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 documentation for the Squid configuration file.
34 This documentation can also be found online at:
35 http://www.squid-cache.org/Doc/config/
37 You may wish to look at the Squid home page and wiki for the
38 FAQ and other documentation:
39 http://www.squid-cache.org/
40 http://wiki.squid-cache.org/SquidFaq
41 http://wiki.squid-cache.org/ConfigExamples
43 This documentation shows what the defaults for various directives
44 happen to be. If you don't need to change the default, you should
45 leave the line out of your squid.conf in most cases.
47 In some cases "none" refers to no default setting at all,
48 while in other cases it refers to the value of the option
49 - the comments for that keyword indicate if this is the case.
54 Configuration options can be included using the "include" directive.
55 Include takes a list of files to include. Quoting and wildcards are
60 include /path/to/included/file/squid.acl.config
62 Includes can be nested up to a hard-coded depth of 16 levels.
63 This arbitrary restriction is to prevent recursive include references
64 from causing Squid entering an infinite loop whilst trying to load
67 Values with byte units
69 Squid accepts size units on some size related directives. All
70 such directives are documented with a default value displaying
73 Units accepted by Squid are:
75 KB - Kilobyte (1024 bytes)
79 Values with spaces, quotes, and other special characters
81 Squid supports directive parameters with spaces, quotes, and other
82 special characters. Surround such parameters with "double quotes". Use
83 the configuration_includes_quoted_values directive to enable or
86 Squid supports reading configuration option parameters from external
87 files using the syntax:
88 parameters("/path/filename")
90 acl whitelist dstdomain parameters("/etc/squid/whitelist.txt")
92 Conditional configuration
94 If-statements can be used to make configuration directives
98 ... regular configuration directives ...
100 ... regular configuration directives ...]
103 The else part is optional. The keywords "if", "else", and "endif"
104 must be typed on their own lines, as if they were regular
105 configuration directives.
107 NOTE: An else-if condition is not supported.
109 These individual conditions types are supported:
112 Always evaluates to true.
114 Always evaluates to false.
115 <integer> = <integer>
116 Equality comparison of two integer numbers.
121 The following SMP-related preprocessor macros can be used.
123 ${process_name} expands to the current Squid process "name"
124 (e.g., squid1, squid2, or cache1).
126 ${process_number} expands to the current Squid process
127 identifier, which is an integer number (e.g., 1, 2, 3) unique
128 across all Squid processes of the current service instance.
130 ${service_name} expands into the current Squid service instance
131 name identifier which is provided by -n on the command line.
135 # options still not yet ported from 2.7 to 3.x
136 NAME: broken_vary_encoding
139 This option is not yet supported by Squid-3.
145 This option is not yet supported by Squid-3.
151 This option is not yet supported by Squid-3.
154 NAME: external_refresh_check
157 This option is not yet supported by Squid-3.
160 NAME: location_rewrite_program location_rewrite_access location_rewrite_children location_rewrite_concurrency
163 This option is not yet supported by Squid-3.
166 NAME: refresh_stale_hit
169 This option is not yet supported by Squid-3.
172 # Options removed in 3.5
173 NAME: hierarchy_stoplist
176 Remove this line. Use always_direct or cache_peer_access ACLs instead if you need to prevent cache_peer use.
182 Remove this line. Use acls with access_log directives to control access logging
188 Remove this line. Use acls with icap_log directives to control icap logging
191 # Options Removed in 3.3
192 NAME: ignore_ims_on_miss
195 Remove this line. The HTTP/1.1 feature is now configured by 'cache_miss_revalidate'.
198 # Options Removed in 3.2
199 NAME: dns_v4_fallback
202 Remove this line. Squid performs a 'Happy Eyeballs' algorithm, the 'fallback' algorithm is no longer relevant.
205 NAME: emulate_httpd_log
208 Replace this with an access_log directive using the format 'common' or 'combined'.
214 Use a regular access.log with ACL limiting it to MISS events.
220 Remove this line. Configure FTP page display using the CSS controls in errorpages.css instead.
223 NAME: ignore_expect_100
226 Remove this line. The HTTP/1.1 feature is now fully supported by default.
232 Remove this option from your config. To log FQDN use %>A in the log format.
235 NAME: log_ip_on_direct
238 Remove this option from your config. To log server or peer names use %<A in the log format.
241 NAME: maximum_single_addr_tries
244 Replaced by connect_retries. The behaviour has changed, please read the documentation before altering.
247 NAME: referer_log referrer_log
250 Replace this with an access_log directive using the format 'referrer'.
256 Remove this line. The feature is supported by default in storage types where update is implemented.
259 NAME: url_rewrite_concurrency
262 Remove this line. Set the 'concurrency=' option of url_rewrite_children instead.
268 Replace this with an access_log directive using the format 'useragent'.
271 # Options Removed in 3.1
275 Remove this line. DNS is no longer tested on startup.
278 NAME: extension_methods
281 Remove this line. All valid methods for HTTP are accepted by default.
284 # 2.7 Options Removed/Replaced in 3.2
289 # 2.7 Options Removed/Replaced in 3.1
297 Remove this line. HTTP/1.1 is supported by default.
300 NAME: upgrade_http0.9
303 Remove this line. ICY/1.0 streaming protocol is supported by default.
306 NAME: zph_local zph_mode zph_option zph_parent zph_sibling
309 Alter these entries. Use the qos_flows directive instead.
312 # Options Removed in 3.0
316 Since squid-3.0 replace with request_header_access or reply_header_access
317 depending on whether you wish to match client requests or server replies.
320 NAME: httpd_accel_no_pmtu_disc
323 Since squid-3.0 use the 'disable-pmtu-discovery' flag on http_port instead.
326 NAME: wais_relay_host
329 Replace this line with 'cache_peer' configuration.
332 NAME: wais_relay_port
335 Replace this line with 'cache_peer' configuration.
339 OPTIONS FOR AUTHENTICATION
340 -----------------------------------------------------------------------------
349 This is used to define parameters for the various authentication
350 schemes supported by Squid.
352 format: auth_param scheme parameter [setting]
354 The order in which authentication schemes are presented to the client is
355 dependent on the order the scheme first appears in config file. IE
356 has a bug (it's not RFC 2617 compliant) in that it will use the basic
357 scheme if basic is the first entry presented, even if more secure
358 schemes are presented. For now use the order in the recommended
359 settings section below. If other browsers have difficulties (don't
360 recognize the schemes offered even if you are using basic) either
361 put basic first, or disable the other schemes (by commenting out their
364 Once an authentication scheme is fully configured, it can only be
365 shutdown by shutting squid down and restarting. Changes can be made on
366 the fly and activated with a reconfigure. I.E. You can change to a
367 different helper, but not unconfigure the helper completely.
369 Please note that while this directive defines how Squid processes
370 authentication it does not automatically activate authentication.
371 To use authentication you must in addition make use of ACLs based
372 on login name in http_access (proxy_auth, proxy_auth_regex or
373 external with %LOGIN used in the format tag). The browser will be
374 challenged for authentication on the first such acl encountered
375 in http_access processing and will also be re-challenged for new
376 login credentials if the request is being denied by a proxy_auth
379 WARNING: authentication can't be used in a transparently intercepting
380 proxy as the client then thinks it is talking to an origin server and
381 not the proxy. This is a limitation of bending the TCP/IP protocol to
382 transparently intercepting port 80, not a limitation in Squid.
383 Ports flagged 'transparent', 'intercept', or 'tproxy' have
384 authentication disabled.
386 === Parameters common to all schemes. ===
389 Specifies the command for the external authenticator.
391 By default, each authentication scheme is not used unless a
392 program is specified.
394 See http://wiki.squid-cache.org/Features/AddonHelpers for
395 more details on helper operations and creating your own.
398 Specifies a string to be append to request line format for
399 the authentication helper. "Quoted" format values may contain
400 spaces and logformat %macros. In theory, any logformat %macro
401 can be used. In practice, a %macro expands as a dash (-) if
402 the helper request is sent before the required macro
403 information is available to Squid.
405 By default, Squid uses request formats provided in
406 scheme-specific examples below (search for %credentials).
408 The expanded key_extras value is added to the Squid credentials
409 cache and, hence, will affect authentication. It can be used to
410 autenticate different users with identical user names (e.g.,
411 when user authentication depends on http_port).
413 Avoid adding frequently changing information to key_extras. For
414 example, if you add user source IP, and it changes frequently
415 in your environment, then max_user_ip ACL is going to treat
416 every user+IP combination as a unique "user", breaking the ACL
417 and wasting a lot of memory on those user records. It will also
418 force users to authenticate from scratch whenever their IP
422 Specifies the protection scope (aka realm name) which is to be
423 reported to the client for the authentication scheme. It is
424 commonly part of the text the user will see when prompted for
425 their username and password.
427 For Basic the default is "Squid proxy-caching web server".
428 For Digest there is no default, this parameter is mandatory.
429 For NTLM and Negotiate this parameter is ignored.
431 "children" numberofchildren [startup=N] [idle=N] [concurrency=N]
433 The maximum number of authenticator processes to spawn. If
434 you start too few Squid will have to wait for them to process
435 a backlog of credential verifications, slowing it down. When
436 password verifications are done via a (slow) network you are
437 likely to need lots of authenticator processes.
439 The startup= and idle= options permit some skew in the exact
440 amount run. A minimum of startup=N will begin during startup
441 and reconfigure. Squid will start more in groups of up to
442 idle=N in an attempt to meet traffic needs and to keep idle=N
443 free above those traffic needs up to the maximum.
445 The concurrency= option sets the number of concurrent requests
446 the helper can process. The default of 0 is used for helpers
447 who only supports one request at a time. Setting this to a
448 number greater than 0 changes the protocol used to include a
449 channel ID field first on the request/response line, allowing
450 multiple requests to be sent to the same helper in parallel
451 without waiting for the response.
453 Concurrency must not be set unless it's known the helper
454 supports the input format with channel-ID fields.
456 NOTE: NTLM and Negotiate schemes do not support concurrency
457 in the Squid code module even though some helpers can.
460 IF HAVE_AUTH_MODULE_BASIC
461 === Basic authentication parameters ===
464 HTTP uses iso-latin-1 as character set, while some
465 authentication backends such as LDAP expects UTF-8. If this is
466 set to on Squid will translate the HTTP iso-latin-1 charset to
467 UTF-8 before sending the username and password to the helper.
469 "credentialsttl" timetolive
470 Specifies how long squid assumes an externally validated
471 username:password pair is valid for - in other words how
472 often the helper program is called for that user. Set this
473 low to force revalidation with short lived passwords.
475 NOTE: setting this high does not impact your susceptibility
476 to replay attacks unless you are using an one-time password
477 system (such as SecureID). If you are using such a system,
478 you will be vulnerable to replay attacks unless you also
479 use the max_user_ip ACL in an http_access rule.
481 "casesensitive" on|off
482 Specifies if usernames are case sensitive. Most user databases
483 are case insensitive allowing the same username to be spelled
484 using both lower and upper case letters, but some are case
485 sensitive. This makes a big difference for user_max_ip ACL
486 processing and similar.
489 IF HAVE_AUTH_MODULE_DIGEST
490 === Digest authentication parameters ===
493 HTTP uses iso-latin-1 as character set, while some
494 authentication backends such as LDAP expects UTF-8. If this is
495 set to on Squid will translate the HTTP iso-latin-1 charset to
496 UTF-8 before sending the username and password to the helper.
498 "nonce_garbage_interval" timeinterval
499 Specifies the interval that nonces that have been issued
500 to client_agent's are checked for validity.
502 "nonce_max_duration" timeinterval
503 Specifies the maximum length of time a given nonce will be
506 "nonce_max_count" number
507 Specifies the maximum number of times a given nonce can be
510 "nonce_strictness" on|off
511 Determines if squid requires strict increment-by-1 behavior
512 for nonce counts, or just incrementing (off - for use when
513 user agents generate nonce counts that occasionally miss 1
514 (ie, 1,2,4,6)). Default off.
516 "check_nonce_count" on|off
517 This directive if set to off can disable the nonce count check
518 completely to work around buggy digest qop implementations in
519 certain mainstream browser versions. Default on to check the
520 nonce count to protect from authentication replay attacks.
522 "post_workaround" on|off
523 This is a workaround to certain buggy browsers who send an
524 incorrect request digest in POST requests when reusing the
525 same nonce as acquired earlier on a GET request.
528 IF HAVE_AUTH_MODULE_NEGOTIATE
529 === Negotiate authentication parameters ===
532 If you experience problems with PUT/POST requests when using
533 the this authentication scheme then you can try setting this
534 to off. This will cause Squid to forcibly close the connection
535 on the initial request where the browser asks which schemes
536 are supported by the proxy.
539 IF HAVE_AUTH_MODULE_NTLM
540 === NTLM authentication parameters ===
543 If you experience problems with PUT/POST requests when using
544 the this authentication scheme then you can try setting this
545 to off. This will cause Squid to forcibly close the connection
546 on the initial request where the browser asks which schemes
547 are supported by the proxy.
550 === Example Configuration ===
552 This configuration displays the recommended authentication scheme
553 order from most to least secure with recommended minimum configuration
554 settings for each scheme:
556 #auth_param negotiate program <uncomment and complete this line to activate>
557 #auth_param negotiate children 20 startup=0 idle=1
558 #auth_param negotiate keep_alive on
560 #auth_param digest program <uncomment and complete this line to activate>
561 #auth_param digest children 20 startup=0 idle=1
562 #auth_param digest realm Squid proxy-caching web server
563 #auth_param digest nonce_garbage_interval 5 minutes
564 #auth_param digest nonce_max_duration 30 minutes
565 #auth_param digest nonce_max_count 50
567 #auth_param ntlm program <uncomment and complete this line to activate>
568 #auth_param ntlm children 20 startup=0 idle=1
569 #auth_param ntlm keep_alive on
571 #auth_param basic program <uncomment and complete this line>
572 #auth_param basic children 5 startup=5 idle=1
573 #auth_param basic realm Squid proxy-caching web server
574 #auth_param basic credentialsttl 2 hours
577 NAME: authenticate_cache_garbage_interval
580 LOC: Config.authenticateGCInterval
582 The time period between garbage collection across the username cache.
583 This is a trade-off between memory utilization (long intervals - say
584 2 days) and CPU (short intervals - say 1 minute). Only change if you
588 NAME: authenticate_ttl
591 LOC: Config.authenticateTTL
593 The time a user & their credentials stay in the logged in
594 user cache since their last request. When the garbage
595 interval passes, all user credentials that have passed their
596 TTL are removed from memory.
599 NAME: authenticate_ip_ttl
601 LOC: Config.authenticateIpTTL
604 If you use proxy authentication and the 'max_user_ip' ACL,
605 this directive controls how long Squid remembers the IP
606 addresses associated with each user. Use a small value
607 (e.g., 60 seconds) if your users might change addresses
608 quickly, as is the case with dialup. You might be safe
609 using a larger value (e.g., 2 hours) in a corporate LAN
610 environment with relatively static address assignments.
615 -----------------------------------------------------------------------------
618 NAME: external_acl_type
619 TYPE: externalAclHelper
620 LOC: Config.externalAclHelperList
623 This option defines external acl classes using a helper program
624 to look up the status
626 external_acl_type name [options] FORMAT.. /path/to/helper [helper arguments..]
630 ttl=n TTL in seconds for cached results (defaults to 3600
633 TTL for cached negative lookups (default same
636 Maximum number of acl helper processes spawned to service
637 external acl lookups of this type. (default 20)
639 Minimum number of acl helper processes to spawn during
640 startup and reconfigure to service external acl lookups
641 of this type. (default 0)
643 Number of acl helper processes to keep ahead of traffic
644 loads. Squid will spawn this many at once whenever load
645 rises above the capabilities of existing processes.
646 Up to the value of children-max. (default 1)
647 concurrency=n concurrency level per process. Only used with helpers
648 capable of processing more than one query at a time.
649 cache=n limit the result cache size, default is 262144.
650 grace=n Percentage remaining of TTL where a refresh of a
651 cached entry should be initiated without needing to
652 wait for a new reply. (default is for no grace period)
653 protocol=2.5 Compatibility mode for Squid-2.5 external acl helpers
654 ipv4 / ipv6 IP protocol used to communicate with this helper.
655 The default is to auto-detect IPv6 and use it when available.
657 FORMAT specifications
659 %LOGIN Authenticated user login name
660 %EXT_USER Username from previous external acl
661 %EXT_LOG Log details from previous external acl
662 %EXT_TAG Tag from previous external acl
663 %IDENT Ident user name
665 %SRCPORT Client source port
668 %PROTO Requested URL scheme
670 %PATH Requested URL path
671 %METHOD Request method
672 %MYADDR Squid interface address
673 %MYPORT Squid http_port number
674 %PATH Requested URL-path (including query-string if any)
675 %USER_CERT SSL User certificate in PEM format
676 %USER_CERTCHAIN SSL User certificate chain in PEM format
677 %USER_CERT_xx SSL User certificate subject attribute xx
678 %USER_CA_CERT_xx SSL User certificate issuer attribute xx
679 %ssl::>sni SSL client SNI sent to Squid
680 %ssl::<cert_subject SSL server certificate DN
681 %ssl::<cert_issuer SSL server certificate issuer DN
683 %>{Header} HTTP request header "Header"
685 HTTP request header "Hdr" list member "member"
687 HTTP request header list member using ; as
688 list separator. ; can be any non-alphanumeric
691 %<{Header} HTTP reply header "Header"
693 HTTP reply header "Hdr" list member "member"
695 HTTP reply header list member using ; as
696 list separator. ; can be any non-alphanumeric
699 %ACL The name of the ACL being tested.
700 %DATA The ACL arguments. If not used then any arguments
701 is automatically added at the end of the line
703 NOTE: this will encode the arguments as one token,
704 whereas the default will pass each separately.
706 %% The percent sign. Useful for helpers which need
707 an unchanging input format.
