IFDEF: MULTICAST_MISS_STREAM
5936 LOC: Config.mcast_miss.encode_key
5937 DEFAULT: XXXXXXXXXXXXXXXX
5939 The URLs that are sent in the multicast miss stream are
5940 encrypted. This is the encryption key.
5943 NAME: mcast_icp_query_timeout
5947 LOC: Config.Timeout.mcast_icp_query
5949 For multicast peers, Squid regularly sends out ICP "probes" to
5950 count how many other peers are listening on the given multicast
5951 address. This value specifies how long Squid should wait to
5952 count all the replies. The default is 2000 msec, or 2
5957 INTERNAL ICON OPTIONS
5958 -----------------------------------------------------------------------------
5961 NAME: icon_directory
5963 LOC: Config.icons.directory
5964 DEFAULT: @DEFAULT_ICON_DIR@
5966 Where the icons are stored. These are normally kept in
5970 NAME: global_internal_static
5972 LOC: Config.onoff.global_internal_static
5975 This directive controls is Squid should intercept all requests for
5976 /squid-internal-static/ no matter which host the URL is requesting
5977 (default on setting), or if nothing special should be done for
5978 such URLs (off setting). The purpose of this directive is to make
5979 icons etc work better in complex cache hierarchies where it may
5980 not always be possible for all corners in the cache mesh to reach
5981 the server generating a directory listing.
5984 NAME: short_icon_urls
5986 LOC: Config.icons.use_short_names
5989 If this is enabled Squid will use short URLs for icons.
5990 If disabled it will revert to the old behavior of including
5991 it's own name and port in the URL.
5993 If you run a complex cache hierarchy with a mix of Squid and
5994 other proxies you may need to disable this directive.
5999 -----------------------------------------------------------------------------
6002 NAME: error_directory
6004 LOC: Config.errorDirectory
6007 If you wish to create your own versions of the default
6008 error files to customize them to suit your company copy
6009 the error/template files to another directory and point
6012 WARNING: This option will disable multi-language support
6013 on error pages if used.
6015 The squid developers are interested in making squid available in
6016 a wide variety of languages. If you are making translations for a
6017 language that Squid does not currently provide please consider
6018 contributing your translation back to the project.
6019 http://wiki.squid-cache.org/Translations
6021 The squid developers working on translations are happy to supply drop-in
6022 translated error files in exchange for any new language contributions.
6025 NAME: error_default_language
6026 IFDEF: USE_ERR_LOCALES
6028 LOC: Config.errorDefaultLanguage
6031 Set the default language which squid will send error pages in
6032 if no existing translation matches the clients language
6035 If unset (default) generic English will be used.
6037 The squid developers are interested in making squid available in
6038 a wide variety of languages. If you are interested in making
6039 translations for any language see the squid wiki for details.
6040 http://wiki.squid-cache.org/Translations
6043 NAME: error_log_languages
6044 IFDEF: USE_ERR_LOCALES
6046 LOC: Config.errorLogMissingLanguages
6049 Log to cache.log what languages users are attempting to
6050 auto-negotiate for translations.
6052 Successful negotiations are not logged. Only failures
6053 have meaning to indicate that Squid may need an upgrade
6054 of its error page translations.
6057 NAME: err_page_stylesheet
6059 LOC: Config.errorStylesheet
6060 DEFAULT: @DEFAULT_CONFIG_DIR@/errorpage.css
6062 CSS Stylesheet to pattern the display of Squid default error pages.
6064 For information on CSS see http://www.w3.org/Style/CSS/
6069 LOC: Config.errHtmlText
6072 HTML text to include in error messages. Make this a "mailto"
6073 URL to your admin address, or maybe just a link to your
6074 organizations Web page.
6076 To include this in your error messages, you must rewrite
6077 the error template files (found in the "errors" directory).
6078 Wherever you want the 'err_html_text' line to appear,
6079 insert a %L tag in the error template file.
6082 NAME: email_err_data
6085 LOC: Config.onoff.emailErrData
6088 If enabled, information about the occurred error will be
6089 included in the mailto links of the ERR pages (if %W is set)
6090 so that the email body contains the data.
6091 Syntax is <A HREF="mailto:%w%W">%w</A>
6096 LOC: Config.denyInfoList
6099 Usage: deny_info err_page_name acl
6100 or deny_info http://... acl
6101 or deny_info TCP_RESET acl
6103 This can be used to return a ERR_ page for requests which
6104 do not pass the 'http_access' rules. Squid remembers the last
6105 acl it evaluated in http_access, and if a 'deny_info' line exists
6106 for that ACL Squid returns a corresponding error page.
6108 The acl is typically the last acl on the http_access deny line which
6109 denied access. The exceptions to this rule are:
6110 - When Squid needs to request authentication credentials. It's then
6111 the first authentication related acl encountered
6112 - When none of the http_access lines matches. It's then the last
6113 acl processed on the last http_access line.
6114 - When the decision to deny access was made by an adaptation service,
6115 the acl name is the corresponding eCAP or ICAP service_name.
6117 NP: If providing your own custom error pages with error_directory
6118 you may also specify them by your custom file name:
6119 Example: deny_info ERR_CUSTOM_ACCESS_DENIED bad_guys
6121 By defaut Squid will send "403 Forbidden". A different 4xx or 5xx
6122 may be specified by prefixing the file name with the code and a colon.
6123 e.g. 404:ERR_CUSTOM_ACCESS_DENIED
6125 Alternatively you can tell Squid to reset the TCP connection
6126 by specifying TCP_RESET.
6128 Or you can specify an error URL or URL pattern. The browsers will
6129 get redirected to the specified URL after formatting tags have
6130 been replaced. Redirect will be done with 302 or 307 according to
6131 HTTP/1.1 specs. A different 3xx code may be specified by prefixing
6132 the URL. e.g. 303:http://example.com/
6135 %a - username (if available. Password NOT included)
6138 %E - Error description
6140 %H - Request domain name
6141 %i - Client IP Address
6143 %o - Message result from external ACL helper
6144 %p - Request Port number
6145 %P - Request Protocol name
6146 %R - Request URL path
6147 %T - Timestamp in RFC 1123 format
6148 %U - Full canonical URL from client
6149 (HTTPS URLs terminate with *)
6150 %u - Full canonical URL from client
6151 %w - Admin email from squid.conf
6153 %% - Literal percent (%) code
6158 OPTIONS INFLUENCING REQUEST FORWARDING
6159 -----------------------------------------------------------------------------
6162 NAME: nonhierarchical_direct
6164 LOC: Config.onoff.nonhierarchical_direct
6167 By default, Squid will send any non-hierarchical requests
6168 (matching hierarchy_stoplist or not cacheable request type) direct
6171 If you set this to off, Squid will prefer to send these
6172 requests to parents.