710 General request syntax:
712 [channel-ID] FORMAT-values [acl-values ...]
715 FORMAT-values consists of transaction details expanded with
716 whitespace separation per the config file FORMAT specification
717 using the FORMAT macros listed above.
719 acl-values consists of any string specified in the referencing
720 config 'acl ... external' line. see the "acl external" directive.
722 Request values sent to the helper are URL escaped to protect
723 each value in requests against whitespaces.
725 If using protocol=2.5 then the request sent to the helper is not
726 URL escaped to protect against whitespace.
728 NOTE: protocol=3.0 is deprecated as no longer necessary.
730 When using the concurrency= option the protocol is changed by
731 introducing a query channel tag in front of the request/response.
732 The query channel tag is a number between 0 and concurrency-1.
733 This value must be echoed back unchanged to Squid as the first part
734 of the response relating to its request.
737 The helper receives lines expanded per the above format specification
738 and for each input line returns 1 line starting with OK/ERR/BH result
739 code and optionally followed by additional keywords with more details.
742 General result syntax:
744 [channel-ID] result keyword=value ...
746 Result consists of one of the codes:
749 the ACL test produced a match.
752 the ACL test does not produce a match.
755 An internal error occurred in the helper, preventing
756 a result being identified.
758 The meaning of 'a match' is determined by your squid.conf
759 access control configuration. See the Squid wiki for details.
763 user= The users name (login)
765 password= The users password (for login= cache_peer option)
767 message= Message describing the reason for this response.
768 Available as %o in error pages.
769 Useful on (ERR and BH results).
771 tag= Apply a tag to a request. Only sets a tag once,
772 does not alter existing tags.
774 log= String to be logged in access.log. Available as
775 %ea in logformat specifications.
777 clt_conn_tag= Associates a TAG with the client TCP connection.
778 Please see url_rewrite_program related documentation for
781 Any keywords may be sent on any response whether OK, ERR or BH.
783 All response keyword values need to be a single token with URL
784 escaping, or enclosed in double quotes (") and escaped using \ on
785 any double quotes or \ characters within the value. The wrapping
786 double quotes are removed before the value is interpreted by Squid.
787 \r and \n are also replace by CR and LF.
789 Some example key values:
793 user="J. \"Bob\" Smith"
800 DEFAULT: ssl::certHasExpired ssl_error X509_V_ERR_CERT_HAS_EXPIRED
801 DEFAULT: ssl::certNotYetValid ssl_error X509_V_ERR_CERT_NOT_YET_VALID
802 DEFAULT: ssl::certDomainMismatch ssl_error SQUID_X509_V_ERR_DOMAIN_MISMATCH
803 DEFAULT: ssl::certUntrusted ssl_error X509_V_ERR_INVALID_CA X509_V_ERR_SELF_SIGNED_CERT_IN_CHAIN X509_V_ERR_UNABLE_TO_VERIFY_LEAF_SIGNATURE X509_V_ERR_UNABLE_TO_GET_ISSUER_CERT X509_V_ERR_UNABLE_TO_GET_ISSUER_CERT_LOCALLY X509_V_ERR_CERT_UNTRUSTED
804 DEFAULT: ssl::certSelfSigned ssl_error X509_V_ERR_DEPTH_ZERO_SELF_SIGNED_CERT
807 DEFAULT: manager url_regex -i ^cache_object:// +i ^https?://[^/]+/squid-internal-mgr/
808 DEFAULT: localhost src 127.0.0.1/32 ::1
809 DEFAULT: to_localhost dst 127.0.0.0/8 0.0.0.0/32 ::1
810 DEFAULT_DOC: ACLs all, manager, localhost, and to_localhost are predefined.
812 Defining an Access List
814 Every access list definition must begin with an aclname and acltype,
815 followed by either type-specific arguments or a quoted filename that
818 acl aclname acltype argument ...
819 acl aclname acltype "file" ...
821 When using "file", the file should contain one item per line.
823 Some acl types supports options which changes their default behaviour.
824 The available options are:
826 -i,+i By default, regular expressions are CASE-SENSITIVE. To make them
827 case-insensitive, use the -i option. To return case-sensitive
828 use the +i option between patterns, or make a new ACL line
831 -n Disable lookups and address type conversions. If lookup or
832 conversion is required because the parameter type (IP or
833 domain name) does not match the message address type (domain
834 name or IP), then the ACL would immediately declare a mismatch
835 without any warnings or lookups.
837 -- Used to stop processing all options, in the case the first acl
838 value has '-' character as first character (for example the '-'
839 is a valid domain name)
841 Some acl types require suspending the current request in order
842 to access some external data source.
843 Those which do are marked with the tag [slow], those which
844 don't are marked as [fast].
845 See http://wiki.squid-cache.org/SquidFaq/SquidAcl
846 for further information
848 ***** ACL TYPES AVAILABLE *****
850 acl aclname src ip-address/mask ... # clients IP address [fast]
851 acl aclname src addr1-addr2/mask ... # range of addresses [fast]
852 acl aclname dst [-n] ip-address/mask ... # URL host's IP address [slow]
853 acl aclname localip ip-address/mask ... # IP address the client connected to [fast]
855 acl aclname arp mac-address ... (xx:xx:xx:xx:xx:xx notation)
856 # The arp ACL requires the special configure option --enable-arp-acl.
857 # Furthermore, the ARP ACL code is not portable to all operating systems.
858 # It works on Linux, Solaris, Windows, FreeBSD, and some
859 # other *BSD variants.
862 # NOTE: Squid can only determine the MAC address for clients that are on
863 # the same subnet. If the client is on a different subnet,
864 # then Squid cannot find out its MAC address.
866 acl aclname srcdomain .foo.com ...
867 # reverse lookup, from client IP [slow]
868 acl aclname dstdomain [-n] .foo.com ...
869 # Destination server from URL [fast]
870 acl aclname srcdom_regex [-i] \.foo\.com ...
871 # regex matching client name [slow]
872 acl aclname dstdom_regex [-n] [-i] \.foo\.com ...
873 # regex matching server [fast]
875 # For dstdomain and dstdom_regex a reverse lookup is tried if a IP
876 # based URL is used and no match is found. The name "none" is used
877 # if the reverse lookup fails.
879 acl aclname src_as number ...
880 acl aclname dst_as number ...
882 # Except for access control, AS numbers can be used for
883 # routing of requests to specific caches. Here's an
884 # example for routing all requests for AS#1241 and only
885 # those to mycache.mydomain.net:
886 # acl asexample dst_as 1241
887 # cache_peer_access mycache.mydomain.net allow asexample
888 # cache_peer_access mycache_mydomain.net deny all
890 acl aclname peername myPeer ...
892 # match against a named cache_peer entry
893 # set unique name= on cache_peer lines for reliable use.
895 acl aclname time [day-abbrevs] [h1:m1-h2:m2]
905 # h1:m1 must be less than h2:m2
907 acl aclname url_regex [-i] ^http:// ...
908 # regex matching on whole URL [fast]
909 acl aclname urllogin [-i] [^a-zA-Z0-9] ...
910 # regex matching on URL login field
911 acl aclname urlpath_regex [-i] \.gif$ ...
912 # regex matching on URL path [fast]
914 acl aclname port 80 70 21 0-1024... # destination TCP port [fast]
916 acl aclname localport 3128 ... # TCP port the client connected to [fast]
917 # NP: for interception mode this is usually '80'
919 acl aclname myportname 3128 ... # http(s)_port name [fast]
921 acl aclname proto HTTP FTP ... # request protocol [fast]
923 acl aclname method GET POST ... # HTTP request method [fast]
925 acl aclname http_status 200 301 500- 400-403 ...
926 # status code in reply [fast]
928 acl aclname browser [-i] regexp ...
929 # pattern match on User-Agent header (see also req_header below) [fast]
931 acl aclname referer_regex [-i] regexp ...
932 # pattern match on Referer header [fast]
933 # Referer is highly unreliable, so use with care
935 acl aclname ident username ...
936 acl aclname ident_regex [-i] pattern ...
937 # string match on ident output [slow]
938 # use REQUIRED to accept any non-null ident.
940 acl aclname proxy_auth [-i] username ...
941 acl aclname proxy_auth_regex [-i] pattern ...
942 # perform http authentication challenge to the client and match against
943 # supplied credentials [slow]
945 # takes a list of allowed usernames.
946 # use REQUIRED to accept any valid username.
948 # Will use proxy authentication in forward-proxy scenarios, and plain
949 # http authenticaiton in reverse-proxy scenarios
951 # NOTE: when a Proxy-Authentication header is sent but it is not
952 # needed during ACL checking the username is NOT logged
955 # NOTE: proxy_auth requires a EXTERNAL authentication program
956 # to check username/password combinations (see
957 # auth_param directive).
959 # NOTE: proxy_auth can't be used in a transparent/intercepting proxy
960 # as the browser needs to be configured for using a proxy in order
961 # to respond to proxy authentication.
963 acl aclname snmp_community string ...
964 # A community string to limit access to your SNMP Agent [fast]
967 # acl snmppublic snmp_community public
969 acl aclname maxconn number
970 # This will be matched when the client's IP address has
971 # more than <number> TCP connections established. [fast]
972 # NOTE: This only measures direct TCP links so X-Forwarded-For
973 # indirect clients are not counted.
975 acl aclname max_user_ip [-s] number
976 # This will be matched when the user attempts to log in from more
977 # than <number> different ip addresses. The authenticate_ip_ttl
978 # parameter controls the timeout on the ip entries. [fast]
979 # If -s is specified the limit is strict, denying browsing
980 # from any further IP addresses until the ttl has expired. Without
981 # -s Squid will just annoy the user by "randomly" denying requests.
982 # (the counter is reset each time the limit is reached and a
984 # NOTE: in acceleration mode or where there is mesh of child proxies,
985 # clients may appear to come from multiple addresses if they are
986 # going through proxy farms, so a limit of 1 may cause user problems.
988 acl aclname random probability
989 # Pseudo-randomly match requests. Based on the probability given.
990 # Probability may be written as a decimal (0.333), fraction (1/3)
991 # or ratio of matches:non-matches (3:5).
993 acl aclname req_mime_type [-i] mime-type ...
994 # regex match against the mime type of the request generated
995 # by the client. Can be used to detect file upload or some
996 # types HTTP tunneling requests [fast]
997 # NOTE: This does NOT match the reply. You cannot use this
998 # to match the returned file type.
1000 acl aclname req_header header-name [-i] any\.regex\.here
1001 # regex match against any of the known request headers. May be
1002 # thought of as a superset of "browser", "referer" and "mime-type"
1005 acl aclname rep_mime_type [-i] mime-type ...
1006 # regex match against the mime type of the reply received by
1007 # squid. Can be used to detect file download or some
1008 # types HTTP tunneling requests. [fast]
1009 # NOTE: This has no effect in http_access rules. It only has
1010 # effect in rules that affect the reply data stream such as
1011 # http_reply_access.
1013 acl aclname rep_header header-name [-i] any\.regex\.here
1014 # regex match against any of the known reply headers. May be
1015 # thought of as a superset of "browser", "referer" and "mime-type"
1018 acl aclname external class_name [arguments...]
1019 # external ACL lookup via a helper class defined by the
1020 # external_acl_type directive [slow]
1022 acl aclname user_cert attribute values...
1023 # match against attributes in a user SSL certificate
1024 # attribute is one of DN/C/O/CN/L/ST [fast]
1026 acl aclname ca_cert attribute values...
1027 # match against attributes a users issuing CA SSL certificate
1028 # attribute is one of DN/C/O/CN/L/ST [fast]
1030 acl aclname ext_user username ...
1031 acl aclname ext_user_regex [-i] pattern ...
1032 # string match on username returned by external acl helper [slow]
1033 # use REQUIRED to accept any non-null user name.
1035 acl aclname tag tagvalue ...
1036 # string match on tag returned by external acl helper [fast]
1037 # DEPRECATED. Only the first tag will match with this ACL.
1038 # Use the 'note' ACL instead for handling multiple tag values.
1040 acl aclname hier_code codename ...
1041 # string match against squid hierarchy code(s); [fast]
1042 # e.g., DIRECT, PARENT_HIT, NONE, etc.
1044 # NOTE: This has no effect in http_access rules. It only has
1045 # effect in rules that affect the reply data stream such as
1046 # http_reply_access.
1048 acl aclname note name [value ...]
1049 # match transaction annotation [fast]
1050 # Without values, matches any annotation with a given name.
1051 # With value(s), matches any annotation with a given name that
1052 # also has one of the given values.
1053 # Names and values are compared using a string equality test.
1054 # Annotation sources include note and adaptation_meta directives
1055 # as well as helper and eCAP responses.
1057 acl aclname adaptation_service service ...
1058 # Matches the name of any icap_service, ecap_service,
1059 # adaptation_service_set, or adaptation_service_chain that Squid
1060 # has used (or attempted to use) for the master transaction.
1061 # This ACL must be defined after the corresponding adaptation
1062 # service is named in squid.conf. This ACL is usable with
1063 # adaptation_meta because it starts matching immediately after
1064 # the service has been selected for adaptation.
1067 acl aclname ssl_error errorname
1068 # match against SSL certificate validation error [fast]
1070 # For valid error names see in @DEFAULT_ERROR_DIR@/templates/error-details.txt
1073 # The following can be used as shortcuts for certificate properties:
1074 # [ssl::]certHasExpired: the "not after" field is in the past
1075 # [ssl::]certNotYetValid: the "not before" field is in the future
1076 # [ssl::]certUntrusted: The certificate issuer is not to be trusted.
1077 # [ssl::]certSelfSigned: The certificate is self signed.
1078 # [ssl::]certDomainMismatch: The certificate CN domain does not
1079 # match the name the name of the host we are connecting to.
1081 # The ssl::certHasExpired, ssl::certNotYetValid, ssl::certDomainMismatch,
1082 # ssl::certUntrusted, and ssl::certSelfSigned can also be used as
1083 # predefined ACLs, just like the 'all' ACL.
1085 # NOTE: The ssl_error ACL is only supported with sslproxy_cert_error,
1086 # sslproxy_cert_sign, and sslproxy_cert_adapt options.
1088 acl aclname server_cert_fingerprint [-sha1] fingerprint
1089 # match against server SSL certificate fingerprint [fast]
1091 # The fingerprint is the digest of the DER encoded version
1092 # of the whole certificate. The user should use the form: XX:XX:...
1093 # Optional argument specifies the digest algorithm to use.
1094 # The SHA1 digest algorithm is the default and is currently
1095 # the only algorithm supported (-sha1).
1097 acl aclname at_step step
1098 # match against the current step during ssl_bump evaluation [fast]
1099 # Never matches and should not be used outside the ssl_bump context.
1101 # At each SslBump step, Squid evaluates ssl_bump directives to find
1102 # the next bumping action (e.g., peek or splice). Valid SslBump step
1103 # values and the corresponding ssl_bump evaluation moments are:
1104 # SslBump1: After getting TCP-level and HTTP CONNECT info.
1105 # SslBump2: After getting SSL Client Hello info.
1106 # SslBump3: After getting SSL Server Hello info.
1108 acl aclname any-of acl1 acl2 ...
1109 # match any one of the acls [fast or slow]
1110 # The first matching ACL stops further ACL evaluation.
1112 # ACLs from multiple any-of lines with the same name are ORed.
1113 # For example, A = (a1 or a2) or (a3 or a4) can be written as
1114 # acl A any-of a1 a2
1115 # acl A any-of a3 a4
1117 # This group ACL is fast if all evaluated ACLs in the group are fast
1118 # and slow otherwise.
1120 acl aclname all-of acl1 acl2 ...
1121 # match all of the acls [fast or slow]
1122 # The first mismatching ACL stops further ACL evaluation.
1124 # ACLs from multiple all-of lines with the same name are ORed.
1125 # For example, B = (b1 and b2) or (b3 and b4) can be written as
1126 # acl B all-of b1 b2
1127 # acl B all-of b3 b4
1129 # This group ACL is fast if all evaluated ACLs in the group are fast
1130 # and slow otherwise.
1133 acl macaddress arp 09:00:2b:23:45:67
1134 acl myexample dst_as 1241
1135 acl password proxy_auth REQUIRED
1136 acl fileupload req_mime_type -i ^multipart/form-data$
1137 acl javascript rep_mime_type -i ^application/x-javascript$
1141 # Recommended minimum configuration:
1144 # Example rule allowing access from your local networks.
1145 # Adapt to list your (internal) IP networks from where browsing
1147 acl localnet src 10.0.0.0/8 # RFC1918 possible internal network
1148 acl localnet src 172.16.0.0/12 # RFC1918 possible internal network
1149 acl localnet src 192.168.0.0/16 # RFC1918 possible internal network
1150 acl localnet src fc00::/7 # RFC 4193 local private network range
1151 acl localnet src fe80::/10 # RFC 4291 link-local (directly plugged) machines
1153 acl SSL_ports port 443
1154 acl Safe_ports port 80 # http
1155 acl Safe_ports port 21 # ftp
1156 acl Safe_ports port 443 # https
1157 acl Safe_ports port 70 # gopher
1158 acl Safe_ports port 210 # wais
1159 acl Safe_ports port 1025-65535 # unregistered ports
1160 acl Safe_ports port 280 # http-mgmt
1161 acl Safe_ports port 488 # gss-http
1162 acl Safe_ports port 591 # filemaker
1163 acl Safe_ports port 777 # multiling http
1164 acl CONNECT method CONNECT
1168 NAME: follow_x_forwarded_for
1170 IFDEF: FOLLOW_X_FORWARDED_FOR
1171 LOC: Config.accessList.followXFF
1172 DEFAULT_IF_NONE: deny all
1173 DEFAULT_DOC: X-Forwarded-For header will be ignored.
1175 Allowing or Denying the X-Forwarded-For header to be followed to
1176 find the original source of a request.
1178 Requests may pass through a chain of several other proxies
1179 before reaching us. The X-Forwarded-For header will contain a
1180 comma-separated list of the IP addresses in the chain, with the
1181 rightmost address being the most recent.
1183 If a request reaches us from a source that is allowed by this
1184 configuration item, then we consult the X-Forwarded-For header
1185 to see where that host received the request from. If the
1186 X-Forwarded-For header contains multiple addresses, we continue
1187 backtracking until we reach an address for which we are not allowed
1188 to follow the X-Forwarded-For header, or until we reach the first
1189 address in the list. For the purpose of ACL used in the
1190 follow_x_forwarded_for directive the src ACL type always matches
1191 the address we are testing and srcdomain matches its rDNS.
1193 The end result of this process is an IP address that we will
1194 refer to as the indirect client address. This address may
1195 be treated as the client address for access control, ICAP, delay
1196 pools and logging, depending on the acl_uses_indirect_client,
1197 icap_uses_indirect_client, delay_pool_uses_indirect_client,
1198 log_uses_indirect_client and tproxy_uses_indirect_client options.
1200 This clause only supports fast acl types.
1201 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1203 SECURITY CONSIDERATIONS:
1205 Any host for which we follow the X-Forwarded-For header
1206 can place incorrect information in the header, and Squid
1207 will use the incorrect information as if it were the
1208 source address of the request. This may enable remote
1209 hosts to bypass any access control restrictions that are
1210 based on the client's source addresses.
1214 acl localhost src 127.0.0.1
1215 acl my_other_proxy srcdomain .proxy.example.com
1216 follow_x_forwarded_for allow localhost
1217 follow_x_forwarded_for allow my_other_proxy
1220 NAME: acl_uses_indirect_client
1223 IFDEF: FOLLOW_X_FORWARDED_FOR
1225 LOC: Config.onoff.acl_uses_indirect_client
1227 Controls whether the indirect client address
1228 (see follow_x_forwarded_for) is used instead of the
1229 direct client address in acl matching.
1231 NOTE: maxconn ACL considers direct TCP links and indirect
1232 clients will always have zero. So no match.
1235 NAME: delay_pool_uses_indirect_client
1238 IFDEF: FOLLOW_X_FORWARDED_FOR&&USE_DELAY_POOLS
1240 LOC: Config.onoff.delay_pool_uses_indirect_client
1242 Controls whether the indirect client address
1243 (see follow_x_forwarded_for) is used instead of the
1244 direct client address in delay pools.
1247 NAME: log_uses_indirect_client
1250 IFDEF: FOLLOW_X_FORWARDED_FOR
1252 LOC: Config.onoff.log_uses_indirect_client
1254 Controls whether the indirect client address
1255 (see follow_x_forwarded_for) is used instead of the
1256 direct client address in the access log.
1259 NAME: tproxy_uses_indirect_client
1262 IFDEF: FOLLOW_X_FORWARDED_FOR&&LINUX_NETFILTER
1264 LOC: Config.onoff.tproxy_uses_indirect_client
1266 Controls whether the indirect client address
1267 (see follow_x_forwarded_for) is used instead of the
1268 direct client address when spoofing the outgoing client.
1270 This has no effect on requests arriving in non-tproxy
1273 SECURITY WARNING: Usage of this option is dangerous
1274 and should not be used trivially. Correct configuration
1275 of follow_x_forewarded_for with a limited set of trusted
1276 sources is required to prevent abuse of your proxy.
1279 NAME: spoof_client_ip
1281 LOC: Config.accessList.spoof_client_ip
1283 DEFAULT_DOC: Allow spoofing on all TPROXY traffic.
1285 Control client IP address spoofing of TPROXY traffic based on
1286 defined access lists.
1288 spoof_client_ip allow|deny [!]aclname ...
1290 If there are no "spoof_client_ip" lines present, the default
1291 is to "allow" spoofing of any suitable request.
1293 Note that the cache_peer "no-tproxy" option overrides this ACL.
1295 This clause supports fast acl types.
1296 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1301 LOC: Config.accessList.http
1302 DEFAULT_IF_NONE: deny all
1303 DEFAULT_DOC: Deny, unless rules exist in squid.conf.
1305 Allowing or Denying access based on defined access lists
1307 Access to the HTTP port:
1308 http_access allow|deny [!]aclname ...
1310 NOTE on default values:
1312 If there are no "access" lines present, the default is to deny
1315 If none of the "access" lines cause a match, the default is the
1316 opposite of the last line in the list. If the last line was
1317 deny, the default is allow. Conversely, if the last line
1318 is allow, the default will be deny. For these reasons, it is a
1319 good idea to have an "deny all" entry at the end of your access
1320 lists to avoid potential confusion.
1322 This clause supports both fast and slow acl types.
1323 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1328 # Recommended minimum Access Permission configuration:
1330 # Deny requests to certain unsafe ports
1331 http_access deny !Safe_ports
1333 # Deny CONNECT to other than secure SSL ports
1334 http_access deny CONNECT !SSL_ports
1336 # Only allow cachemgr access from localhost
1337 http_access allow localhost manager
1338 http_access deny manager
1340 # We strongly recommend the following be uncommented to protect innocent
1341 # web applications running on the proxy server who think the only
1342 # one who can access services on "localhost" is a local user
1343 #http_access deny to_localhost
1346 # INSERT YOUR OWN RULE(S) HERE TO ALLOW ACCESS FROM YOUR CLIENTS
1349 # Example rule allowing access from your local networks.
1350 # Adapt localnet in the ACL section to list your (internal) IP networks
1351 # from where browsing should be allowed
1352 http_access allow localnet
1353 http_access allow localhost
1355 # And finally deny all other access to this proxy
1356 http_access deny all
1360 NAME: adapted_http_access http_access2
1362 LOC: Config.accessList.adapted_http
1364 DEFAULT_DOC: Allow, unless rules exist in squid.conf.
1366 Allowing or Denying access based on defined access lists
1368 Essentially identical to http_access, but runs after redirectors
1369 and ICAP/eCAP adaptation. Allowing access control based on their
1372 If not set then only http_access is used.
1375 NAME: http_reply_access
1377 LOC: Config.accessList.reply
1379 DEFAULT_DOC: Allow, unless rules exist in squid.conf.
1381 Allow replies to client requests. This is complementary to http_access.
1383 http_reply_access allow|deny [!] aclname ...
1385 NOTE: if there are no access lines present, the default is to allow
1388 If none of the access lines cause a match the opposite of the
1389 last line will apply. Thus it is good practice to end the rules
1390 with an "allow all" or "deny all" entry.
1392 This clause supports both fast and slow acl types.
1393 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1398 LOC: Config.accessList.icp
1400 DEFAULT_DOC: Deny, unless rules exist in squid.conf.
1402 Allowing or Denying access to the ICP port based on defined
1405 icp_access allow|deny [!]aclname ...
1407 NOTE: The default if no icp_access lines are present is to
1408 deny all traffic. This default may cause problems with peers
1411 This clause only supports fast acl types.
1412 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1414 # Allow ICP queries from local networks only
1415 #icp_access allow localnet
1416 #icp_access deny all
1422 LOC: Config.accessList.htcp
1424 DEFAULT_DOC: Deny, unless rules exist in squid.conf.
1426 Allowing or Denying access to the HTCP port based on defined
1429 htcp_access allow|deny [!]aclname ...
1431 See also htcp_clr_access for details on access control for
1432 cache purge (CLR) HTCP messages.
1434 NOTE: The default if no htcp_access lines are present is to
1435 deny all traffic. This default may cause problems with peers
1436 using the htcp option.
1438 This clause only supports fast acl types.
1439 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1441 # Allow HTCP queries from local networks only
1442 #htcp_access allow localnet
1443 #htcp_access deny all
1446 NAME: htcp_clr_access
1449 LOC: Config.accessList.htcp_clr
1451 DEFAULT_DOC: Deny, unless rules exist in squid.conf.
1453 Allowing or Denying access to purge content using HTCP based
1454 on defined access lists.
1455 See htcp_access for details on general HTCP access control.
1457 htcp_clr_access allow|deny [!]aclname ...
1459 This clause only supports fast acl types.
1460 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1462 # Allow HTCP CLR requests from trusted peers
1463 acl htcp_clr_peer src 192.0.2.2 2001:DB8::2
1464 htcp_clr_access allow htcp_clr_peer
1465 htcp_clr_access deny all
1470 LOC: Config.accessList.miss
1472 DEFAULT_DOC: Allow, unless rules exist in squid.conf.
1474 Determins whether network access is permitted when satisfying a request.
1477 to force your neighbors to use you as a sibling instead of
1480 acl localclients src 192.0.2.0/24 2001:DB8::a:0/64
1481 miss_access deny !localclients
1482 miss_access allow all
1484 This means only your local clients are allowed to fetch relayed/MISS
1485 replies from the network and all other clients can only fetch cached
1488 The default for this setting allows all clients who passed the
1489 http_access rules to relay via this proxy.
1491 This clause only supports fast acl types.
1492 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1495 NAME: ident_lookup_access
1499 DEFAULT_DOC: Unless rules exist in squid.conf, IDENT is not fetched.
1500 LOC: Ident::TheConfig.identLookup
1502 A list of ACL elements which, if matched, cause an ident
1503 (RFC 931) lookup to be performed for this request. For
1504 example, you might choose to always perform ident lookups
1505 for your main multi-user Unix boxes, but not for your Macs
1506 and PCs. By default, ident lookups are not performed for
1509 To enable ident lookups for specific client addresses, you
1510 can follow this example:
1512 acl ident_aware_hosts src 198.168.1.0/24
1513 ident_lookup_access allow ident_aware_hosts
1514 ident_lookup_access deny all
1516 Only src type ACL checks are fully supported. A srcdomain
1517 ACL might work at times, but it will not always provide
1520 This clause only supports fast acl types.
1521 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1524 NAME: reply_body_max_size
1525 COMMENT: size [acl acl...]
1528 DEFAULT_DOC: No limit is applied.
1529 LOC: Config.ReplyBodySize
1531 This option specifies the maximum size of a reply body. It can be
1532 used to prevent users from downloading very large files, such as
1533 MP3's and movies. When the reply headers are received, the
1534 reply_body_max_size lines are processed, and the first line where
1535 all (if any) listed ACLs are true is used as the maximum body size
1538 This size is checked twice. First when we get the reply headers,
1539 we check the content-length value. If the content length value exists
1540 and is larger than the allowed size, the request is denied and the
1541 user receives an error message that says "the request or reply
1542 is too large." If there is no content-length, and the reply
1543 size exceeds this limit, the client's connection is just closed
1544 and they will receive a partial reply.
1546 WARNING: downstream caches probably can not detect a partial reply
1547 if there is no content-length header, so they will cache
1548 partial responses and give them out as hits. You should NOT
1549 use this option if you have downstream caches.
1551 WARNING: A maximum size smaller than the size of squid's error messages
1552 will cause an infinite loop and crash squid. Ensure that the smallest
1553 non-zero value you use is greater that the maximum header size plus
1554 the size of your largest error page.
1556 If you set this parameter none (the default), there will be
1559 Configuration Format is:
1560 reply_body_max_size SIZE UNITS [acl ...]
1562 reply_body_max_size 10 MB
1568 -----------------------------------------------------------------------------
1571 NAME: http_port ascii_port
1576 Usage: port [mode] [options]
1577 hostname:port [mode] [options]
1578 1.2.3.4:port [mode] [options]
1580 The socket addresses where Squid will listen for HTTP client
1581 requests. You may specify multiple socket addresses.
1582 There are three forms: port alone, hostname with port, and
1583 IP address with port. If you specify a hostname or IP
1584 address, Squid binds the socket to that specific
1585 address. Most likely, you do not need to bind to a specific
1586 address, so you can use the port number alone.
1588 If you are running Squid in accelerator mode, you
1589 probably want to listen on port 80 also, or instead.
1591 The -a command line option may be used to specify additional
1592 port(s) where Squid listens for proxy request. Such ports will
1593 be plain proxy ports with no options.
1595 You may specify multiple socket addresses on multiple lines.
1599 intercept Support for IP-Layer interception of
1600 outgoing requests without browser settings.
1601 NP: disables authentication and IPv6 on the port.
1603 tproxy Support Linux TPROXY for spoofing outgoing
1604 connections using the client IP address.
1605 NP: disables authentication and maybe IPv6 on the port.
1607 accel Accelerator / reverse proxy mode
1609 ssl-bump For each CONNECT request allowed by ssl_bump ACLs,
1610 establish secure connection with the client and with
1611 the server, decrypt HTTPS messages as they pass through
1612 Squid, and treat them as unencrypted HTTP messages,
1613 becoming the man-in-the-middle.
1615 The ssl_bump option is required to fully enable
1616 bumping of CONNECT requests.
1618 Omitting the mode flag causes default forward proxy mode to be used.
1621 Accelerator Mode Options:
1623 defaultsite=domainname
1624 What to use for the Host: header if it is not present
1625 in a request. Determines what site (not origin server)
1626 accelerators should consider the default.
1628 no-vhost Disable using HTTP/1.1 Host header for virtual domain support.
1630 protocol= Protocol to reconstruct accelerated and intercepted
1631 requests with. Defaults to HTTP/1.1 for http_port and
1632 HTTPS/1.1 for https_port.
1633 When an unsupported value is configured Squid will
1634 produce a FATAL error.
1635 Values: HTTP or HTTP/1.1, HTTPS or HTTPS/1.1
1637 vport Virtual host port support. Using the http_port number
1638 instead of the port passed on Host: headers.
1640 vport=NN Virtual host port support. Using the specified port
1641 number instead of the port passed on Host: headers.
1644 Act as if this Squid is the origin server.
1645 This currently means generate new Date: and Expires:
1646 headers on HIT instead of adding Age:.
1648 ignore-cc Ignore request Cache-Control headers.
1650 WARNING: This option violates HTTP specifications if
1651 used in non-accelerator setups.
1653 allow-direct Allow direct forwarding in accelerator mode. Normally
1654 accelerated requests are denied direct forwarding as if
1655 never_direct was used.
1657 WARNING: this option opens accelerator mode to security
1658 vulnerabilities usually only affecting in interception
1659 mode. Make sure to protect forwarding with suitable
1660 http_access rules when using this.
1663 SSL Bump Mode Options:
1664 In addition to these options ssl-bump requires TLS/SSL options.
1666 generate-host-certificates[=<on|off>]
1667 Dynamically create SSL server certificates for the
1668 destination hosts of bumped CONNECT requests.When
1669 enabled, the cert and key options are used to sign
1670 generated certificates. Otherwise generated
1671 certificate will be selfsigned.
1672 If there is a CA certificate lifetime of the generated
1673 certificate equals lifetime of the CA certificate. If
1674 generated certificate is selfsigned lifetime is three
1676 This option is enabled by default when ssl-bump is used.
1677 See the ssl-bump option above for more information.
1679 dynamic_cert_mem_cache_size=SIZE
1680 Approximate total RAM size spent on cached generated
1681 certificates. If set to zero, caching is disabled. The
1682 default value is 4MB.
1686 cert= Path to SSL certificate (PEM format).
1688 key= Path to SSL private key file (PEM format)
1689 if not specified, the certificate file is
1690 assumed to be a combined certificate and
1693 version= The version of SSL/TLS supported
1694 1 automatic (default)
1701 cipher= Colon separated list of supported ciphers.
1702 NOTE: some ciphers such as EDH ciphers depend on
1703 additional settings. If those settings are
1704 omitted the ciphers may be silently ignored
1705 by the OpenSSL library.
1707 options= Various SSL implementation options. The most important
1709 NO_SSLv2 Disallow the use of SSLv2
1710 NO_SSLv3 Disallow the use of SSLv3
1711 NO_TLSv1 Disallow the use of TLSv1.0
1712 NO_TLSv1_1 Disallow the use of TLSv1.1
1713 NO_TLSv1_2 Disallow the use of TLSv1.2
1714 SINGLE_DH_USE Always create a new key when using
1715 temporary/ephemeral DH key exchanges
1716 ALL Enable various bug workarounds
1717 suggested as "harmless" by OpenSSL
1718 Be warned that this reduces SSL/TLS
1719 strength to some attacks.
1720 See OpenSSL SSL_CTX_set_options documentation for a
1721 complete list of options.
1723 clientca= File containing the list of CAs to use when
1724 requesting a client certificate.
1726 cafile= File containing additional CA certificates to
1727 use when verifying client certificates. If unset
1728 clientca will be used.
1730 capath= Directory containing additional CA certificates
1731 and CRL lists to use when verifying client certificates.
1733 crlfile= File of additional CRL lists to use when verifying
1734 the client certificate, in addition to CRLs stored in
1735 the capath. Implies VERIFY_CRL flag below.
1737 dhparams= File containing DH parameters for temporary/ephemeral
1738 DH key exchanges. See OpenSSL documentation for details
1739 on how to create this file.
1740 WARNING: EDH ciphers will be silently disabled if this
1743 sslflags= Various flags modifying the use of SSL:
1745 Don't request client certificates
1746 immediately, but wait until acl processing
1747 requires a certificate (not yet implemented).
1749 Don't use the default CA lists built in
1752 Don't allow for session reuse. Each connection
1753 will result in a new SSL session.
1755 Verify CRL lists when accepting client
1758 Verify CRL lists for all certificates in the
1759 client certificate chain.
1761 sslcontext= SSL session ID context identifier.
1765 connection-auth[=on|off]
1766 use connection-auth=off to tell Squid to prevent
1767 forwarding Microsoft connection oriented authentication
1768 (NTLM, Negotiate and Kerberos)
1770 disable-pmtu-discovery=
1771 Control Path-MTU discovery usage:
1772 off lets OS decide on what to do (default).
1773 transparent disable PMTU discovery when transparent
1775 always disable always PMTU discovery.