6174 Note that in most configurations, by turning this off you will only
6175 add latency to these request without any improvement in global hit
6178 If you are inside an firewall see never_direct instead of
6184 LOC: Config.onoff.prefer_direct
6187 Normally Squid tries to use parents for most requests. If you for some
6188 reason like it to first try going direct and only use a parent if
6189 going direct fails set this to on.
6191 By combining nonhierarchical_direct off and prefer_direct on you
6192 can set up Squid to use a parent as a backup path if going direct
6195 Note: If you want Squid to use parents for all requests see
6196 the never_direct directive. prefer_direct only modifies how Squid
6197 acts on cacheable requests.
6202 LOC: Config.accessList.AlwaysDirect
6205 Usage: always_direct allow|deny [!]aclname ...
6207 Here you can use ACL elements to specify requests which should
6208 ALWAYS be forwarded by Squid to the origin servers without using
6209 any peers. For example, to always directly forward requests for
6210 local servers ignoring any parents or siblings you may have use
6213 acl local-servers dstdomain my.domain.net
6214 always_direct allow local-servers
6216 To always forward FTP requests directly, use
6219 always_direct allow FTP
6221 NOTE: There is a similar, but opposite option named
6222 'never_direct'. You need to be aware that "always_direct deny
6223 foo" is NOT the same thing as "never_direct allow foo". You
6224 may need to use a deny rule to exclude a more-specific case of
6225 some other rule. Example:
6227 acl local-external dstdomain external.foo.net
6228 acl local-servers dstdomain .foo.net
6229 always_direct deny local-external
6230 always_direct allow local-servers
6232 NOTE: If your goal is to make the client forward the request
6233 directly to the origin server bypassing Squid then this needs
6234 to be done in the client configuration. Squid configuration
6235 can only tell Squid how Squid should fetch the object.
6237 NOTE: This directive is not related to caching. The replies
6238 is cached as usual even if you use always_direct. To not cache
6239 the replies see the 'cache' directive.
6241 This clause supports both fast and slow acl types.
6242 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
6247 LOC: Config.accessList.NeverDirect
6250 Usage: never_direct allow|deny [!]aclname ...
6252 never_direct is the opposite of always_direct. Please read
6253 the description for always_direct if you have not already.
6255 With 'never_direct' you can use ACL elements to specify
6256 requests which should NEVER be forwarded directly to origin
6257 servers. For example, to force the use of a proxy for all
6258 requests, except those in your local domain use something like:
6260 acl local-servers dstdomain .foo.net
6261 never_direct deny local-servers
6262 never_direct allow all
6264 or if Squid is inside a firewall and there are local intranet
6265 servers inside the firewall use something like:
6267 acl local-intranet dstdomain .foo.net
6268 acl local-external dstdomain external.foo.net
6269 always_direct deny local-external
6270 always_direct allow local-intranet
6271 never_direct allow all
6273 This clause supports both fast and slow acl types.
6274 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
6278 ADVANCED NETWORKING OPTIONS
6279 -----------------------------------------------------------------------------
6282 NAME: incoming_icp_average
6285 LOC: Config.comm_incoming.icp_average
6288 NAME: incoming_http_average
6291 LOC: Config.comm_incoming.http_average
6294 NAME: incoming_dns_average
6297 LOC: Config.comm_incoming.dns_average
6300 NAME: min_icp_poll_cnt
6303 LOC: Config.comm_incoming.icp_min_poll
6306 NAME: min_dns_poll_cnt
6309 LOC: Config.comm_incoming.dns_min_poll
6312 NAME: min_http_poll_cnt
6315 LOC: Config.comm_incoming.http_min_poll
6317 Heavy voodoo here. I can't even believe you are reading this.
6318 Are you crazy? Don't even think about adjusting these unless
6319 you understand the algorithms in comm_select.c first!
6325 LOC: Config.accept_filter
6329 The name of an accept(2) filter to install on Squid's
6330 listen socket(s). This feature is perhaps specific to
6331 FreeBSD and requires support in the kernel.
6333 The 'httpready' filter delays delivering new connections
6334 to Squid until a full HTTP request has been received.
6335 See the accf_http(9) man page for details.
6337 The 'dataready' filter delays delivering new connections
6338 to Squid until there is some data to process.
6339 See the accf_dataready(9) man page for details.
6343 The 'data' filter delays delivering of new connections
6344 to Squid until there is some data to process by TCP_ACCEPT_DEFER.
6345 You may optionally specify a number of seconds to wait by
6346 'data=N' where N is the number of seconds. Defaults to 30
6347 if not specified. See the tcp(7) man page for details.
6350 accept_filter httpready
6355 NAME: client_ip_max_connections
6357 LOC: Config.client_ip_max_connections
6360 Set an absolute limit on the number of connections a single
6361 client IP can use. Any more than this and Squid will begin to drop
6362 new connections from the client until it closes some links.
6364 Note that this is a global limit. It affects all HTTP, HTCP, Gopher and FTP
6365 connections from the client. For finer control use the ACL access controls.
6367 Requires client_db to be enabled (the default).
6369 WARNING: This may noticably slow down traffic received via external proxies
6370 or NAT devices and cause them to rebound error messages back to their clients.
6373 NAME: tcp_recv_bufsize
6377 LOC: Config.tcpRcvBufsz
6379 Size of receive buffer to set for TCP sockets. Probably just
6380 as easy to change your kernel's default. Set to zero to use
6381 the default buffer size.