1777 In many setups of transparently intercepting proxies
1778 Path-MTU discovery can not work on traffic towards the
1779 clients. This is the case when the intercepting device
1780 does not fully track connections and fails to forward
1781 ICMP must fragment messages to the cache server. If you
1782 have such setup and experience that certain clients
1783 sporadically hang or never complete requests set
1784 disable-pmtu-discovery option to 'transparent'.
1786 name= Specifies a internal name for the port. Defaults to
1787 the port specification (port or addr:port)
1789 tcpkeepalive[=idle,interval,timeout]
1790 Enable TCP keepalive probes of idle connections.
1791 In seconds; idle is the initial time before TCP starts
1792 probing the connection, interval how often to probe, and
1793 timeout the time before giving up.
1795 If you run Squid on a dual-homed machine with an internal
1796 and an external interface we recommend you to specify the
1797 internal address:port in http_port. This way Squid will only be
1798 visible on the internal address.
1802 # Squid normally listens to port 3128
1803 http_port @DEFAULT_HTTP_PORT@
1813 Usage: [ip:]port cert=certificate.pem [key=key.pem] [mode] [options...]
1815 The socket address where Squid will listen for client requests made
1816 over TLS or SSL connections. Commonly referred to as HTTPS.
1818 This is most useful for situations where you are running squid in
1819 accelerator mode and you want to do the SSL work at the accelerator level.
1821 You may specify multiple socket addresses on multiple lines,
1822 each with their own SSL certificate and/or options.
1826 accel Accelerator / reverse proxy mode
1828 intercept Support for IP-Layer interception of
1829 outgoing requests without browser settings.
1830 NP: disables authentication and IPv6 on the port.
1832 tproxy Support Linux TPROXY for spoofing outgoing
1833 connections using the client IP address.
1834 NP: disables authentication and maybe IPv6 on the port.
1836 ssl-bump For each intercepted connection allowed by ssl_bump
1837 ACLs, establish a secure connection with the client and with
1838 the server, decrypt HTTPS messages as they pass through
1839 Squid, and treat them as unencrypted HTTP messages,
1840 becoming the man-in-the-middle.
1842 An "ssl_bump server-first" match is required to
1843 fully enable bumping of intercepted SSL connections.
1845 Requires tproxy or intercept.
1847 Omitting the mode flag causes default forward proxy mode to be used.
1850 See http_port for a list of generic options
1855 cert= Path to SSL certificate (PEM format).
1857 key= Path to SSL private key file (PEM format)
1858 if not specified, the certificate file is
1859 assumed to be a combined certificate and
1862 version= The version of SSL/TLS supported
1863 1 automatic (default)
1868 cipher= Colon separated list of supported ciphers.
1870 options= Various SSL engine options. The most important
1872 NO_SSLv2 Disallow the use of SSLv2
1873 NO_SSLv3 Disallow the use of SSLv3
1874 NO_TLSv1 Disallow the use of TLSv1
1875 SINGLE_DH_USE Always create a new key when using
1876 temporary/ephemeral DH key exchanges
1877 See src/ssl_support.c or OpenSSL SSL_CTX_set_options
1878 documentation for a complete list of options.
1880 clientca= File containing the list of CAs to use when
1881 requesting a client certificate.
1883 cafile= File containing additional CA certificates to
1884 use when verifying client certificates. If unset
1885 clientca will be used.
1887 capath= Directory containing additional CA certificates
1888 and CRL lists to use when verifying client certificates.
1890 crlfile= File of additional CRL lists to use when verifying
1891 the client certificate, in addition to CRLs stored in
1892 the capath. Implies VERIFY_CRL flag below.
1894 dhparams= File containing DH parameters for temporary/ephemeral
1897 sslflags= Various flags modifying the use of SSL:
1899 Don't request client certificates
1900 immediately, but wait until acl processing
1901 requires a certificate (not yet implemented).
1903 Don't use the default CA lists built in
1906 Don't allow for session reuse. Each connection
1907 will result in a new SSL session.
1909 Verify CRL lists when accepting client
1912 Verify CRL lists for all certificates in the
1913 client certificate chain.
1915 sslcontext= SSL session ID context identifier.
1917 generate-host-certificates[=<on|off>]
1918 Dynamically create SSL server certificates for the
1919 destination hosts of bumped SSL requests.When
1920 enabled, the cert and key options are used to sign
1921 generated certificates. Otherwise generated
1922 certificate will be selfsigned.
1923 If there is CA certificate life time of generated
1924 certificate equals lifetime of CA certificate. If
1925 generated certificate is selfsigned lifetime is three
1927 This option is enabled by default when SslBump is used.
1928 See the sslBump option above for more information.
1930 dynamic_cert_mem_cache_size=SIZE
1931 Approximate total RAM size spent on cached generated
1932 certificates. If set to zero, caching is disabled. The
1933 default value is 4MB.
1935 See http_port for a list of available options.
1938 NAME: tcp_outgoing_tos tcp_outgoing_ds tcp_outgoing_dscp
1941 LOC: Ip::Qos::TheConfig.tosToServer
1943 Allows you to select a TOS/Diffserv value for packets outgoing
1944 on the server side, based on an ACL.
1946 tcp_outgoing_tos ds-field [!]aclname ...
1948 Example where normal_service_net uses the TOS value 0x00
1949 and good_service_net uses 0x20
1951 acl normal_service_net src 10.0.0.0/24
1952 acl good_service_net src 10.0.1.0/24
1953 tcp_outgoing_tos 0x00 normal_service_net
1954 tcp_outgoing_tos 0x20 good_service_net
1956 TOS/DSCP values really only have local significance - so you should
1957 know what you're specifying. For more information, see RFC2474,
1958 RFC2475, and RFC3260.
1960 The TOS/DSCP byte must be exactly that - a octet value 0 - 255, or
1961 "default" to use whatever default your host has. Note that in
1962 practice often only multiples of 4 is usable as the two rightmost bits
1963 have been redefined for use by ECN (RFC 3168 section 23.1).
1965 Processing proceeds in the order specified, and stops at first fully
1968 Only fast ACLs are supported.
1971 NAME: clientside_tos
1974 LOC: Ip::Qos::TheConfig.tosToClient
1976 Allows you to select a TOS/Diffserv value for packets being transmitted
1977 on the client-side, based on an ACL.
1979 clientside_tos ds-field [!]aclname ...
1981 Example where normal_service_net uses the TOS value 0x00
1982 and good_service_net uses 0x20
1984 acl normal_service_net src 10.0.0.0/24
1985 acl good_service_net src 10.0.1.0/24
1986 clientside_tos 0x00 normal_service_net
1987 clientside_tos 0x20 good_service_net
1989 Note: This feature is incompatible with qos_flows. Any TOS values set here
1990 will be overwritten by TOS values in qos_flows.
1993 NAME: tcp_outgoing_mark
1995 IFDEF: SO_MARK&&USE_LIBCAP
1997 LOC: Ip::Qos::TheConfig.nfmarkToServer
1999 Allows you to apply a Netfilter mark value to outgoing packets
2000 on the server side, based on an ACL.
2002 tcp_outgoing_mark mark-value [!]aclname ...
2004 Example where normal_service_net uses the mark value 0x00
2005 and good_service_net uses 0x20
2007 acl normal_service_net src 10.0.0.0/24
2008 acl good_service_net src 10.0.1.0/24
2009 tcp_outgoing_mark 0x00 normal_service_net
2010 tcp_outgoing_mark 0x20 good_service_net
2012 Only fast ACLs are supported.
2015 NAME: clientside_mark
2017 IFDEF: SO_MARK&&USE_LIBCAP
2019 LOC: Ip::Qos::TheConfig.nfmarkToClient
2021 Allows you to apply a Netfilter mark value to packets being transmitted
2022 on the client-side, based on an ACL.
2024 clientside_mark mark-value [!]aclname ...
2026 Example where normal_service_net uses the mark value 0x00
2027 and good_service_net uses 0x20
2029 acl normal_service_net src 10.0.0.0/24
2030 acl good_service_net src 10.0.1.0/24
2031 clientside_mark 0x00 normal_service_net
2032 clientside_mark 0x20 good_service_net
2034 Note: This feature is incompatible with qos_flows. Any mark values set here
2035 will be overwritten by mark values in qos_flows.
2042 LOC: Ip::Qos::TheConfig
2044 Allows you to select a TOS/DSCP value to mark outgoing
2045 connections to the client, based on where the reply was sourced.
2046 For platforms using netfilter, allows you to set a netfilter mark
2047 value instead of, or in addition to, a TOS value.
2049 By default this functionality is disabled. To enable it with the default
2050 settings simply use "qos_flows mark" or "qos_flows tos". Default
2051 settings will result in the netfilter mark or TOS value being copied
2052 from the upstream connection to the client. Note that it is the connection
2053 CONNMARK value not the packet MARK value that is copied.
2055 It is not currently possible to copy the mark or TOS value from the
2056 client to the upstream connection request.
2058 TOS values really only have local significance - so you should
2059 know what you're specifying. For more information, see RFC2474,
2060 RFC2475, and RFC3260.
2062 The TOS/DSCP byte must be exactly that - a octet value 0 - 255. Note that
2063 in practice often only multiples of 4 is usable as the two rightmost bits
2064 have been redefined for use by ECN (RFC 3168 section 23.1).
2066 Mark values can be any unsigned 32-bit integer value.
2068 This setting is configured by setting the following values:
2070 tos|mark Whether to set TOS or netfilter mark values
2072 local-hit=0xFF Value to mark local cache hits.
2074 sibling-hit=0xFF Value to mark hits from sibling peers.
2076 parent-hit=0xFF Value to mark hits from parent peers.
2078 miss=0xFF[/mask] Value to mark cache misses. Takes precedence
2079 over the preserve-miss feature (see below), unless
2080 mask is specified, in which case only the bits
2081 specified in the mask are written.
2083 The TOS variant of the following features are only possible on Linux
2084 and require your kernel to be patched with the TOS preserving ZPH
2085 patch, available from http://zph.bratcheda.org
2086 No patch is needed to preserve the netfilter mark, which will work
2087 with all variants of netfilter.
2089 disable-preserve-miss
2090 This option disables the preservation of the TOS or netfilter
2091 mark. By default, the existing TOS or netfilter mark value of
2092 the response coming from the remote server will be retained
2093 and masked with miss-mark.
2094 NOTE: in the case of a netfilter mark, the mark must be set on
2095 the connection (using the CONNMARK target) not on the packet
2099 Allows you to mask certain bits in the TOS or mark value
2100 received from the remote server, before copying the value to
2101 the TOS sent towards clients.
2102 Default for tos: 0xFF (TOS from server is not changed).
2103 Default for mark: 0xFFFFFFFF (mark from server is not changed).
2105 All of these features require the --enable-zph-qos compilation flag
2106 (enabled by default). Netfilter marking also requires the
2107 libnetfilter_conntrack libraries (--with-netfilter-conntrack) and
2108 libcap 2.09+ (--with-libcap).
2112 NAME: tcp_outgoing_address
2115 DEFAULT_DOC: Address selection is performed by the operating system.
2116 LOC: Config.accessList.outgoing_address
2118 Allows you to map requests to different outgoing IP addresses
2119 based on the username or source address of the user making
2122 tcp_outgoing_address ipaddr [[!]aclname] ...
2125 Forwarding clients with dedicated IPs for certain subnets.
2127 acl normal_service_net src 10.0.0.0/24
2128 acl good_service_net src 10.0.2.0/24
2130 tcp_outgoing_address 2001:db8::c001 good_service_net
2131 tcp_outgoing_address 10.1.0.2 good_service_net
2133 tcp_outgoing_address 2001:db8::beef normal_service_net
2134 tcp_outgoing_address 10.1.0.1 normal_service_net
2136 tcp_outgoing_address 2001:db8::1
2137 tcp_outgoing_address 10.1.0.3
2139 Processing proceeds in the order specified, and stops at first fully
2142 Squid will add an implicit IP version test to each line.
2143 Requests going to IPv4 websites will use the outgoing 10.1.0.* addresses.
2144 Requests going to IPv6 websites will use the outgoing 2001:db8:* addresses.
2147 NOTE: The use of this directive using client dependent ACLs is
2148 incompatible with the use of server side persistent connections. To
2149 ensure correct results it is best to set server_persistent_connections
2150 to off when using this directive in such configurations.
2152 NOTE: The use of this directive to set a local IP on outgoing TCP links
2153 is incompatible with using TPROXY to set client IP out outbound TCP links.
2154 When needing to contact peers use the no-tproxy cache_peer option and the
2155 client_dst_passthru directive re-enable normal forwarding such as this.
2159 NAME: host_verify_strict
2162 LOC: Config.onoff.hostStrictVerify
2164 Regardless of this option setting, when dealing with intercepted
2165 traffic, Squid always verifies that the destination IP address matches
2166 the Host header domain or IP (called 'authority form URL').
2168 This enforcement is performed to satisfy a MUST-level requirement in
2169 RFC 2616 section 14.23: "The Host field value MUST represent the naming
2170 authority of the origin server or gateway given by the original URL".
2173 Squid always responds with an HTTP 409 (Conflict) error
2174 page and logs a security warning if there is no match.
2176 Squid verifies that the destination IP address matches
2177 the Host header for forward-proxy and reverse-proxy traffic
2178 as well. For those traffic types, Squid also enables the
2179 following checks, comparing the corresponding Host header
2180 and Request-URI components:
2182 * The host names (domain or IP) must be identical,
2183 but valueless or missing Host header disables all checks.
2184 For the two host names to match, both must be either IP
2187 * Port numbers must be identical, but if a port is missing
2188 the scheme-default port is assumed.
2191 When set to OFF (the default):
2192 Squid allows suspicious requests to continue but logs a
2193 security warning and blocks caching of the response.
2195 * Forward-proxy traffic is not checked at all.
2197 * Reverse-proxy traffic is not checked at all.
2199 * Intercepted traffic which passes verification is handled
2200 according to client_dst_passthru.
2202 * Intercepted requests which fail verification are sent
2203 to the client original destination instead of DIRECT.
2204 This overrides 'client_dst_passthru off'.
2206 For now suspicious intercepted CONNECT requests are always
2207 responded to with an HTTP 409 (Conflict) error page.
2212 As described in CVE-2009-0801 when the Host: header alone is used
2213 to determine the destination of a request it becomes trivial for
2214 malicious scripts on remote websites to bypass browser same-origin
2215 security policy and sandboxing protections.
2217 The cause of this is that such applets are allowed to perform their
2218 own HTTP stack, in which case the same-origin policy of the browser
2219 sandbox only verifies that the applet tries to contact the same IP
2220 as from where it was loaded at the IP level. The Host: header may
2221 be different from the connected IP and approved origin.
2225 NAME: client_dst_passthru
2228 LOC: Config.onoff.client_dst_passthru
2230 With NAT or TPROXY intercepted traffic Squid may pass the request
2231 directly to the original client destination IP or seek a faster
2232 source using the HTTP Host header.
2234 Using Host to locate alternative servers can provide faster
2235 connectivity with a range of failure recovery options.
2236 But can also lead to connectivity trouble when the client and
2237 server are attempting stateful interactions unaware of the proxy.
2239 This option (on by default) prevents alternative DNS entries being
2240 located to send intercepted traffic DIRECT to an origin server.
2241 The clients original destination IP and port will be used instead.
2243 Regardless of this option setting, when dealing with intercepted
2244 traffic Squid will verify the Host: header and any traffic which
2245 fails Host verification will be treated as if this option were ON.
2247 see host_verify_strict for details on the verification process.
2252 -----------------------------------------------------------------------------
2255 NAME: ssl_unclean_shutdown
2259 LOC: Config.SSL.unclean_shutdown
2261 Some browsers (especially MSIE) bugs out on SSL shutdown
2268 LOC: Config.SSL.ssl_engine
2271 The OpenSSL engine to use. You will need to set this if you
2272 would like to use hardware SSL acceleration for example.
2275 NAME: sslproxy_client_certificate
2278 LOC: Config.ssl_client.cert
2281 Client SSL Certificate to use when proxying https:// URLs
2284 NAME: sslproxy_client_key
2287 LOC: Config.ssl_client.key
2290 Client SSL Key to use when proxying https:// URLs
2293 NAME: sslproxy_version
2296 DEFAULT_DOC: automatic SSL/TLS version negotiation
2297 LOC: Config.ssl_client.version
2300 SSL version level to use when proxying https:// URLs
2302 The versions of SSL/TLS supported:
2304 1 automatic (default)
2312 NAME: sslproxy_options
2315 LOC: Config.ssl_client.options
2318 SSL implementation options to use when proxying https:// URLs
2320 The most important being:
2322 NO_SSLv2 Disallow the use of SSLv2
2323 NO_SSLv3 Disallow the use of SSLv3
2324 NO_TLSv1 Disallow the use of TLSv1.0
2325 NO_TLSv1_1 Disallow the use of TLSv1.1
2326 NO_TLSv1_2 Disallow the use of TLSv1.2
2328 Always create a new key when using temporary/ephemeral
2331 Disable use of RFC5077 session tickets. Some servers
2332 may have problems understanding the TLS extension due
2333 to ambiguous specification in RFC4507.
2334 ALL Enable various bug workarounds suggested as "harmless"
2335 by OpenSSL. Be warned that this may reduce SSL/TLS
2336 strength to some attacks.
2338 See the OpenSSL SSL_CTX_set_options documentation for a
2339 complete list of possible options.