6386 -----------------------------------------------------------------------------
6393 LOC: Adaptation::Icap::TheConfig.onoff
6396 If you want to enable the ICAP module support, set this to on.
6399 NAME: icap_connect_timeout
6402 LOC: Adaptation::Icap::TheConfig.connect_timeout_raw
6405 This parameter specifies how long to wait for the TCP connect to
6406 the requested ICAP server to complete before giving up and either
6407 terminating the HTTP transaction or bypassing the failure.
6409 The default for optional services is peer_connect_timeout.
6410 The default for essential services is connect_timeout.
6411 If this option is explicitly set, its value applies to all services.
6414 NAME: icap_io_timeout
6418 LOC: Adaptation::Icap::TheConfig.io_timeout_raw
6421 This parameter specifies how long to wait for an I/O activity on
6422 an established, active ICAP connection before giving up and
6423 either terminating the HTTP transaction or bypassing the
6426 The default is read_timeout.
6429 NAME: icap_service_failure_limit
6430 COMMENT: limit [in memory-depth time-units]
6431 TYPE: icap_service_failure_limit
6433 LOC: Adaptation::Icap::TheConfig
6436 The limit specifies the number of failures that Squid tolerates
6437 when establishing a new TCP connection with an ICAP service. If
6438 the number of failures exceeds the limit, the ICAP service is
6439 not used for new ICAP requests until it is time to refresh its
6442 A negative value disables the limit. Without the limit, an ICAP
6443 service will not be considered down due to connectivity failures
6444 between ICAP OPTIONS requests.
6446 Squid forgets ICAP service failures older than the specified
6447 value of memory-depth. The memory fading algorithm
6448 is approximate because Squid does not remember individual
6449 errors but groups them instead, splitting the option
6450 value into ten time slots of equal length.
6452 When memory-depth is 0 and by default this option has no
6453 effect on service failure expiration.
6455 Squid always forgets failures when updating service settings
6456 using an ICAP OPTIONS transaction, regardless of this option
6460 # suspend service usage after 10 failures in 5 seconds:
6461 icap_service_failure_limit 10 in 5 seconds
6464 NAME: icap_service_revival_delay
6467 LOC: Adaptation::Icap::TheConfig.service_revival_delay
6470 The delay specifies the number of seconds to wait after an ICAP
6471 OPTIONS request failure before requesting the options again. The
6472 failed ICAP service is considered "down" until fresh OPTIONS are
6475 The actual delay cannot be smaller than the hardcoded minimum
6476 delay of 30 seconds.
6479 NAME: icap_preview_enable
6483 LOC: Adaptation::Icap::TheConfig.preview_enable
6486 The ICAP Preview feature allows the ICAP server to handle the
6487 HTTP message by looking only at the beginning of the message body
6488 or even without receiving the body at all. In some environments,
6489 previews greatly speedup ICAP processing.
6491 During an ICAP OPTIONS transaction, the server may tell Squid what
6492 HTTP messages should be previewed and how big the preview should be.
6493 Squid will not use Preview if the server did not request one.
6495 To disable ICAP Preview for all ICAP services, regardless of
6496 individual ICAP server OPTIONS responses, set this option to "off".
6498 icap_preview_enable off
6501 NAME: icap_preview_size
6504 LOC: Adaptation::Icap::TheConfig.preview_size
6507 The default size of preview data to be sent to the ICAP server.
6508 -1 means no preview. This value might be overwritten on a per server
6509 basis by OPTIONS requests.
6512 NAME: icap_206_enable
6516 LOC: Adaptation::Icap::TheConfig.allow206_enable
6519 206 (Partial Content) responses is an ICAP extension that allows the
6520 ICAP agents to optionally combine adapted and original HTTP message
6521 content. The decision to combine is postponed until the end of the
6522 ICAP response. Squid supports Partial Content extension by default.
6524 Activation of the Partial Content extension is negotiated with each
6525 ICAP service during OPTIONS exchange. Most ICAP servers should handle
6526 negotation correctly even if they do not support the extension, but
6527 some might fail. To disable Partial Content support for all ICAP
6528 services and to avoid any negotiation, set this option to "off".
6534 NAME: icap_default_options_ttl
6537 LOC: Adaptation::Icap::TheConfig.default_options_ttl
6540 The default TTL value for ICAP OPTIONS responses that don't have
6541 an Options-TTL header.
6544 NAME: icap_persistent_connections
6548 LOC: Adaptation::Icap::TheConfig.reuse_connections
6551 Whether or not Squid should use persistent connections to
6555 NAME: adaptation_send_client_ip icap_send_client_ip
6557 IFDEF: USE_ADAPTATION
6559 LOC: Adaptation::Config::send_client_ip
6562 If enabled, Squid shares HTTP client IP information with adaptation
6563 services. For ICAP, Squid adds the X-Client-IP header to ICAP requests.
6564 For eCAP, Squid sets the libecap::metaClientIp transaction option.
6566 See also: adaptation_uses_indirect_client
6569 NAME: adaptation_send_username icap_send_client_username
6571 IFDEF: USE_ADAPTATION
6573 LOC: Adaptation::Config::send_username
6576 This sends authenticated HTTP client username (if available) to
6577 the adaptation service.
6579 For ICAP, the username value is encoded based on the
6580 icap_client_username_encode option and is sent using the header
6581 specified by the icap_client_username_header option.
6584 NAME: icap_client_username_header
6587 LOC: Adaptation::Icap::TheConfig.client_username_header
6588 DEFAULT: X-Client-Username
6590 ICAP request header name to use for adaptation_send_username.
6593 NAME: icap_client_username_encode
6597 LOC: Adaptation::Icap::TheConfig.client_username_encode
6600 Whether to base64 encode the authenticated client username.
6604 TYPE: icap_service_type
6606 LOC: Adaptation::Icap::TheConfig
6609 Defines a single ICAP service using the following format:
6611 icap_service id vectoring_point uri [option ...]
6614 an opaque identifier or name which is used to direct traffic to
6615 this specific service. Must be unique among all adaptation
6616 services in squid.conf.
6618 vectoring_point: reqmod_precache|reqmod_postcache|respmod_precache|respmod_postcache
6619 This specifies at which point of transaction processing the
6620 ICAP service should be activated. *_postcache vectoring points
6621 are not yet supported.
6623 uri: icap://servername:port/servicepath
6624 ICAP server and service location.
6626 ICAP does not allow a single service to handle both REQMOD and RESPMOD
6627 transactions. Squid does not enforce that requirement. You can specify
6628 services with the same service_url and different vectoring_points. You
6629 can even specify multiple identical services as long as their
6630 service_names differ.
6633 Service options are separated by white space. ICAP services support
6634 the following name=value options:
6637 If set to 'on' or '1', the ICAP service is treated as
6638 optional. If the service cannot be reached or malfunctions,
6639 Squid will try to ignore any errors and process the message as
6640 if the service was not enabled. No all ICAP errors can be
6641 bypassed. If set to 0, the ICAP service is treated as
6642 essential and all ICAP errors will result in an error page
6643 returned to the HTTP client.
6645 Bypass is off by default: services are treated as essential.
6648 If set to 'on' or '1', the ICAP service is allowed to
6649 dynamically change the current message adaptation plan by
6650 returning a chain of services to be used next. The services
6651 are specified using the X-Next-Services ICAP response header
6652 value, formatted as a comma-separated list of service names.
6653 Each named service should be configured in squid.conf. Other
6654 services are ignored. An empty X-Next-Services value results
6655 in an empty plan which ends the current adaptation.
6657 Dynamic adaptation plan may cross or cover multiple supported
6658 vectoring points in their natural processing order.
6660 Routing is not allowed by default: the ICAP X-Next-Services
6661 response header is ignored.
6664 Only has effect on split-stack systems. The default on those systems
6665 is to use IPv4-only connections. When set to 'on' this option will
6666 make Squid use IPv6-only connections to contact this ICAP service.
6668 on-overload=block|bypass|wait|force
6669 If the service Max-Connections limit has been reached, do
6670 one of the following for each new ICAP transaction:
6671 * block: send an HTTP error response to the client
6672 * bypass: ignore the "over-connected" ICAP service
6673 * wait: wait (in a FIFO queue) for an ICAP connection slot
6674 * force: proceed, ignoring the Max-Connections limit
6676 In SMP mode with N workers, each worker assumes the service
6677 connection limit is Max-Connections/N, even though not all
6678 workers may use a given service.
6680 The default value is "bypass" if service is bypassable,
6681 otherwise it is set to "wait".
6685 Use the given number as the Max-Connections limit, regardless
6686 of the Max-Connections value given by the service, if any.
6688 Older icap_service format without optional named parameters is
6689 deprecated but supported for backward compatibility.
6692 icap_service svcBlocker reqmod_precache icap://icap1.mydomain.net:1344/reqmod bypass=0
6693 icap_service svcLogger reqmod_precache icap://icap2.mydomain.net:1344/respmod routing=on
6697 TYPE: icap_class_type
6702 This deprecated option was documented to define an ICAP service
6703 chain, even though it actually defined a set of similar, redundant
6704 services, and the chains were not supported.
6706 To define a set of redundant services, please use the
6707 adaptation_service_set directive. For service chains, use
6708 adaptation_service_chain.
6712 TYPE: icap_access_type
6717 This option is deprecated. Please use adaptation_access, which
6718 has the same ICAP functionality, but comes with better
6719 documentation, and eCAP support.
6724 -----------------------------------------------------------------------------
6731 LOC: Adaptation::Ecap::TheConfig.onoff
6734 Controls whether eCAP support is enabled.
6738 TYPE: ecap_service_type
6740 LOC: Adaptation::Ecap::TheConfig
6743 Defines a single eCAP service
6745 ecap_service id vectoring_point uri [option ...]
6748 an opaque identifier or name which is used to direct traffic to
6749 this specific service. Must be unique among all adaptation
6750 services in squid.conf.
6752 vectoring_point: reqmod_precache|reqmod_postcache|respmod_precache|respmod_postcache
6753 This specifies at which point of transaction processing the
6754 eCAP service should be activated. *_postcache vectoring points
6755 are not yet supported.
6757 uri: ecap://vendor/service_name?custom&cgi=style¶meters=optional
6758 Squid uses the eCAP service URI to match this configuration
6759 line with one of the dynamically loaded services. Each loaded
6760 eCAP service must have a unique URI. Obtain the right URI from
6761 the service provider.
6764 Service options are separated by white space. eCAP services support
6765 the following name=value options:
6768 If set to 'on' or '1', the eCAP service is treated as optional.
6769 If the service cannot be reached or malfunctions, Squid will try
6770 to ignore any errors and process the message as if the service
6771 was not enabled. No all eCAP errors can be bypassed.
6772 If set to 'off' or '0', the eCAP service is treated as essential
6773 and all eCAP errors will result in an error page returned to the
6776 Bypass is off by default: services are treated as essential.
6779 If set to 'on' or '1', the eCAP service is allowed to
6780 dynamically change the current message adaptation plan by
6781 returning a chain of services to be used next.
6783 Dynamic adaptation plan may cross or cover multiple supported
6784 vectoring points in their natural processing order.
6786 Routing is not allowed by default.
6788 Older ecap_service format without optional named parameters is
6789 deprecated but supported for backward compatibility.
6793 ecap_service s1 reqmod_precache ecap://filters.R.us/leakDetector?on_error=block bypass=off
6794 ecap_service s2 respmod_precache ecap://filters.R.us/virusFilter config=/etc/vf.cfg bypass=on
6797 NAME: loadable_modules
6799 IFDEF: USE_LOADABLE_MODULES
6800 LOC: Config.loadable_module_names
6803 Instructs Squid to load the specified dynamic module(s) or activate
6804 preloaded module(s).
6806 loadable_modules @DEFAULT_PREFIX@/lib/MinimalAdapter.so
6810 MESSAGE ADAPTATION OPTIONS
6811 -----------------------------------------------------------------------------
6814 NAME: adaptation_service_set
6815 TYPE: adaptation_service_set_type
6816 IFDEF: USE_ADAPTATION
6821 Configures an ordered set of similar, redundant services. This is
6822 useful when hot standby or backup adaptation servers are available.
6824 adaptation_service_set set_name service_name1 service_name2 ...