2342 NAME: sslproxy_cipher
2345 LOC: Config.ssl_client.cipher
2348 SSL cipher list to use when proxying https:// URLs
2350 Colon separated list of supported ciphers.
2353 NAME: sslproxy_cafile
2356 LOC: Config.ssl_client.cafile
2359 file containing CA certificates to use when verifying server
2360 certificates while proxying https:// URLs
2363 NAME: sslproxy_capath
2366 LOC: Config.ssl_client.capath
2369 directory containing CA certificates to use when verifying
2370 server certificates while proxying https:// URLs
2373 NAME: sslproxy_session_ttl
2376 LOC: Config.SSL.session_ttl
2379 Sets the timeout value for SSL sessions
2382 NAME: sslproxy_session_cache_size
2385 LOC: Config.SSL.sessionCacheSize
2388 Sets the cache size to use for ssl session
2393 TYPE: sslproxy_ssl_bump
2394 LOC: Config.accessList.ssl_bump
2395 DEFAULT_DOC: Become a TCP tunnel without decrypting proxied traffic.
2398 This option is consulted when a CONNECT request is received on
2399 an http_port (or a new connection is intercepted at an
2400 https_port), provided that port was configured with an ssl-bump
2401 flag. The subsequent data on the connection is either treated as
2402 HTTPS and decrypted OR tunneled at TCP level without decryption,
2403 depending on the first matching bumping "action".
2405 ssl_bump <action> [!]acl ...
2407 The following bumping actions are currently supported:
2410 Become a TCP tunnel without decrypting proxied traffic.
2411 This is the default action.
2414 Establish a secure connection with the server and, using a
2415 mimicked server certificate, with the client.
2418 Receive client (step SslBump1) or server (step SslBump2)
2419 certificate while preserving the possibility of splicing the
2420 connection. Peeking at the server certificate (during step 2)
2421 usually precludes bumping of the connection at step 3.
2424 Receive client (step SslBump1) or server (step SslBump2)
2425 certificate while preserving the possibility of bumping the
2426 connection. Staring at the server certificate (during step 2)
2427 usually precludes splicing of the connection at step 3.
2430 Close client and server connections.
2432 Backward compatibility actions available at step SslBump1:
2435 Bump the connection. Establish a secure connection with the
2436 client first, then connect to the server. This old mode does
2437 not allow Squid to mimic server SSL certificate and does not
2438 work with intercepted SSL connections.
2441 Bump the connection. Establish a secure connection with the
2442 server first, then establish a secure connection with the
2443 client, using a mimicked server certificate. Works with both
2444 CONNECT requests and intercepted SSL connections, but does
2445 not allow to make decisions based on SSL handshake info.
2448 Decide whether to bump or splice the connection based on
2449 client-to-squid and server-to-squid SSL hello messages.
2453 Same as the "splice" action.
2455 All ssl_bump rules are evaluated at each of the supported bumping
2456 steps. Rules with actions that are impossible at the current step are
2457 ignored. The first matching ssl_bump action wins and is applied at the
2458 end of the current step. If no rules match, the splice action is used.
2459 See the at_step ACL for a list of the supported SslBump steps.
2461 This clause supports both fast and slow acl types.
2462 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
2464 See also: http_port ssl-bump, https_port ssl-bump, and acl at_step.
2467 # Example: Bump all requests except those originating from
2468 # localhost or those going to example.com.
2470 acl broken_sites dstdomain .example.com
2471 ssl_bump splice localhost
2472 ssl_bump splice broken_sites
2476 NAME: sslproxy_flags
2479 LOC: Config.ssl_client.flags
2482 Various flags modifying the use of SSL while proxying https:// URLs:
2483 DONT_VERIFY_PEER Accept certificates that fail verification.
2484 For refined control, see sslproxy_cert_error.
2485 NO_DEFAULT_CA Don't use the default CA list built in
2489 NAME: sslproxy_cert_error
2492 DEFAULT_DOC: Server certificate errors terminate the transaction.
2493 LOC: Config.ssl_client.cert_error
2496 Use this ACL to bypass server certificate validation errors.
2498 For example, the following lines will bypass all validation errors
2499 when talking to servers for example.com. All other
2500 validation errors will result in ERR_SECURE_CONNECT_FAIL error.
2502 acl BrokenButTrustedServers dstdomain example.com
2503 sslproxy_cert_error allow BrokenButTrustedServers
2504 sslproxy_cert_error deny all
2506 This clause only supports fast acl types.
2507 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
2508 Using slow acl types may result in server crashes
2510 Without this option, all server certificate validation errors
2511 terminate the transaction to protect Squid and the client.
2513 SQUID_X509_V_ERR_INFINITE_VALIDATION error cannot be bypassed
2514 but should not happen unless your OpenSSL library is buggy.
2517 Bypassing validation errors is dangerous because an
2518 error usually implies that the server cannot be trusted
2519 and the connection may be insecure.
2521 See also: sslproxy_flags and DONT_VERIFY_PEER.
2524 NAME: sslproxy_cert_sign
2527 POSTSCRIPTUM: signUntrusted ssl::certUntrusted
2528 POSTSCRIPTUM: signSelf ssl::certSelfSigned
2529 POSTSCRIPTUM: signTrusted all
2530 TYPE: sslproxy_cert_sign
2531 LOC: Config.ssl_client.cert_sign
2534 sslproxy_cert_sign <signing algorithm> acl ...
2536 The following certificate signing algorithms are supported:
2539 Sign using the configured CA certificate which is usually
2540 placed in and trusted by end-user browsers. This is the
2541 default for trusted origin server certificates.
2544 Sign to guarantee an X509_V_ERR_CERT_UNTRUSTED browser error.
2545 This is the default for untrusted origin server certificates
2546 that are not self-signed (see ssl::certUntrusted).
2549 Sign using a self-signed certificate with the right CN to
2550 generate a X509_V_ERR_DEPTH_ZERO_SELF_SIGNED_CERT error in the
2551 browser. This is the default for self-signed origin server
2552 certificates (see ssl::certSelfSigned).
2554 This clause only supports fast acl types.
2556 When sslproxy_cert_sign acl(s) match, Squid uses the corresponding
2557 signing algorithm to generate the certificate and ignores all
2558 subsequent sslproxy_cert_sign options (the first match wins). If no
2559 acl(s) match, the default signing algorithm is determined by errors
2560 detected when obtaining and validating the origin server certificate.
2562 WARNING: SQUID_X509_V_ERR_DOMAIN_MISMATCH and ssl:certDomainMismatch can
2563 be used with sslproxy_cert_adapt, but if and only if Squid is bumping a
2564 CONNECT request that carries a domain name. In all other cases (CONNECT
2565 to an IP address or an intercepted SSL connection), Squid cannot detect
2566 the domain mismatch at certificate generation time when
2567 bump-server-first is used.
2570 NAME: sslproxy_cert_adapt
2573 TYPE: sslproxy_cert_adapt
2574 LOC: Config.ssl_client.cert_adapt
2577 sslproxy_cert_adapt <adaptation algorithm> acl ...
2579 The following certificate adaptation algorithms are supported:
2582 Sets the "Not After" property to the "Not After" property of
2583 the CA certificate used to sign generated certificates.
2586 Sets the "Not Before" property to the "Not Before" property of
2587 the CA certificate used to sign generated certificates.
2589 setCommonName or setCommonName{CN}
2590 Sets Subject.CN property to the host name specified as a
2591 CN parameter or, if no explicit CN parameter was specified,
2592 extracted from the CONNECT request. It is a misconfiguration
2593 to use setCommonName without an explicit parameter for
2594 intercepted or tproxied SSL connections.
2596 This clause only supports fast acl types.
2598 Squid first groups sslproxy_cert_adapt options by adaptation algorithm.
2599 Within a group, when sslproxy_cert_adapt acl(s) match, Squid uses the
2600 corresponding adaptation algorithm to generate the certificate and
2601 ignores all subsequent sslproxy_cert_adapt options in that algorithm's
2602 group (i.e., the first match wins within each algorithm group). If no
2603 acl(s) match, the default mimicking action takes place.
2605 WARNING: SQUID_X509_V_ERR_DOMAIN_MISMATCH and ssl:certDomainMismatch can
2606 be used with sslproxy_cert_adapt, but if and only if Squid is bumping a
2607 CONNECT request that carries a domain name. In all other cases (CONNECT
2608 to an IP address or an intercepted SSL connection), Squid cannot detect
2609 the domain mismatch at certificate generation time when
2610 bump-server-first is used.
2613 NAME: sslpassword_program
2616 LOC: Config.Program.ssl_password
2619 Specify a program used for entering SSL key passphrases
2620 when using encrypted SSL certificate keys. If not specified
2621 keys must either be unencrypted, or Squid started with the -N
2622 option to allow it to query interactively for the passphrase.
2624 The key file name is given as argument to the program allowing
2625 selection of the right password if you have multiple encrypted
2630 OPTIONS RELATING TO EXTERNAL SSL_CRTD
2631 -----------------------------------------------------------------------------
2634 NAME: sslcrtd_program
2637 DEFAULT: @DEFAULT_SSL_CRTD@ -s @DEFAULT_SSL_DB_DIR@ -M 4MB
2638 LOC: Ssl::TheConfig.ssl_crtd
2640 Specify the location and options of the executable for ssl_crtd process.
2641 @DEFAULT_SSL_CRTD@ program requires -s and -M parameters
2642 For more information use:
2643 @DEFAULT_SSL_CRTD@ -h
2646 NAME: sslcrtd_children
2647 TYPE: HelperChildConfig
2649 DEFAULT: 32 startup=5 idle=1
2650 LOC: Ssl::TheConfig.ssl_crtdChildren
2652 The maximum number of processes spawn to service ssl server.
2653 The maximum this may be safely set to is 32.
2655 The startup= and idle= options allow some measure of skew in your
2660 Sets the minimum number of processes to spawn when Squid
2661 starts or reconfigures. When set to zero the first request will
2662 cause spawning of the first child process to handle it.
2664 Starting too few children temporary slows Squid under load while it
2665 tries to spawn enough additional processes to cope with traffic.
2669 Sets a minimum of how many processes Squid is to try and keep available
2670 at all times. When traffic begins to rise above what the existing
2671 processes can handle this many more will be spawned up to the maximum
2672 configured. A minimum setting of 1 is required.
2674 You must have at least one ssl_crtd process.
2677 NAME: sslcrtvalidator_program
2681 LOC: Ssl::TheConfig.ssl_crt_validator
2683 Specify the location and options of the executable for ssl_crt_validator
2686 Usage: sslcrtvalidator_program [ttl=n] [cache=n] path ...
2689 ttl=n TTL in seconds for cached results. The default is 60 secs
2690 cache=n limit the result cache size. The default value is 2048
2693 NAME: sslcrtvalidator_children
2694 TYPE: HelperChildConfig
2696 DEFAULT: 32 startup=5 idle=1 concurrency=1
2697 LOC: Ssl::TheConfig.ssl_crt_validator_Children
2699 The maximum number of processes spawn to service SSL server.
2700 The maximum this may be safely set to is 32.
2702 The startup= and idle= options allow some measure of skew in your
2707 Sets the minimum number of processes to spawn when Squid
2708 starts or reconfigures. When set to zero the first request will
2709 cause spawning of the first child process to handle it.
2711 Starting too few children temporary slows Squid under load while it
2712 tries to spawn enough additional processes to cope with traffic.
2716 Sets a minimum of how many processes Squid is to try and keep available
2717 at all times. When traffic begins to rise above what the existing
2718 processes can handle this many more will be spawned up to the maximum
2719 configured. A minimum setting of 1 is required.
2723 The number of requests each certificate validator helper can handle in
2724 parallel. A value of 0 indicates the certficate validator does not
2725 support concurrency. Defaults to 1.
2727 When this directive is set to a value >= 1 then the protocol
2728 used to communicate with the helper is modified to include
2729 a request ID in front of the request/response. The request
2730 ID from the request must be echoed back with the response
2733 You must have at least one ssl_crt_validator process.
2737 OPTIONS WHICH AFFECT THE NEIGHBOR SELECTION ALGORITHM
2738 -----------------------------------------------------------------------------
2746 To specify other caches in a hierarchy, use the format:
2748 cache_peer hostname type http-port icp-port [options]
2753 # hostname type port port options
2754 # -------------------- -------- ----- ----- -----------
2755 cache_peer parent.foo.net parent 3128 3130 default
2756 cache_peer sib1.foo.net sibling 3128 3130 proxy-only
2757 cache_peer sib2.foo.net sibling 3128 3130 proxy-only
2758 cache_peer example.com parent 80 0 default
2759 cache_peer cdn.example.com sibling 3128 0
2761 type: either 'parent', 'sibling', or 'multicast'.
2763 proxy-port: The port number where the peer accept HTTP requests.
2764 For other Squid proxies this is usually 3128
2765 For web servers this is usually 80
2767 icp-port: Used for querying neighbor caches about objects.
2768 Set to 0 if the peer does not support ICP or HTCP.
2769 See ICP and HTCP options below for additional details.
2772 ==== ICP OPTIONS ====
2774 You MUST also set icp_port and icp_access explicitly when using these options.
2775 The defaults will prevent peer traffic using ICP.
2778 no-query Disable ICP queries to this neighbor.
2781 Indicates the named peer is a member of a multicast group.
2782 ICP queries will not be sent directly to the peer, but ICP
2783 replies will be accepted from it.
2785 closest-only Indicates that, for ICP_OP_MISS replies, we'll only forward
2786 CLOSEST_PARENT_MISSes and never FIRST_PARENT_MISSes.
2789 To only send ICP queries to this neighbor infrequently.
2790 This is used to keep the neighbor round trip time updated
2791 and is usually used in conjunction with weighted-round-robin.
2794 ==== HTCP OPTIONS ====
2796 You MUST also set htcp_port and htcp_access explicitly when using these options.
2797 The defaults will prevent peer traffic using HTCP.
2800 htcp Send HTCP, instead of ICP, queries to the neighbor.
2801 You probably also want to set the "icp-port" to 4827
2802 instead of 3130. This directive accepts a comma separated
2803 list of options described below.
2805 htcp=oldsquid Send HTCP to old Squid versions (2.5 or earlier).
2807 htcp=no-clr Send HTCP to the neighbor but without
2808 sending any CLR requests. This cannot be used with
2811 htcp=only-clr Send HTCP to the neighbor but ONLY CLR requests.
2812 This cannot be used with no-clr.
2815 Send HTCP to the neighbor including CLRs but only when
2816 they do not result from PURGE requests.
2819 Forward any HTCP CLR requests this proxy receives to the peer.
2822 ==== PEER SELECTION METHODS ====
2824 The default peer selection method is ICP, with the first responding peer
2825 being used as source. These options can be used for better load balancing.
2828 default This is a parent cache which can be used as a "last-resort"
2829 if a peer cannot be located by any of the peer-selection methods.
2830 If specified more than once, only the first is used.
2832 round-robin Load-Balance parents which should be used in a round-robin
2833 fashion in the absence of any ICP queries.
2834 weight=N can be used to add bias.
2836 weighted-round-robin
2837 Load-Balance parents which should be used in a round-robin
2838 fashion with the frequency of each parent being based on the
2839 round trip time. Closer parents are used more often.
2840 Usually used for background-ping parents.
2841 weight=N can be used to add bias.
2843 carp Load-Balance parents which should be used as a CARP array.
2844 The requests will be distributed among the parents based on the
2845 CARP load balancing hash function based on their weight.
2847 userhash Load-balance parents based on the client proxy_auth or ident username.
2849 sourcehash Load-balance parents based on the client source IP.
2852 To be used only for cache peers of type "multicast".
2853 ALL members of this multicast group have "sibling"
2854 relationship with it, not "parent". This is to a multicast
2855 group when the requested object would be fetched only from
2856 a "parent" cache, anyway. It's useful, e.g., when
2857 configuring a pool of redundant Squid proxies, being
2858 members of the same multicast group.
2861 ==== PEER SELECTION OPTIONS ====
2863 weight=N use to affect the selection of a peer during any weighted
2864 peer-selection mechanisms.
2865 The weight must be an integer; default is 1,
2866 larger weights are favored more.
2867 This option does not affect parent selection if a peering
2868 protocol is not in use.
2870 basetime=N Specify a base amount to be subtracted from round trip
2872 It is subtracted before division by weight in calculating
2873 which parent to fectch from. If the rtt is less than the
2874 base time the rtt is set to a minimal value.
2876 ttl=N Specify a TTL to use when sending multicast ICP queries
2878 Only useful when sending to a multicast group.
2879 Because we don't accept ICP replies from random
2880 hosts, you must configure other group members as
2881 peers with the 'multicast-responder' option.
2883 no-delay To prevent access to this neighbor from influencing the
2886 digest-url=URL Tell Squid to fetch the cache digest (if digests are
2887 enabled) for this host from the specified URL rather
2888 than the Squid default location.
2891 ==== CARP OPTIONS ====
2893 carp-key=key-specification
2894 use a different key than the full URL to hash against the peer.
2895 the key-specification is a comma-separated list of the keywords
2896 scheme, host, port, path, params
2897 Order is not important.
2899 ==== ACCELERATOR / REVERSE-PROXY OPTIONS ====
2901 originserver Causes this parent to be contacted as an origin server.
2902 Meant to be used in accelerator setups when the peer
2906 Set the Host header of requests forwarded to this peer.
2907 Useful in accelerator setups where the server (peer)
2908 expects a certain domain name but clients may request
2909 others. ie example.com or www.example.com
2911 no-digest Disable request of cache digests.
2914 Disables requesting ICMP RTT database (NetDB).
2917 ==== AUTHENTICATION OPTIONS ====
2920 If this is a personal/workgroup proxy and your parent
2921 requires proxy authentication.
2923 Note: The string can include URL escapes (i.e. %20 for
2924 spaces). This also means % must be written as %%.
2927 Send login details received from client to this peer.
2928 Both Proxy- and WWW-Authorization headers are passed
2929 without alteration to the peer.
2930 Authentication is not required by Squid for this to work.
2932 Note: This will pass any form of authentication but
2933 only Basic auth will work through a proxy unless the
2934 connection-auth options are also used.
2936 login=PASS Send login details received from client to this peer.
2937 Authentication is not required by this option.
2939 If there are no client-provided authentication headers
2940 to pass on, but username and password are available
2941 from an external ACL user= and password= result tags
2942 they may be sent instead.
2944 Note: To combine this with proxy_auth both proxies must
2945 share the same user database as HTTP only allows for
2946 a single login (one for proxy, one for origin server).