6826 The named services are used in the set declaration order. The first
6827 applicable adaptation service from the set is used first. The next
6828 applicable service is tried if and only if the transaction with the
6829 previous service fails and the message waiting to be adapted is still
6832 When adaptation starts, broken services are ignored as if they were
6833 not a part of the set. A broken service is a down optional service.
6835 The services in a set must be attached to the same vectoring point
6836 (e.g., pre-cache) and use the same adaptation method (e.g., REQMOD).
6838 If all services in a set are optional then adaptation failures are
6839 bypassable. If all services in the set are essential, then a
6840 transaction failure with one service may still be retried using
6841 another service from the set, but when all services fail, the master
6842 transaction fails as well.
6844 A set may contain a mix of optional and essential services, but that
6845 is likely to lead to surprising results because broken services become
6846 ignored (see above), making previously bypassable failures fatal.
6847 Technically, it is the bypassability of the last failed service that
6850 See also: adaptation_access adaptation_service_chain
6853 adaptation_service_set svcBlocker urlFilterPrimary urlFilterBackup
6854 adaptation service_set svcLogger loggerLocal loggerRemote
6857 NAME: adaptation_service_chain
6858 TYPE: adaptation_service_chain_type
6859 IFDEF: USE_ADAPTATION
6864 Configures a list of complementary services that will be applied
6865 one-by-one, forming an adaptation chain or pipeline. This is useful
6866 when Squid must perform different adaptations on the same message.
6868 adaptation_service_chain chain_name service_name1 svc_name2 ...
6870 The named services are used in the chain declaration order. The first
6871 applicable adaptation service from the chain is used first. The next
6872 applicable service is applied to the successful adaptation results of
6873 the previous service in the chain.
6875 When adaptation starts, broken services are ignored as if they were
6876 not a part of the chain. A broken service is a down optional service.
6878 Request satisfaction terminates the adaptation chain because Squid
6879 does not currently allow declaration of RESPMOD services at the
6880 "reqmod_precache" vectoring point (see icap_service or ecap_service).
6882 The services in a chain must be attached to the same vectoring point
6883 (e.g., pre-cache) and use the same adaptation method (e.g., REQMOD).
6885 A chain may contain a mix of optional and essential services. If an
6886 essential adaptation fails (or the failure cannot be bypassed for
6887 other reasons), the master transaction fails. Otherwise, the failure
6888 is bypassed as if the failed adaptation service was not in the chain.
6890 See also: adaptation_access adaptation_service_set
6893 adaptation_service_chain svcRequest requestLogger urlFilter leakDetector
6896 NAME: adaptation_access
6897 TYPE: adaptation_access_type
6898 IFDEF: USE_ADAPTATION
6902 Sends an HTTP transaction to an ICAP or eCAP adaptation service.
6904 adaptation_access service_name allow|deny [!]aclname...
6905 adaptation_access set_name allow|deny [!]aclname...
6907 At each supported vectoring point, the adaptation_access
6908 statements are processed in the order they appear in this
6909 configuration file. Statements pointing to the following services
6910 are ignored (i.e., skipped without checking their ACL):
6912 - services serving different vectoring points
6913 - "broken-but-bypassable" services
6914 - "up" services configured to ignore such transactions
6915 (e.g., based on the ICAP Transfer-Ignore header).
6917 When a set_name is used, all services in the set are checked
6918 using the same rules, to find the first applicable one. See
6919 adaptation_service_set for details.
6921 If an access list is checked and there is a match, the
6922 processing stops: For an "allow" rule, the corresponding
6923 adaptation service is used for the transaction. For a "deny"
6924 rule, no adaptation service is activated.
6926 It is currently not possible to apply more than one adaptation
6927 service at the same vectoring point to the same HTTP transaction.
6929 See also: icap_service and ecap_service
6932 adaptation_access service_1 allow all
6935 NAME: adaptation_service_iteration_limit
6937 IFDEF: USE_ADAPTATION
6938 LOC: Adaptation::Config::service_iteration_limit
6941 Limits the number of iterations allowed when applying adaptation
6942 services to a message. If your longest adaptation set or chain
6943 may have more than 16 services, increase the limit beyond its
6944 default value of 16. If detecting infinite iteration loops sooner
6945 is critical, make the iteration limit match the actual number
6946 of services in your longest adaptation set or chain.
6948 Infinite adaptation loops are most likely with routing services.
6950 See also: icap_service routing=1
6953 NAME: adaptation_masterx_shared_names
6955 IFDEF: USE_ADAPTATION
6956 LOC: Adaptation::Config::masterx_shared_name
6959 For each master transaction (i.e., the HTTP request and response
6960 sequence, including all related ICAP and eCAP exchanges), Squid
6961 maintains a table of metadata. The table entries are (name, value)
6962 pairs shared among eCAP and ICAP exchanges. The table is destroyed
6963 with the master transaction.
6965 This option specifies the table entry names that Squid must accept
6966 from and forward to the adaptation transactions.
6968 An ICAP REQMOD or RESPMOD transaction may set an entry in the
6969 shared table by returning an ICAP header field with a name
6970 specified in adaptation_masterx_shared_names.
6972 An eCAP REQMOD or RESPMOD transaction may set an entry in the
6973 shared table by implementing the libecap::visitEachOption() API
6974 to provide an option with a name specified in
6975 adaptation_masterx_shared_names.
6977 Squid will store and forward the set entry to subsequent adaptation
6978 transactions within the same master transaction scope.
6980 Only one shared entry name is supported at this time.
6983 # share authentication information among ICAP services
6984 adaptation_masterx_shared_names X-Subscriber-ID
6990 LOC: Adaptation::Icap::TheConfig.repeat
6991 DEFAULT_IF_NONE: deny all
6993 This ACL determines which retriable ICAP transactions are
6994 retried. Transactions that received a complete ICAP response
6995 and did not have to consume or produce HTTP bodies to receive
6996 that response are usually retriable.
6998 icap_retry allow|deny [!]aclname ...
7000 Squid automatically retries some ICAP I/O timeouts and errors
7001 due to persistent connection race conditions.
7003 See also: icap_retry_limit
7006 NAME: icap_retry_limit
7009 LOC: Adaptation::Icap::TheConfig.repeat_limit
7012 Limits the number of retries allowed. When set to zero (default),
7013 no retries are allowed.
7015 Communication errors due to persistent connection race
7016 conditions are unavoidable, automatically retried, and do not
7017 count against this limit.
7019 See also: icap_retry
7025 -----------------------------------------------------------------------------
7028 NAME: check_hostnames
7031 LOC: Config.onoff.check_hostnames
7033 For security and stability reasons Squid can check
7034 hostnames for Internet standard RFC compliance. If you want
7035 Squid to perform these checks turn this directive on.
7038 NAME: allow_underscore
7041 LOC: Config.onoff.allow_underscore
7043 Underscore characters is not strictly allowed in Internet hostnames
7044 but nevertheless used by many sites. Set this to off if you want
7045 Squid to be strict about the standard.
7046 This check is performed only when check_hostnames is set to on.
7049 NAME: cache_dns_program
7051 IFDEF: USE_DNSSERVERS
7052 DEFAULT: @DEFAULT_DNSSERVER@
7053 LOC: Config.Program.dnsserver
7055 Specify the location of the executable for dnslookup process.
7059 TYPE: HelperChildConfig
7060 IFDEF: USE_DNSSERVERS
7061 DEFAULT: 32 startup=1 idle=1
7062 LOC: Config.dnsChildren
7064 The maximum number of processes spawn to service DNS name lookups.
7065 If you limit it too few Squid will have to wait for them to process
7066 a backlog of requests, slowing it down. If you allow too many they
7067 will use RAM and other system resources noticably.
7068 The maximum this may be safely set to is 32.
7070 The startup= and idle= options allow some measure of skew in your
7075 Sets a minimum of how many processes are to be spawned when Squid
7076 starts or reconfigures. When set to zero the first request will
7077 cause spawning of the first child process to handle it.
7079 Starting too few will cause an initial slowdown in traffic as Squid
7080 attempts to simultaneously spawn enough processes to cope.
7084 Sets a minimum of how many processes Squid is to try and keep available
7085 at all times. When traffic begins to rise above what the existing
7086 processes can handle this many more will be spawned up to the maximum
7087 configured. A minimum setting of 1 is required.
7090 NAME: dns_retransmit_interval
7093 LOC: Config.Timeout.idns_retransmit
7094 IFDEF: !USE_DNSSERVERS
7096 Initial retransmit interval for DNS queries. The interval is
7097 doubled each time all configured DNS servers have been tried.
7103 LOC: Config.Timeout.idns_query
7104 IFDEF: !USE_DNSSERVERS
7106 DNS Query timeout. If no response is received to a DNS query
7107 within this time all DNS servers for the queried domain
7108 are assumed to be unavailable.
7111 NAME: dns_packet_max
7114 LOC: Config.dns.packet_max
7115 IFDEF: !USE_DNSSERVERS
7117 Maximum number of bytes packet size to advertise via EDNS.
7118 Set to "none" to disable EDNS large packet support.
7120 For legacy reasons DNS UDP replies will default to 512 bytes which
7121 is too small for many responses. EDNS provides a means for Squid to
7122 negotiate receiving larger responses back immediately without having
7123 to failover with repeat requests. Responses larger than this limit
7124 will retain the old behaviour of failover to TCP DNS.
7126 Squid has no real fixed limit internally, but allowing packet sizes
7127 over 1500 bytes requires network jumbogram support and is usually not
7130 WARNING: The RFC also indicates that some older resolvers will reply
7131 with failure of the whole request if the extension is added. Some
7132 resolvers have already been identified which will reply with mangled
7133 EDNS response on occasion. Usually in response to many-KB jumbogram
7134 sizes being advertised by Squid.
7135 Squid will currently treat these both as an unable-to-resolve domain
7136 even if it would be resolvable without EDNS.
7143 LOC: Config.onoff.res_defnames
7145 Normally the RES_DEFNAMES resolver option is disabled
7146 (see res_init(3)). This prevents caches in a hierarchy
7147 from interpreting single-component hostnames locally. To allow
7148 Squid to handle single-component names, enable this option.
7151 NAME: dns_nameservers
7154 LOC: Config.dns_nameservers
7156 Use this if you want to specify a list of DNS name servers
7157 (IP addresses) to use instead of those given in your
7158 /etc/resolv.conf file.
7159 On Windows platforms, if no value is specified here or in
7160 the /etc/resolv.conf file, the list of DNS name servers are
7161 taken from the Windows registry, both static and dynamic DHCP
7162 configurations are supported.
7164 Example: dns_nameservers 10.0.0.1 192.172.0.4
7169 DEFAULT: @DEFAULT_HOSTS@
7170 LOC: Config.etcHostsPath
7172 Location of the host-local IP name-address associations
7173 database. Most Operating Systems have such a file on different
7175 - Un*X & Linux: /etc/hosts
7176 - Windows NT/2000: %SystemRoot%\system32\drivers\etc\hosts
7177 (%SystemRoot% value install default is c:\winnt)
7178 - Windows XP/2003: %SystemRoot%\system32\drivers\etc\hosts
7179 (%SystemRoot% value install default is c:\windows)
7180 - Windows 9x/Me: %windir%\hosts
7181 (%windir% value is usually c:\windows)
7182 - Cygwin: /etc/hosts
7184 The file contains newline-separated definitions, in the
7185 form ip_address_in_dotted_form name [name ...] names are
7186 whitespace-separated. Lines beginning with an hash (#)
7187 character are comments.
7189 The file is checked at startup and upon configuration.
7190 If set to 'none', it won't be checked.
7191 If append_domain is used, that domain will be added to
7192 domain-local (i.e. not containing any dot character) host
7198 LOC: Config.appendDomain
7201 Appends local domain name to hostnames without any dots in
7202 them. append_domain must begin with a period.
7204 Be warned there are now Internet names with no dots in
7205 them using only top-domain names, so setting this may
7206 cause some Internet sites to become unavailable.
7209 append_domain .yourdomain.com
7212 NAME: ignore_unknown_nameservers
7214 LOC: Config.onoff.ignore_unknown_nameservers
7216 IFDEF: !USE_DNSSERVERS
7218 By default Squid checks that DNS responses are received
7219 from the same IP addresses they are sent to. If they
7220 don't match, Squid ignores the response and writes a warning
7221 message to cache.log. You can allow responses from unknown
7222 nameservers by setting this option to 'off'.