2947 Also be warned this will expose your users proxy
2948 password to the peer. USE WITH CAUTION
2951 Send the username to the upstream cache, but with a
2952 fixed password. This is meant to be used when the peer
2953 is in another administrative domain, but it is still
2954 needed to identify each user.
2955 The star can optionally be followed by some extra
2956 information which is added to the username. This can
2957 be used to identify this proxy to the peer, similar to
2958 the login=username:password option above.
2961 If this is a personal/workgroup proxy and your parent
2962 requires a secure proxy authentication.
2963 The first principal from the default keytab or defined by
2964 the environment variable KRB5_KTNAME will be used.
2966 WARNING: The connection may transmit requests from multiple
2967 clients. Negotiate often assumes end-to-end authentication
2968 and a single-client. Which is not strictly true here.
2970 login=NEGOTIATE:principal_name
2971 If this is a personal/workgroup proxy and your parent
2972 requires a secure proxy authentication.
2973 The principal principal_name from the default keytab or
2974 defined by the environment variable KRB5_KTNAME will be
2977 WARNING: The connection may transmit requests from multiple
2978 clients. Negotiate often assumes end-to-end authentication
2979 and a single-client. Which is not strictly true here.
2981 connection-auth=on|off
2982 Tell Squid that this peer does or not support Microsoft
2983 connection oriented authentication, and any such
2984 challenges received from there should be ignored.
2985 Default is auto to automatically determine the status
2989 ==== SSL / HTTPS / TLS OPTIONS ====
2991 ssl Encrypt connections to this peer with SSL/TLS.
2993 sslcert=/path/to/ssl/certificate
2994 A client SSL certificate to use when connecting to
2997 sslkey=/path/to/ssl/key
2998 The private SSL key corresponding to sslcert above.
2999 If 'sslkey' is not specified 'sslcert' is assumed to
3000 reference a combined file containing both the
3001 certificate and the key.
3003 sslversion=1|2|3|4|5|6
3004 The SSL version to use when connecting to this peer
3005 1 = automatic (default)
3012 sslcipher=... The list of valid SSL ciphers to use when connecting
3015 ssloptions=... Specify various SSL implementation options:
3017 NO_SSLv2 Disallow the use of SSLv2
3018 NO_SSLv3 Disallow the use of SSLv3
3019 NO_TLSv1 Disallow the use of TLSv1.0
3020 NO_TLSv1_1 Disallow the use of TLSv1.1
3021 NO_TLSv1_2 Disallow the use of TLSv1.2
3023 Always create a new key when using
3024 temporary/ephemeral DH key exchanges
3025 ALL Enable various bug workarounds
3026 suggested as "harmless" by OpenSSL
3027 Be warned that this reduces SSL/TLS
3028 strength to some attacks.
3030 See the OpenSSL SSL_CTX_set_options documentation for a
3033 sslcafile=... A file containing additional CA certificates to use
3034 when verifying the peer certificate.
3036 sslcapath=... A directory containing additional CA certificates to
3037 use when verifying the peer certificate.
3039 sslcrlfile=... A certificate revocation list file to use when
3040 verifying the peer certificate.
3042 sslflags=... Specify various flags modifying the SSL implementation:
3045 Accept certificates even if they fail to
3048 Don't use the default CA list built in
3051 Don't verify the peer certificate
3052 matches the server name
3054 ssldomain= The peer name as advertised in it's certificate.
3055 Used for verifying the correctness of the received peer
3056 certificate. If not specified the peer hostname will be
3060 Enable the "Front-End-Https: On" header needed when
3061 using Squid as a SSL frontend in front of Microsoft OWA.
3062 See MS KB document Q307347 for details on this header.
3063 If set to auto the header will only be added if the
3064 request is forwarded as a https:// URL.
3067 ==== GENERAL OPTIONS ====
3070 A peer-specific connect timeout.
3071 Also see the peer_connect_timeout directive.
3073 connect-fail-limit=N
3074 How many times connecting to a peer must fail before
3075 it is marked as down. Standby connection failures
3076 count towards this limit. Default is 10.
3078 allow-miss Disable Squid's use of only-if-cached when forwarding
3079 requests to siblings. This is primarily useful when
3080 icp_hit_stale is used by the sibling. To extensive use
3081 of this option may result in forwarding loops, and you
3082 should avoid having two-way peerings with this option.
3083 For example to deny peer usage on requests from peer
3084 by denying cache_peer_access if the source is a peer.
3086 max-conn=N Limit the number of concurrent connections the Squid
3087 may open to this peer, including already opened idle
3088 and standby connections. There is no peer-specific
3089 connection limit by default.
3091 A peer exceeding the limit is not used for new
3092 requests unless a standby connection is available.
3094 max-conn currently works poorly with idle persistent
3095 connections: When a peer reaches its max-conn limit,
3096 and there are idle persistent connections to the peer,
3097 the peer may not be selected because the limiting code
3098 does not know whether Squid can reuse those idle
3101 standby=N Maintain a pool of N "hot standby" connections to an
3102 UP peer, available for requests when no idle
3103 persistent connection is available (or safe) to use.
3104 By default and with zero N, no such pool is maintained.
3105 N must not exceed the max-conn limit (if any).
3107 At start or after reconfiguration, Squid opens new TCP
3108 standby connections until there are N connections
3109 available and then replenishes the standby pool as
3110 opened connections are used up for requests. A used
3111 connection never goes back to the standby pool, but
3112 may go to the regular idle persistent connection pool
3113 shared by all peers and origin servers.
3115 Squid never opens multiple new standby connections
3116 concurrently. This one-at-a-time approach minimizes
3117 flooding-like effect on peers. Furthermore, just a few
3118 standby connections should be sufficient in most cases
3119 to supply most new requests with a ready-to-use
3122 Standby connections obey server_idle_pconn_timeout.
3123 For the feature to work as intended, the peer must be
3124 configured to accept and keep them open longer than
3125 the idle timeout at the connecting Squid, to minimize
3126 race conditions typical to idle used persistent
3127 connections. Default request_timeout and
3128 server_idle_pconn_timeout values ensure such a
3131 name=xxx Unique name for the peer.
3132 Required if you have multiple peers on the same host
3133 but different ports.
3134 This name can be used in cache_peer_access and similar
3135 directives to dentify the peer.
3136 Can be used by outgoing access controls through the
3139 no-tproxy Do not use the client-spoof TPROXY support when forwarding
3140 requests to this peer. Use normal address selection instead.
3141 This overrides the spoof_client_ip ACL.
3143 proxy-only objects fetched from the peer will not be stored locally.
3147 NAME: cache_peer_domain cache_host_domain
3152 Use to limit the domains for which a neighbor cache will be
3156 cache_peer_domain cache-host domain [domain ...]
3157 cache_peer_domain cache-host !domain
3159 For example, specifying
3161 cache_peer_domain parent.foo.net .edu
3163 has the effect such that UDP query packets are sent to
3164 'bigserver' only when the requested object exists on a
3165 server in the .edu domain. Prefixing the domainname
3166 with '!' means the cache will be queried for objects
3169 NOTE: * Any number of domains may be given for a cache-host,
3170 either on the same or separate lines.
3171 * When multiple domains are given for a particular
3172 cache-host, the first matched domain is applied.
3173 * Cache hosts with no domain restrictions are queried
3175 * There are no defaults.
3176 * There is also a 'cache_peer_access' tag in the ACL
3180 NAME: cache_peer_access
3185 Similar to 'cache_peer_domain' but provides more flexibility by
3189 cache_peer_access cache-host allow|deny [!]aclname ...
3191 The syntax is identical to 'http_access' and the other lists of
3192 ACL elements. See the comments for 'http_access' below, or
3193 the Squid FAQ (http://wiki.squid-cache.org/SquidFaq/SquidAcl).
3196 NAME: neighbor_type_domain
3197 TYPE: hostdomaintype
3199 DEFAULT_DOC: The peer type from cache_peer directive is used for all requests to that peer.
3202 Modify the cache_peer neighbor type when passing requests
3203 about specific domains to the peer.
3206 neighbor_type_domain neighbor parent|sibling domain domain ...
3209 cache_peer foo.example.com parent 3128 3130
3210 neighbor_type_domain foo.example.com sibling .au .de
3212 The above configuration treats all requests to foo.example.com as a
3213 parent proxy unless the request is for a .au or .de ccTLD domain name.
3216 NAME: dead_peer_timeout
3220 LOC: Config.Timeout.deadPeer
3222 This controls how long Squid waits to declare a peer cache
3223 as "dead." If there are no ICP replies received in this
3224 amount of time, Squid will declare the peer dead and not
3225 expect to receive any further ICP replies. However, it
3226 continues to send ICP queries, and will mark the peer as
3227 alive upon receipt of the first subsequent ICP reply.
3229 This timeout also affects when Squid expects to receive ICP
3230 replies from peers. If more than 'dead_peer' seconds have
3231 passed since the last ICP reply was received, Squid will not
3232 expect to receive an ICP reply on the next query. Thus, if
3233 your time between requests is greater than this timeout, you
3234 will see a lot of requests sent DIRECT to origin servers
3235 instead of to your parents.
3238 NAME: forward_max_tries
3241 LOC: Config.forward_max_tries
3243 Controls how many different forward paths Squid will try
3244 before giving up. See also forward_timeout.
3246 NOTE: connect_retries (default: none) can make each of these
3247 possible forwarding paths be tried multiple times.
3251 MEMORY CACHE OPTIONS
3252 -----------------------------------------------------------------------------
3259 LOC: Config.memMaxSize
3261 NOTE: THIS PARAMETER DOES NOT SPECIFY THE MAXIMUM PROCESS SIZE.
3262 IT ONLY PLACES A LIMIT ON HOW MUCH ADDITIONAL MEMORY SQUID WILL
3263 USE AS A MEMORY CACHE OF OBJECTS. SQUID USES MEMORY FOR OTHER
3264 THINGS AS WELL. SEE THE SQUID FAQ SECTION 8 FOR DETAILS.
3266 'cache_mem' specifies the ideal amount of memory to be used
3268 * In-Transit objects
3270 * Negative-Cached objects
3272 Data for these objects are stored in 4 KB blocks. This
3273 parameter specifies the ideal upper limit on the total size of
3274 4 KB blocks allocated. In-Transit objects take the highest
3277 In-transit objects have priority over the others. When
3278 additional space is needed for incoming data, negative-cached
3279 and hot objects will be released. In other words, the
3280 negative-cached and hot objects will fill up any unused space
3281 not needed for in-transit objects.
3283 If circumstances require, this limit will be exceeded.
3284 Specifically, if your incoming request rate requires more than
3285 'cache_mem' of memory to hold in-transit objects, Squid will
3286 exceed this limit to satisfy the new requests. When the load
3287 decreases, blocks will be freed until the high-water mark is
3288 reached. Thereafter, blocks will be used to store hot
3291 If shared memory caching is enabled, Squid does not use the shared
3292 cache space for in-transit objects, but they still consume as much
3293 local memory as they need. For more details about the shared memory
3294 cache, see memory_cache_shared.
3297 NAME: maximum_object_size_in_memory
3301 LOC: Config.Store.maxInMemObjSize
3303 Objects greater than this size will not be attempted to kept in
3304 the memory cache. This should be set high enough to keep objects
3305 accessed frequently in memory to improve performance whilst low
3306 enough to keep larger objects from hoarding cache_mem.
3309 NAME: memory_cache_shared
3312 LOC: Config.memShared
3314 DEFAULT_DOC: "on" where supported if doing memory caching with multiple SMP workers.
3316 Controls whether the memory cache is shared among SMP workers.
3318 The shared memory cache is meant to occupy cache_mem bytes and replace
3319 the non-shared memory cache, although some entities may still be
3320 cached locally by workers for now (e.g., internal and in-transit
3321 objects may be served from a local memory cache even if shared memory
3322 caching is enabled).
3324 By default, the memory cache is shared if and only if all of the
3325 following conditions are satisfied: Squid runs in SMP mode with
3326 multiple workers, cache_mem is positive, and Squid environment
3327 supports required IPC primitives (e.g., POSIX shared memory segments
3328 and GCC-style atomic operations).
3330 To avoid blocking locks, shared memory uses opportunistic algorithms
3331 that do not guarantee that every cachable entity that could have been
3332 shared among SMP workers will actually be shared.
3334 Currently, entities exceeding 32KB in size cannot be shared.
3337 NAME: memory_cache_mode
3341 DEFAULT_DOC: Keep the most recently fetched objects in memory
3343 Controls which objects to keep in the memory cache (cache_mem)
3345 always Keep most recently fetched objects in memory (default)
3347 disk Only disk cache hits are kept in memory, which means
3348 an object must first be cached on disk and then hit
3349 a second time before cached in memory.
3351 network Only objects fetched from network is kept in memory
3354 NAME: memory_replacement_policy
3356 LOC: Config.memPolicy
3359 The memory replacement policy parameter determines which
3360 objects are purged from memory when memory space is needed.
3362 See cache_replacement_policy for details on algorithms.
3367 -----------------------------------------------------------------------------
3370 NAME: cache_replacement_policy
3372 LOC: Config.replPolicy
3375 The cache replacement policy parameter determines which
3376 objects are evicted (replaced) when disk space is needed.
3378 lru : Squid's original list based LRU policy
3379 heap GDSF : Greedy-Dual Size Frequency
3380 heap LFUDA: Least Frequently Used with Dynamic Aging
3381 heap LRU : LRU policy implemented using a heap
3383 Applies to any cache_dir lines listed below this directive.
3385 The LRU policies keeps recently referenced objects.
3387 The heap GDSF policy optimizes object hit rate by keeping smaller
3388 popular objects in cache so it has a better chance of getting a
3389 hit. It achieves a lower byte hit rate than LFUDA though since
3390 it evicts larger (possibly popular) objects.
3392 The heap LFUDA policy keeps popular objects in cache regardless of
3393 their size and thus optimizes byte hit rate at the expense of
3394 hit rate since one large, popular object will prevent many
3395 smaller, slightly less popular objects from being cached.
3397 Both policies utilize a dynamic aging mechanism that prevents
3398 cache pollution that can otherwise occur with frequency-based
3399 replacement policies.
3401 NOTE: if using the LFUDA replacement policy you should increase
3402 the value of maximum_object_size above its default of 4 MB to
3403 to maximize the potential byte hit rate improvement of LFUDA.
3405 For more information about the GDSF and LFUDA cache replacement
3406 policies see http://www.hpl.hp.com/techreports/1999/HPL-1999-69.html
3407 and http://fog.hpl.external.hp.com/techreports/98/HPL-98-173.html.
3410 NAME: minimum_object_size
3414 DEFAULT_DOC: no limit
3415 LOC: Config.Store.minObjectSize
3417 Objects smaller than this size will NOT be saved on disk. The
3418 value is specified in bytes, and the default is 0 KB, which
3419 means all responses can be stored.
3422 NAME: maximum_object_size
3426 LOC: Config.Store.maxObjectSize
3428 Set the default value for max-size parameter on any cache_dir.
3429 The value is specified in bytes, and the default is 4 MB.
3431 If you wish to get a high BYTES hit ratio, you should probably
3432 increase this (one 32 MB object hit counts for 3200 10KB
3435 If you wish to increase hit ratio more than you want to
3436 save bandwidth you should leave this low.
3438 NOTE: if using the LFUDA replacement policy you should increase
3439 this value to maximize the byte hit rate improvement of LFUDA!
3440 See cache_replacement_policy for a discussion of this policy.
3446 DEFAULT_DOC: No disk cache. Store cache ojects only in memory.
3447 LOC: Config.cacheSwap
3450 cache_dir Type Directory-Name Fs-specific-data [options]
3452 You can specify multiple cache_dir lines to spread the
3453 cache among different disk partitions.
3455 Type specifies the kind of storage system to use. Only "ufs"
3456 is built by default. To enable any of the other storage systems
3457 see the --enable-storeio configure option.
3459 'Directory' is a top-level directory where cache swap
3460 files will be stored. If you want to use an entire disk
3461 for caching, this can be the mount-point directory.
3462 The directory must exist and be writable by the Squid
3463 process. Squid will NOT create this directory for you.
3465 In SMP configurations, cache_dir must not precede the workers option
3466 and should use configuration macros or conditionals to give each
3467 worker interested in disk caching a dedicated cache directory.
3470 ==== The ufs store type ====
3472 "ufs" is the old well-known Squid storage format that has always
3476 cache_dir ufs Directory-Name Mbytes L1 L2 [options]
3478 'Mbytes' is the amount of disk space (MB) to use under this
3479 directory. The default is 100 MB. Change this to suit your
3480 configuration. Do NOT put the size of your disk drive here.
3481 Instead, if you want Squid to use the entire disk drive,
3482 subtract 20% and use that value.
3484 'L1' is the number of first-level subdirectories which
3485 will be created under the 'Directory'. The default is 16.
3487 'L2' is the number of second-level subdirectories which
3488 will be created under each first-level directory. The default
3492 ==== The aufs store type ====
3494 "aufs" uses the same storage format as "ufs", utilizing
3495 POSIX-threads to avoid blocking the main Squid process on
3496 disk-I/O. This was formerly known in Squid as async-io.
3499 cache_dir aufs Directory-Name Mbytes L1 L2 [options]
3501 see argument descriptions under ufs above
3504 ==== The diskd store type ====
3506 "diskd" uses the same storage format as "ufs", utilizing a
3507 separate process to avoid blocking the main Squid process on
3511 cache_dir diskd Directory-Name Mbytes L1 L2 [options] [Q1=n] [Q2=n]
3513 see argument descriptions under ufs above
3515 Q1 specifies the number of unacknowledged I/O requests when Squid
3516 stops opening new files. If this many messages are in the queues,
3517 Squid won't open new files. Default is 64
3519 Q2 specifies the number of unacknowledged messages when Squid
3520 starts blocking. If this many messages are in the queues,
3521 Squid blocks until it receives some replies. Default is 72
3523 When Q1 < Q2 (the default), the cache directory is optimized
3524 for lower response time at the expense of a decrease in hit
3525 ratio. If Q1 > Q2, the cache directory is optimized for
3526 higher hit ratio at the expense of an increase in response
3530 ==== The rock store type ====
3533 cache_dir rock Directory-Name Mbytes [options]
3535 The Rock Store type is a database-style storage. All cached
3536 entries are stored in a "database" file, using fixed-size slots.