7225 NAME: dns_v4_fallback
7228 LOC: Config.onoff.dns_require_A
7229 IFDEF: !USE_DNSSERVERS
7231 Standard practice with DNS is to lookup either A or AAAA records
7232 and use the results if it succeeds. Only looking up the other if
7233 the first attempt fails or otherwise produces no results.
7235 That policy however will cause squid to produce error pages for some
7236 servers that advertise AAAA but are unreachable over IPv6.
7238 If this is ON squid will always lookup both AAAA and A, using both.
7239 If this is OFF squid will lookup AAAA and only try A if none found.
7241 WARNING: There are some possibly unwanted side-effects with this on:
7242 *) Doubles the load placed by squid on the DNS network.
7243 *) May negatively impact connection delay times.
7249 LOC: Config.dns.v4_first
7250 IFDEF: !USE_DNSSERVERS
7252 With the IPv6 Internet being as fast or faster than IPv4 Internet
7253 for most networks Squid prefers to contact websites over IPv6.
7255 This option reverses the order of preference to make Squid contact
7256 dual-stack websites over IPv4 first. Squid will still perform both
7257 IPv6 and IPv4 DNS lookups before connecting.
7260 This option will restrict the situations under which IPv6
7261 connectivity is used (and tested). Hiding network problems
7262 which would otherwise be detected and warned about.
7266 COMMENT: (number of entries)
7269 LOC: Config.ipcache.size
7276 LOC: Config.ipcache.low
7283 LOC: Config.ipcache.high
7285 The size, low-, and high-water marks for the IP cache.
7288 NAME: fqdncache_size
7289 COMMENT: (number of entries)
7292 LOC: Config.fqdncache.size
7294 Maximum number of FQDN cache entries.
7299 -----------------------------------------------------------------------------
7306 LOC: Config.onoff.mem_pools
7308 If set, Squid will keep pools of allocated (but unused) memory
7309 available for future use. If memory is a premium on your
7310 system and you believe your malloc library outperforms Squid
7311 routines, disable this.
7314 NAME: memory_pools_limit
7318 LOC: Config.MemPools.limit
7320 Used only with memory_pools on:
7321 memory_pools_limit 50 MB
7323 If set to a non-zero value, Squid will keep at most the specified
7324 limit of allocated (but unused) memory in memory pools. All free()
7325 requests that exceed this limit will be handled by your malloc
7326 library. Squid does not pre-allocate any memory, just safe-keeps
7327 objects that otherwise would be free()d. Thus, it is safe to set
7328 memory_pools_limit to a reasonably high value even if your
7329 configuration will use less memory.
7331 If set to none, Squid will keep all memory it can. That is, there
7332 will be no limit on the total amount of memory used for safe-keeping.
7334 To disable memory allocation optimization, do not set
7335 memory_pools_limit to 0 or none. Set memory_pools to "off" instead.
7337 An overhead for maintaining memory pools is not taken into account
7338 when the limit is checked. This overhead is close to four bytes per
7339 object kept. However, pools may actually _save_ memory because of
7340 reduced memory thrashing in your malloc library.
7344 COMMENT: on|off|transparent|truncate|delete
7347 LOC: opt_forwarded_for
7349 If set to "on", Squid will append your client's IP address
7350 in the HTTP requests it forwards. By default it looks like:
7352 X-Forwarded-For: 192.1.2.3
7354 If set to "off", it will appear as
7356 X-Forwarded-For: unknown
7358 If set to "transparent", Squid will not alter the
7359 X-Forwarded-For header in any way.
7361 If set to "delete", Squid will delete the entire
7362 X-Forwarded-For header.
7364 If set to "truncate", Squid will remove all existing
7365 X-Forwarded-For entries, and place the client IP as the sole entry.
7368 NAME: cachemgr_passwd
7369 TYPE: cachemgrpasswd
7371 LOC: Config.passwd_list
7373 Specify passwords for cachemgr operations.
7375 Usage: cachemgr_passwd password action action ...
7377 Some valid actions are (see cache manager menu for a full list):
7417 * Indicates actions which will not be performed without a
7418 valid password, others can be performed if not listed here.
7420 To disable an action, set the password to "disable".
7421 To allow performing an action without a password, set the
7424 Use the keyword "all" to set the same password for all actions.
7427 cachemgr_passwd secret shutdown
7428 cachemgr_passwd lesssssssecret info stats/objects
7429 cachemgr_passwd disable all
7436 LOC: Config.onoff.client_db
7438 If you want to disable collecting per-client statistics,
7439 turn off client_db here.
7442 NAME: refresh_all_ims
7446 LOC: Config.onoff.refresh_all_ims
7448 When you enable this option, squid will always check
7449 the origin server for an update when a client sends an
7450 If-Modified-Since request. Many browsers use IMS
7451 requests when the user requests a reload, and this
7452 ensures those clients receive the latest version.
7454 By default (off), squid may return a Not Modified response
7455 based on the age of the cached version.
7458 NAME: reload_into_ims
7459 IFDEF: USE_HTTP_VIOLATIONS
7463 LOC: Config.onoff.reload_into_ims
7465 When you enable this option, client no-cache or ``reload''
7466 requests will be changed to If-Modified-Since requests.
7467 Doing this VIOLATES the HTTP standard. Enabling this
7468 feature could make you liable for problems which it
7471 see also refresh_pattern for a more selective approach.
7474 NAME: connect_retries
7476 LOC: Config.connect_retries
7479 This sets the maximum number of connection attempts made for each
7480 TCP connection. The connect_retries attempts must all still
7481 complete within the connection timeout period.
7483 The default is not to re-try if the first connection attempt fails.
7484 The (not recommended) maximum is 10 tries.
7486 A warning message will be generated if it is set to a too-high
7487 value and the configured value will be over-ridden.
7489 Note: These re-tries are in addition to forward_max_tries
7490 which limit how many different addresses may be tried to find
7494 NAME: retry_on_error
7496 LOC: Config.retry.onerror
7499 If set to ON Squid will automatically retry requests when
7500 receiving an error response with status 403 (Forbidden),
7501 500 (Internal Error), 501 or 503 (Service not available).
7502 Status 502 and 504 (Gateway errors) are always retried.
7504 This is mainly useful if you are in a complex cache hierarchy to
7505 work around access control errors.
7507 NOTE: This retry will attempt to find another working destination.
7508 Which is different from the server which just failed.
7511 NAME: as_whois_server
7513 LOC: Config.as_whois_server
7514 DEFAULT: whois.ra.net
7516 WHOIS server to query for AS numbers. NOTE: AS numbers are
7517 queried only when Squid starts up, not for every request.
7522 LOC: Config.onoff.offline
7525 Enable this option and Squid will never try to validate cached
7529 NAME: uri_whitespace
7530 TYPE: uri_whitespace
7531 LOC: Config.uri_whitespace
7534 What to do with requests that have whitespace characters in the
7537 strip: The whitespace characters are stripped out of the URL.
7538 This is the behavior recommended by RFC2396.
7539 deny: The request is denied. The user receives an "Invalid
7541 allow: The request is allowed and the URI is not changed. The
7542 whitespace characters remain in the URI. Note the
7543 whitespace is passed to redirector processes if they
7545 encode: The request is allowed and the whitespace characters are
7546 encoded according to RFC1738. This could be considered
7547 a violation of the HTTP/1.1
7548 RFC because proxies are not allowed to rewrite URI's.
7549 chop: The request is allowed and the URI is chopped at the
7550 first whitespace. This might also be considered a
7556 LOC: Config.chroot_dir
7559 Specifies a directory where Squid should do a chroot() while
7560 initializing. This also causes Squid to fully drop root
7561 privileges after initializing. This means, for example, if you
7562 use a HTTP port less than 1024 and try to reconfigure, you may
7563 get an error saying that Squid can not open the port.
7566 NAME: balance_on_multiple_ip
7568 LOC: Config.onoff.balance_on_multiple_ip
7571 Modern IP resolvers in squid sort lookup results by preferred access.
7572 By default squid will use these IP in order and only rotates to
7573 the next listed when the most preffered fails.
7575 Some load balancing servers based on round robin DNS have been
7576 found not to preserve user session state across requests
7577 to different IP addresses.
7579 Enabling this directive Squid rotates IP's per request.
7582 NAME: pipeline_prefetch
7584 LOC: Config.onoff.pipeline_prefetch
7587 To boost the performance of pipelined requests to closer
7588 match that of a non-proxied environment Squid can try to fetch
7589 up to two requests in parallel from a pipeline.
7591 Defaults to off for bandwidth management and access logging
7594 WARNING: pipelining breaks NTLM and Negotiate/Kerberos authentication.
7597 NAME: high_response_time_warning
7600 LOC: Config.warnings.high_rptm
7603 If the one-minute median response time exceeds this value,
7604 Squid prints a WARNING with debug level 0 to get the
7605 administrators attention. The value is in milliseconds.
7608 NAME: high_page_fault_warning
7610 LOC: Config.warnings.high_pf
7613 If the one-minute average page fault rate exceeds this
7614 value, Squid prints a WARNING with debug level 0 to get
7615 the administrators attention. The value is in page faults
7619 NAME: high_memory_warning
7621 LOC: Config.warnings.high_memory
7624 If the memory usage (as determined by mallinfo) exceeds
7625 this amount, Squid prints a WARNING with debug level 0 to get
7626 the administrators attention.
7629 NAME: sleep_after_fork
7630 COMMENT: (microseconds)
7632 LOC: Config.sleep_after_fork
7635 When this is set to a non-zero value, the main Squid process
7636 sleeps the specified number of microseconds after a fork()
7637 system call. This sleep may help the situation where your
7638 system reports fork() failures due to lack of (virtual)
7639 memory. Note, however, if you have a lot of child
7640 processes, these sleep delays will add up and your
7641 Squid will not service requests for some amount of time
7642 until all the child processes have been started.
7643 On Windows value less then 1000 (1 milliseconds) are
7647 NAME: windows_ipaddrchangemonitor
7648 IFDEF: _SQUID_MSWIN_
7652 LOC: Config.onoff.WIN32_IpAddrChangeMonitor
7654 On Windows Squid by default will monitor IP address changes and will
7655 reconfigure itself after any detected event. This is very useful for
7656 proxies connected to internet with dial-up interfaces.
7657 In some cases (a Proxy server acting as VPN gateway is one) it could be
7658 desiderable to disable this behaviour setting this to 'off'.
7659 Note: after changing this, Squid service must be restarted.
7664 IFDEF: USE_SQUID_EUI
7666 LOC: Eui::TheConfig.euiLookup
7668 Whether to lookup the EUI or MAC address of a connected client.
7671 NAME: max_filedescriptors max_filedesc
7674 LOC: Config.max_filedescriptors
7676 The maximum number of filedescriptors supported.
7678 The default "0" means Squid inherits the current ulimit setting.
7680 Note: Changing this requires a restart of Squid. Also
7681 not all comm loops supports large values.
7689 Number of main Squid processes or "workers" to fork and maintain.
7690 0: "no daemon" mode, like running "squid -N ..."
7691 1: "no SMP" mode, start one main Squid process daemon (default)
7692 N: start N main Squid process daemons (i.e., SMP mode)
7694 In SMP mode, each worker does nearly all what a single Squid daemon
7695 does (e.g., listen on http_port and forward HTTP requests).
7698 NAME: cpu_affinity_map
7699 TYPE: CpuAffinityMap
7700 LOC: Config.cpuAffinityMap
7703 Usage: cpu_affinity_map process_numbers=P1,P2,... cores=C1,C2,...
7705 Sets 1:1 mapping between Squid processes and CPU cores. For example,
7707 cpu_affinity_map process_numbers=1,2,3,4 cores=1,3,5,7
7709 affects processes 1 through 4 only and places them on the first
7710 four even cores, starting with core #1.
7712 CPU cores are numbered starting from 1. Requires support for
7713 sched_getaffinity(2) and sched_setaffinity(2) system calls.
7715 Multiple cpu_affinity_map options are merged.