3537 A single entry occupies one or more slots.
3539 If possible, Squid using Rock Store creates a dedicated kid
3540 process called "disker" to avoid blocking Squid worker(s) on disk
3541 I/O. One disker kid is created for each rock cache_dir. Diskers
3542 are created only when Squid, running in daemon mode, has support
3543 for the IpcIo disk I/O module.
3545 swap-timeout=msec: Squid will not start writing a miss to or
3546 reading a hit from disk if it estimates that the swap operation
3547 will take more than the specified number of milliseconds. By
3548 default and when set to zero, disables the disk I/O time limit
3549 enforcement. Ignored when using blocking I/O module because
3550 blocking synchronous I/O does not allow Squid to estimate the
3551 expected swap wait time.
3553 max-swap-rate=swaps/sec: Artificially limits disk access using
3554 the specified I/O rate limit. Swap out requests that
3555 would cause the average I/O rate to exceed the limit are
3556 delayed. Individual swap in requests (i.e., hits or reads) are
3557 not delayed, but they do contribute to measured swap rate and
3558 since they are placed in the same FIFO queue as swap out
3559 requests, they may wait longer if max-swap-rate is smaller.
3560 This is necessary on file systems that buffer "too
3561 many" writes and then start blocking Squid and other processes
3562 while committing those writes to disk. Usually used together
3563 with swap-timeout to avoid excessive delays and queue overflows
3564 when disk demand exceeds available disk "bandwidth". By default
3565 and when set to zero, disables the disk I/O rate limit
3566 enforcement. Currently supported by IpcIo module only.
3568 slot-size=bytes: The size of a database "record" used for
3569 storing cached responses. A cached response occupies at least
3570 one slot and all database I/O is done using individual slots so
3571 increasing this parameter leads to more disk space waste while
3572 decreasing it leads to more disk I/O overheads. Should be a
3573 multiple of your operating system I/O page size. Defaults to
3574 16KBytes. A housekeeping header is stored with each slot and
3575 smaller slot-sizes will be rejected. The header is smaller than
3579 ==== COMMON OPTIONS ====
3581 no-store no new objects should be stored to this cache_dir.
3583 min-size=n the minimum object size in bytes this cache_dir
3584 will accept. It's used to restrict a cache_dir
3585 to only store large objects (e.g. AUFS) while
3586 other stores are optimized for smaller objects
3590 max-size=n the maximum object size in bytes this cache_dir
3592 The value in maximum_object_size directive sets
3593 the default unless more specific details are
3594 available (ie a small store capacity).
3596 Note: To make optimal use of the max-size limits you should order
3597 the cache_dir lines with the smallest max-size value first.
3601 # Uncomment and adjust the following to add a disk cache directory.
3602 #cache_dir ufs @DEFAULT_SWAP_DIR@ 100 16 256
3606 NAME: store_dir_select_algorithm
3608 LOC: Config.store_dir_select_algorithm
3611 How Squid selects which cache_dir to use when the response
3612 object will fit into more than one.
3614 Regardless of which algorithm is used the cache_dir min-size
3615 and max-size parameters are obeyed. As such they can affect
3616 the selection algorithm by limiting the set of considered
3623 This algorithm is suited to caches with similar cache_dir
3624 sizes and disk speeds.
3626 The disk with the least I/O pending is selected.
3627 When there are multiple disks with the same I/O load ranking
3628 the cache_dir with most available capacity is selected.
3630 When a mix of cache_dir sizes are configured the faster disks
3631 have a naturally lower I/O loading and larger disks have more
3632 capacity. So space used to store objects and data throughput
3633 may be very unbalanced towards larger disks.
3638 This algorithm is suited to caches with unequal cache_dir
3641 Each cache_dir is selected in a rotation. The next suitable
3644 Available cache_dir capacity is only considered in relation
3645 to whether the object will fit and meets the min-size and
3646 max-size parameters.
3648 Disk I/O loading is only considered to prevent overload on slow
3649 disks. This algorithm does not spread objects by size, so any
3650 I/O loading per-disk may appear very unbalanced and volatile.
3652 If several cache_dirs use similar min-size, max-size, or other
3653 limits to to reject certain responses, then do not group such
3654 cache_dir lines together, to avoid round-robin selection bias
3655 towards the first cache_dir after the group. Instead, interleave
3656 cache_dir lines from different groups. For example:
3658 store_dir_select_algorithm round-robin
3659 cache_dir rock /hdd1 ... min-size=100000
3660 cache_dir rock /ssd1 ... max-size=99999
3661 cache_dir rock /hdd2 ... min-size=100000
3662 cache_dir rock /ssd2 ... max-size=99999
3663 cache_dir rock /hdd3 ... min-size=100000
3664 cache_dir rock /ssd3 ... max-size=99999
3667 NAME: max_open_disk_fds
3669 LOC: Config.max_open_disk_fds
3671 DEFAULT_DOC: no limit
3673 To avoid having disk as the I/O bottleneck Squid can optionally
3674 bypass the on-disk cache if more than this amount of disk file
3675 descriptors are open.
3677 A value of 0 indicates no limit.
3680 NAME: cache_swap_low
3681 COMMENT: (percent, 0-100)
3684 LOC: Config.Swap.lowWaterMark
3686 The low-water mark for cache object replacement.
3687 Replacement begins when the swap (disk) usage is above the
3688 low-water mark and attempts to maintain utilization near the
3689 low-water mark. As swap utilization gets close to high-water
3690 mark object eviction becomes more aggressive. If utilization is
3691 close to the low-water mark less replacement is done each time.
3693 Defaults are 90% and 95%. If you have a large cache, 5% could be
3694 hundreds of MB. If this is the case you may wish to set these
3695 numbers closer together.
3697 See also cache_swap_high
3700 NAME: cache_swap_high
3701 COMMENT: (percent, 0-100)
3704 LOC: Config.Swap.highWaterMark
3706 The high-water mark for cache object replacement.
3707 Replacement begins when the swap (disk) usage is above the
3708 low-water mark and attempts to maintain utilization near the
3709 low-water mark. As swap utilization gets close to high-water
3710 mark object eviction becomes more aggressive. If utilization is
3711 close to the low-water mark less replacement is done each time.
3713 Defaults are 90% and 95%. If you have a large cache, 5% could be
3714 hundreds of MB. If this is the case you may wish to set these
3715 numbers closer together.
3717 See also cache_swap_low
3722 -----------------------------------------------------------------------------
3729 DEFAULT_DOC: The format definitions squid, common, combined, referrer, useragent are built in.
3733 logformat <name> <format specification>
3735 Defines an access log format.
3737 The <format specification> is a string with embedded % format codes
3739 % format codes all follow the same basic structure where all but
3740 the formatcode is optional. Output strings are automatically escaped
3741 as required according to their context and the output format
3742 modifiers are usually not needed, but can be specified if an explicit
3743 output format is desired.
3745 % ["|[|'|#] [-] [[0]width] [{argument}] formatcode
3747 " output in quoted string format
3748 [ output in squid text log format as used by log_mime_hdrs
3749 # output in URL quoted format
3754 width minimum and/or maximum field width:
3755 [width_min][.width_max]
3756 When minimum starts with 0, the field is zero-padded.
3757 String values exceeding maximum width are truncated.
3759 {arg} argument such as header name etc
3763 % a literal % character
3764 sn Unique sequence number per log line entry
3765 err_code The ID of an error response served by Squid or
3766 a similar internal error identifier.
3767 err_detail Additional err_code-dependent error information.
3768 note The annotation specified by the argument. Also
3769 logs the adaptation meta headers set by the
3770 adaptation_meta configuration parameter.
3771 If no argument given all annotations logged.
3772 The argument may include a separator to use with
3775 By default, multiple note values are separated with ","
3776 and multiple notes are separated with "\r\n".
3777 When logging named notes with %{name}note, the
3778 explicitly configured separator is used between note
3779 values. When logging all notes with %note, the
3780 explicitly configured separator is used between
3781 individual notes. There is currently no way to
3782 specify both value and notes separators when logging
3783 all notes with %note.
3785 Connection related format codes:
3787 >a Client source IP address
3789 >p Client source port
3790 >eui Client source EUI (MAC address, EUI-48 or EUI-64 identifier)
3791 >la Local IP address the client connected to
3792 >lp Local port number the client connected to
3793 >qos Client connection TOS/DSCP value set by Squid
3794 >nfmark Client connection netfilter mark set by Squid
3796 la Local listening IP address the client connection was connected to.
3797 lp Local listening port number the client connection was connected to.
3799 <a Server IP address of the last server or peer connection
3800 <A Server FQDN or peer name
3801 <p Server port number of the last server or peer connection
3802 <la Local IP address of the last server or peer connection
3803 <lp Local port number of the last server or peer connection
3804 <qos Server connection TOS/DSCP value set by Squid
3805 <nfmark Server connection netfilter mark set by Squid
3807 Time related format codes:
3809 ts Seconds since epoch
3810 tu subsecond time (milliseconds)
3811 tl Local time. Optional strftime format argument
3812 default %d/%b/%Y:%H:%M:%S %z
3813 tg GMT time. Optional strftime format argument
3814 default %d/%b/%Y:%H:%M:%S %z
3815 tr Response time (milliseconds)
3816 dt Total time spent making DNS lookups (milliseconds)
3817 tS Approximate master transaction start time in
3818 <full seconds since epoch>.<fractional seconds> format.
3819 Currently, Squid considers the master transaction
3820 started when a complete HTTP request header initiating
3821 the transaction is received from the client. This is
3822 the same value that Squid uses to calculate transaction
3823 response time when logging %tr to access.log. Currently,
3824 Squid uses millisecond resolution for %tS values,
3825 similar to the default access.log "current time" field
3828 Access Control related format codes:
3830 et Tag returned by external acl
3831 ea Log string returned by external acl
3832 un User name (any available)
3833 ul User name from authentication
3834 ue User name from external acl helper
3835 ui User name from ident
3836 us User name from SSL
3837 credentials Client credentials. The exact meaning depends on
3838 the authentication scheme: For Basic authentication,
3839 it is the password; for Digest, the realm sent by the
3840 client; for NTLM and Negotiate, the client challenge
3841 or client credentials prefixed with "YR " or "KK ".
3843 HTTP related format codes:
3847 [http::]rm Request method (GET/POST etc)
3848 [http::]>rm Request method from client
3849 [http::]<rm Request method sent to server or peer
3850 [http::]ru Request URL from client (historic, filtered for logging)
3851 [http::]>ru Request URL from client
3852 [http::]<ru Request URL sent to server or peer
3853 [http::]>rs Request URL scheme from client
3854 [http::]<rs Request URL scheme sent to server or peer
3855 [http::]>rd Request URL domain from client
3856 [http::]>rd Request URL domain sent to server or peer
3857 [http::]>rP Request URL port from client
3858 [http::]<rP Request URL port sent to server or peer
3859 [http::]rp Request URL path excluding hostname
3860 [http::]>rp Request URL path excluding hostname from client
3861 [http::]<rp Request URL path excluding hostname sent to server or peer
3862 [http::]rv Request protocol version
3863 [http::]>rv Request protocol version from client
3864 [http::]<rv Request protocol version sent to server or peer
3866 [http::]>h Original received request header.
3867 Usually differs from the request header sent by
3868 Squid, although most fields are often preserved.
3869 Accepts optional header field name/value filter
3870 argument using name[:[separator]element] format.
3871 [http::]>ha Received request header after adaptation and
3872 redirection (pre-cache REQMOD vectoring point).
3873 Usually differs from the request header sent by
3874 Squid, although most fields are often preserved.
3875 Optional header name argument as for >h
3880 [http::]<Hs HTTP status code received from the next hop
3881 [http::]>Hs HTTP status code sent to the client
3883 [http::]<h Reply header. Optional header name argument
3886 [http::]mt MIME content type
3891 [http::]st Total size of request + reply traffic with client
3892 [http::]>st Total size of request received from client.
3893 Excluding chunked encoding bytes.
3894 [http::]<st Total size of reply sent to client (after adaptation)
3896 [http::]>sh Size of request headers received from client
3897 [http::]<sh Size of reply headers sent to client (after adaptation)
3899 [http::]<sH Reply high offset sent
3900 [http::]<sS Upstream object size
3902 [http::]<bs Number of HTTP-equivalent message body bytes
3903 received from the next hop, excluding chunked
3904 transfer encoding and control messages.
3905 Generated FTP/Gopher listings are treated as
3911 [http::]<pt Peer response time in milliseconds. The timer starts
3912 when the last request byte is sent to the next hop
3913 and stops when the last response byte is received.
3914 [http::]<tt Total server-side time in milliseconds. The timer
3915 starts with the first connect request (or write I/O)
3916 sent to the first selected peer. The timer stops
3917 with the last I/O with the last peer.
3919 Squid handling related format codes:
3921 Ss Squid request status (TCP_MISS etc)
3922 Sh Squid hierarchy status (DEFAULT_PARENT etc)
3924 SSL-related format codes:
3926 ssl::bump_mode SslBump decision for the transaction:
3928 For CONNECT requests that initiated bumping of
3929 a connection and for any request received on
3930 an already bumped connection, Squid logs the
3931 corresponding SslBump mode ("server-first" or
3932 "client-first"). See the ssl_bump option for
3933 more information about these modes.
3935 A "none" token is logged for requests that
3936 triggered "ssl_bump" ACL evaluation matching
3937 either a "none" rule or no rules at all.
3939 In all other cases, a single dash ("-") is
3942 ssl::>sni SSL client SNI sent to Squid. Available only
3943 after the peek, stare, or splice SSL bumping
3946 If ICAP is enabled, the following code becomes available (as
3947 well as ICAP log codes documented with the icap_log option):
3949 icap::tt Total ICAP processing time for the HTTP
3950 transaction. The timer ticks when ICAP
3951 ACLs are checked and when ICAP
3952 transaction is in progress.
3954 If adaptation is enabled the following three codes become available:
3956 adapt::<last_h The header of the last ICAP response or
3957 meta-information from the last eCAP
3958 transaction related to the HTTP transaction.
3959 Like <h, accepts an optional header name
3962 adapt::sum_trs Summed adaptation transaction response
3963 times recorded as a comma-separated list in
3964 the order of transaction start time. Each time
3965 value is recorded as an integer number,
3966 representing response time of one or more
3967 adaptation (ICAP or eCAP) transaction in
3968 milliseconds. When a failed transaction is
3969 being retried or repeated, its time is not
3970 logged individually but added to the
3971 replacement (next) transaction. See also:
3974 adapt::all_trs All adaptation transaction response times.
3975 Same as adaptation_strs but response times of
3976 individual transactions are never added
3977 together. Instead, all transaction response
3978 times are recorded individually.
3980 You can prefix adapt::*_trs format codes with adaptation
3981 service name in curly braces to record response time(s) specific
3982 to that service. For example: %{my_service}adapt::sum_trs
3984 If SSL is enabled, the following formating codes become available:
3986 %ssl::>cert_subject The Subject field of the received client
3987 SSL certificate or a dash ('-') if Squid has
3988 received an invalid/malformed certificate or
3989 no certificate at all. Consider encoding the
3990 logged value because Subject often has spaces.
3992 %ssl::>cert_issuer The Issuer field of the received client
3993 SSL certificate or a dash ('-') if Squid has
3994 received an invalid/malformed certificate or
3995 no certificate at all. Consider encoding the
3996 logged value because Issuer often has spaces.
3998 The default formats available (which do not need re-defining) are:
4000 logformat squid %ts.%03tu %6tr %>a %Ss/%03>Hs %<st %rm %ru %[un %Sh/%<a %mt
4001 logformat common %>a %[ui %[un [%tl] "%rm %ru HTTP/%rv" %>Hs %<st %Ss:%Sh
4002 logformat combined %>a %[ui %[un [%tl] "%rm %ru HTTP/%rv" %>Hs %<st "%{Referer}>h" "%{User-Agent}>h" %Ss:%Sh
4003 logformat referrer %ts.%03tu %>a %{Referer}>h %ru
4004 logformat useragent %>a [%tl] "%{User-Agent}>h"
4006 NOTE: When the log_mime_hdrs directive is set to ON.
4007 The squid, common and combined formats have a safely encoded copy
4008 of the mime headers appended to each line within a pair of brackets.
4010 NOTE: The common and combined formats are not quite true to the Apache definition.
4011 The logs from Squid contain an extra status and hierarchy code appended.
4015 NAME: access_log cache_access_log
4017 LOC: Config.Log.accesslogs
4018 DEFAULT_IF_NONE: daemon:@DEFAULT_ACCESS_LOG@ squid
4020 Configures whether and how Squid logs HTTP and ICP transactions.
4021 If access logging is enabled, a single line is logged for every
4022 matching HTTP or ICP request. The recommended directive formats are:
4024 access_log <module>:<place> [option ...] [acl acl ...]
4025 access_log none [acl acl ...]
4027 The following directive format is accepted but may be deprecated:
4028 access_log <module>:<place> [<logformat name> [acl acl ...]]
4030 In most cases, the first ACL name must not contain the '=' character
4031 and should not be equal to an existing logformat name. You can always
4032 start with an 'all' ACL to work around those restrictions.
4034 Will log to the specified module:place using the specified format (which
4035 must be defined in a logformat directive) those entries which match
4036 ALL the acl's specified (which must be defined in acl clauses).
4037 If no acl is specified, all requests will be logged to this destination.
4039 ===== Available options for the recommended directive format =====
4041 logformat=name Names log line format (either built-in or
4042 defined by a logformat directive). Defaults
4045 buffer-size=64KB Defines approximate buffering limit for log
4046 records (see buffered_logs). Squid should not
4047 keep more than the specified size and, hence,
4048 should flush records before the buffer becomes
4049 full to avoid overflows under normal
4050 conditions (the exact flushing algorithm is
4051 module-dependent though). The on-error option
4052 controls overflow handling.
4054 on-error=die|drop Defines action on unrecoverable errors. The
4055 'drop' action ignores (i.e., does not log)
4056 affected log records. The default 'die' action
4057 kills the affected worker. The drop action
4058 support has not been tested for modules other
4061 ===== Modules Currently available =====
4063 none Do not log any requests matching these ACL.
4064 Do not specify Place or logformat name.
4066 stdio Write each log line to disk immediately at the completion of
4068 Place: the filename and path to be written.
4070 daemon Very similar to stdio. But instead of writing to disk the log
4071 line is passed to a daemon helper for asychronous handling instead.
4072 Place: varies depending on the daemon.
4074 log_file_daemon Place: the file name and path to be written.
4076 syslog To log each request via syslog facility.
4077 Place: The syslog facility and priority level for these entries.
4078 Place Format: facility.priority
4080 where facility could be any of:
4081 authpriv, daemon, local0 ... local7 or user.
4083 And priority could be any of:
4084 err, warning, notice, info, debug.
4086 udp To send each log line as text data to a UDP receiver.
4087 Place: The destination host name or IP and port.
4088 Place Format: //host:port
4090 tcp To send each log line as text data to a TCP receiver.
4091 Lines may be accumulated before sending (see buffered_logs).
4092 Place: The destination host name or IP and port.
4093 Place Format: //host:port
4096 access_log daemon:@DEFAULT_ACCESS_LOG@ squid
4102 LOC: Config.Log.icaplogs
4105 ICAP log files record ICAP transaction summaries, one line per
4108 The icap_log option format is:
4109 icap_log <filepath> [<logformat name> [acl acl ...]]
4110 icap_log none [acl acl ...]]
4112 Please see access_log option documentation for details. The two
4113 kinds of logs share the overall configuration approach and many
4116 ICAP processing of a single HTTP message or transaction may
4117 require multiple ICAP transactions. In such cases, multiple
4118 ICAP transaction log lines will correspond to a single access
4121 ICAP log uses logformat codes that make sense for an ICAP
4122 transaction. Header-related codes are applied to the HTTP header
4123 embedded in an ICAP server response, with the following caveats:
4124 For REQMOD, there is no HTTP response header unless the ICAP
4125 server performed request satisfaction. For RESPMOD, the HTTP
4126 request header is the header sent to the ICAP server. For
4127 OPTIONS, there are no HTTP headers.
4129 The following format codes are also available for ICAP logs:
4131 icap::<A ICAP server IP address. Similar to <A.
4133 icap::<service_name ICAP service name from the icap_service
4134 option in Squid configuration file.
4136 icap::ru ICAP Request-URI. Similar to ru.
4138 icap::rm ICAP request method (REQMOD, RESPMOD, or
4139 OPTIONS). Similar to existing rm.
4141 icap::>st Bytes sent to the ICAP server (TCP payload
4142 only; i.e., what Squid writes to the socket).
4144 icap::<st Bytes received from the ICAP server (TCP
4145 payload only; i.e., what Squid reads from
4148 icap::<bs Number of message body bytes received from the
4149 ICAP server. ICAP message body, if any, usually
4150 includes encapsulated HTTP message headers and
4151 possibly encapsulated HTTP message body. The
4152 HTTP body part is dechunked before its size is
4155 icap::tr Transaction response time (in
4156 milliseconds). The timer starts when
4157 the ICAP transaction is created and
4158 stops when the transaction is completed.
4161 icap::tio Transaction I/O time (in milliseconds). The
4162 timer starts when the first ICAP request
4163 byte is scheduled for sending. The timers
4164 stops when the last byte of the ICAP response
4167 icap::to Transaction outcome: ICAP_ERR* for all
4168 transaction errors, ICAP_OPT for OPTION
4169 transactions, ICAP_ECHO for 204
4170 responses, ICAP_MOD for message
4171 modification, and ICAP_SAT for request
4172 satisfaction. Similar to Ss.
4174 icap::Hs ICAP response status code. Similar to Hs.
4176 icap::>h ICAP request header(s). Similar to >h.
4178 icap::<h ICAP response header(s). Similar to <h.
4180 The default ICAP log format, which can be used without an explicit
4181 definition, is called icap_squid:
4183 logformat icap_squid %ts.%03tu %6icap::tr %>a %icap::to/%03icap::Hs %icap::<size %icap::rm %icap::ru% %un -/%icap::<A -
4185 See also: logformat, log_icap, and %adapt::<last_h
4188 NAME: logfile_daemon
4190 DEFAULT: @DEFAULT_LOGFILED@
4191 LOC: Log::TheConfig.logfile_daemon
4193 Specify the path to the logfile-writing daemon. This daemon is
4194 used to write the access and store logs, if configured.
4196 Squid sends a number of commands to the log daemon:
4197 L<data>\n - logfile data
4202 r<n>\n - set rotate count to <n>
4203 b<n>\n - 1 = buffer output, 0 = don't buffer output
4205 No responses is expected.
4208 NAME: stats_collection
4210 LOC: Config.accessList.stats_collection
4212 DEFAULT_DOC: Allow logging for all transactions.
4213 COMMENT: allow|deny acl acl...
4215 This options allows you to control which requests gets accounted
4216 in performance counters.
4218 This clause only supports fast acl types.
4219 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
4222 NAME: cache_store_log
4225 LOC: Config.Log.store
4227 Logs the activities of the storage manager. Shows which
4228 objects are ejected from the cache, and which objects are
4229 saved and for how long.
4230 There are not really utilities to analyze this data, so you can safely
4231 disable it (the default).
4233 Store log uses modular logging outputs. See access_log for the list
4234 of modules supported.
4237 cache_store_log stdio:@DEFAULT_STORE_LOG@
4238 cache_store_log daemon:@DEFAULT_STORE_LOG@
4241 NAME: cache_swap_state cache_swap_log
4243 LOC: Config.Log.swap
4245 DEFAULT_DOC: Store the journal inside its cache_dir
4247 Location for the cache "swap.state" file. This index file holds
4248 the metadata of objects saved on disk. It is used to rebuild
4249 the cache during startup. Normally this file resides in each
4250 'cache_dir' directory, but you may specify an alternate
4251 pathname here. Note you must give a full filename, not just
4252 a directory. Since this is the index for the whole object
4253 list you CANNOT periodically rotate it!
4255 If %s can be used in the file name it will be replaced with a
4256 a representation of the cache_dir name where each / is replaced
4257 with '.'. This is needed to allow adding/removing cache_dir
4258 lines when cache_swap_log is being used.
4260 If have more than one 'cache_dir', and %s is not used in the name
4261 these swap logs will have names such as:
4267 The numbered extension (which is added automatically)
4268 corresponds to the order of the 'cache_dir' lines in this
4269 configuration file. If you change the order of the 'cache_dir'
4270 lines in this file, these index files will NOT correspond to
4271 the correct 'cache_dir' entry (unless you manually rename
4272 them). We recommend you do NOT use this option. It is
4273 better to keep these index files in each 'cache_dir' directory.
4276 NAME: logfile_rotate
4279 LOC: Config.Log.rotateNumber
4281 Specifies the number of logfile rotations to make when you
4282 type 'squid -k rotate'. The default is 10, which will rotate
4283 with extensions 0 through 9. Setting logfile_rotate to 0 will
4284 disable the file name rotation, but the logfiles are still closed
4285 and re-opened. This will enable you to rename the logfiles
4286 yourself just before sending the rotate signal.
4288 Note, the 'squid -k rotate' command normally sends a USR1
4289 signal to the running squid process. In certain situations
4290 (e.g. on Linux with Async I/O), USR1 is used for other
4291 purposes, so -k rotate uses another signal. It is best to get
4292 in the habit of using 'squid -k rotate' instead of 'kill -USR1
4295 Note, from Squid-3.1 this option is only a default for cache.log,
4296 that log can be rotated separately by using debug_options.
4301 DEFAULT: @DEFAULT_MIME_TABLE@
4302 LOC: Config.mimeTablePathname
4304 Path to Squid's icon configuration file.
4306 You shouldn't need to change this, but the default file contains
4307 examples and formatting information if you do.
4313 LOC: Config.onoff.log_mime_hdrs
4316 The Cache can record both the request and the response MIME
4317 headers for each HTTP transaction. The headers are encoded
4318 safely and will appear as two bracketed fields at the end of
4319 the access log (for either the native or httpd-emulated log
4320 formats). To enable this logging set log_mime_hdrs to 'on'.
4325 DEFAULT: @DEFAULT_PID_FILE@
4326 LOC: Config.pidFilename
4328 A filename to write the process-id to. To disable, enter "none".
4331 NAME: client_netmask
4333 LOC: Config.Addrs.client_netmask
4335 DEFAULT_DOC: Log full client IP address
4337 A netmask for client addresses in logfiles and cachemgr output.
4338 Change this to protect the privacy of your cache clients.
4339 A netmask of 255.255.255.0 will log all IP's in that range with
4340 the last digit set to '0'.
4343 NAME: strip_query_terms
4345 LOC: Config.onoff.strip_query_terms
4348 By default, Squid strips query terms from requested URLs before
4349 logging. This protects your user's privacy and reduces log size.
4351 When investigating HIT/MISS or other caching behaviour you
4352 will need to disable this to see the full URL used by Squid.
4359 LOC: Config.onoff.buffered_logs
4361 Whether to write/send access_log records ASAP or accumulate them and
4362 then write/send them in larger chunks. Buffering may improve
4363 performance because it decreases the number of I/Os. However,
4364 buffering increases the delay before log records become available to
4365 the final recipient (e.g., a disk file or logging daemon) and,
4366 hence, increases the risk of log records loss.
4368 Note that even when buffered_logs are off, Squid may have to buffer
4369 records if it cannot write/send them immediately due to pending I/Os
4370 (e.g., the I/O writing the previous log record) or connectivity loss.
4372 Currently honored by 'daemon' and 'tcp' access_log modules only.
4375 NAME: netdb_filename
4377 DEFAULT: stdio:@DEFAULT_NETDB_FILE@
4378 LOC: Config.netdbFilename
4381 Where Squid stores it's netdb journal.
4382 When enabled this journal preserves netdb state between restarts.
4384 To disable, enter "none".
4388 OPTIONS FOR TROUBLESHOOTING
4389 -----------------------------------------------------------------------------
4394 DEFAULT_IF_NONE: @DEFAULT_CACHE_LOG@
4395 LOC: Debug::cache_log
4397 Squid administrative logging file.
4399 This is where general information about Squid behavior goes. You can
4400 increase the amount of data logged to this file and how often it is
4401 rotated with "debug_options"
4407 DEFAULT_DOC: Log all critical and important messages.
4408 LOC: Debug::debugOptions
4410 Logging options are set as section,level where each source file
4411 is assigned a unique section. Lower levels result in less
4412 output, Full debugging (level 9) can result in a very large
4413 log file, so be careful.
4415 The magic word "ALL" sets debugging levels for all sections.
4416 The default is to run with "ALL,1" to record important warnings.
4418 The rotate=N option can be used to keep more or less of these logs
4419 than would otherwise be kept by logfile_rotate.
4420 For most uses a single log should be enough to monitor current
4421 events affecting Squid.
4426 LOC: Config.coredump_dir
4427 DEFAULT_IF_NONE: none
4428 DEFAULT_DOC: Use the directory from where Squid was started.
4430 By default Squid leaves core files in the directory from where
4431 it was started. If you set 'coredump_dir' to a directory
4432 that exists, Squid will chdir() to that directory at startup
4433 and coredump files will be left there.
4437 # Leave coredumps in the first cache dir
4438 coredump_dir @DEFAULT_SWAP_DIR@
4444 OPTIONS FOR FTP GATEWAYING
4445 -----------------------------------------------------------------------------
4451 LOC: Config.Ftp.anon_user
4453 If you want the anonymous login password to be more informative
4454 (and enable the use of picky FTP servers), set this to something
4455 reasonable for your domain, like wwwuser@somewhere.net
4457 The reason why this is domainless by default is the
4458 request can be made on the behalf of a user in any domain,
4459 depending on how the cache is used.
4460 Some FTP server also validate the email address is valid
4461 (for example perl.com).
4467 LOC: Config.Ftp.passive
4469 If your firewall does not allow Squid to use passive
4470 connections, turn off this option.
4472 Use of ftp_epsv_all option requires this to be ON.
4478 LOC: Config.Ftp.epsv_all
4480 FTP Protocol extensions permit the use of a special "EPSV ALL" command.
4482 NATs may be able to put the connection on a "fast path" through the
4483 translator, as the EPRT command will never be used and therefore,
4484 translation of the data portion of the segments will never be needed.
4486 When a client only expects to do two-way FTP transfers this may be
4488 If squid finds that it must do a three-way FTP transfer after issuing
4489 an EPSV ALL command, the FTP session will fail.
4491 If you have any doubts about this option do not use it.
4492 Squid will nicely attempt all other connection methods.
4494 Requires ftp_passive to be ON (default) for any effect.
4500 LOC: Config.accessList.ftp_epsv
4502 FTP Protocol extensions permit the use of a special "EPSV" command.
4504 NATs may be able to put the connection on a "fast path" through the
4505 translator using EPSV, as the EPRT command will never be used
4506 and therefore, translation of the data portion of the segments
4507 will never be needed.
4509 EPSV is often required to interoperate with FTP servers on IPv6
4510 networks. On the other hand, it may break some IPv4 servers.
4512 By default, EPSV may try EPSV with any FTP server. To fine tune
4513 that decision, you may restrict EPSV to certain clients or servers
4516 ftp_epsv allow|deny al1 acl2 ...
4518 WARNING: Disabling EPSV may cause problems with external NAT and IPv6.
4520 Only fast ACLs are supported.
4521 Requires ftp_passive to be ON (default) for any effect.
4527 LOC: Config.Ftp.eprt
4529 FTP Protocol extensions permit the use of a special "EPRT" command.
4531 This extension provides a protocol neutral alternative to the
4532 IPv4-only PORT command. When supported it enables active FTP data
4533 channels over IPv6 and efficient NAT handling.
4535 Turning this OFF will prevent EPRT being attempted and will skip
4536 straight to using PORT for IPv4 servers.
4538 Some devices are known to not handle this extension correctly and
4539 may result in crashes. Devices which suport EPRT enough to fail
4540 cleanly will result in Squid attempting PORT anyway. This directive
4541 should only be disabled when EPRT results in device failures.
4543 WARNING: Doing so will convert Squid back to the old behavior with all
4544 the related problems with external NAT devices/layers and IPv4-only FTP.
4547 NAME: ftp_sanitycheck
4550 LOC: Config.Ftp.sanitycheck
4552 For security and data integrity reasons Squid by default performs
4553 sanity checks of the addresses of FTP data connections ensure the
4554 data connection is to the requested server. If you need to allow
4555 FTP connections to servers using another IP address for the data
4556 connection turn this off.
4559 NAME: ftp_telnet_protocol
4562 LOC: Config.Ftp.telnet
4564 The FTP protocol is officially defined to use the telnet protocol
4565 as transport channel for the control connection. However, many
4566 implementations are broken and does not respect this aspect of
4569 If you have trouble accessing files with ASCII code 255 in the
4570 path or similar problems involving this ASCII code you can
4571 try setting this directive to off. If that helps, report to the
4572 operator of the FTP server in question that their FTP server
4573 is broken and does not follow the FTP standard.
4577 OPTIONS FOR EXTERNAL SUPPORT PROGRAMS
4578 -----------------------------------------------------------------------------
4583 DEFAULT: @DEFAULT_DISKD@
4584 LOC: Config.Program.diskd
4586 Specify the location of the diskd executable.
4587 Note this is only useful if you have compiled in
4588 diskd as one of the store io modules.
4591 NAME: unlinkd_program
4594 DEFAULT: @DEFAULT_UNLINKD@
4595 LOC: Config.Program.unlinkd
4597 Specify the location of the executable for file deletion process.
4600 NAME: pinger_program
4602 DEFAULT: @DEFAULT_PINGER@
4603 LOC: Config.pinger.program
4606 Specify the location of the executable for the pinger process.
4612 LOC: Config.pinger.enable
4615 Control whether the pinger is active at run-time.
4616 Enables turning ICMP pinger on and off with a simple
4617 squid -k reconfigure.
4622 OPTIONS FOR URL REWRITING
4623 -----------------------------------------------------------------------------
4626 NAME: url_rewrite_program redirect_program
4628 LOC: Config.Program.redirect
4631 Specify the location of the executable URL rewriter to use.
4632 Since they can perform almost any function there isn't one included.
4634 For each requested URL, the rewriter will receive on line with the format
4636 [channel-ID <SP>] URL [<SP> extras]<NL>
4638 See url_rewrite_extras on how to send "extras" with optional values to
4640 After processing the request the helper must reply using the following format:
4642 [channel-ID <SP>] result [<SP> kv-pairs]
4644 The result code can be:
4646 OK status=30N url="..."
4647 Redirect the URL to the one supplied in 'url='.
4648 'status=' is optional and contains the status code to send
4649 the client in Squids HTTP response. It must be one of the
4650 HTTP redirect status codes: 301, 302, 303, 307, 308.
4651 When no status is given Squid will use 302.
4653 OK rewrite-url="..."
4654 Rewrite the URL to the one supplied in 'rewrite-url='.
4655 The new URL is fetched directly by Squid and returned to
4656 the client as the response to its request.
4659 When neither of url= and rewrite-url= are sent Squid does
4663 Do not change the URL.
4666 An internal error occurred in the helper, preventing
4667 a result being identified. The 'message=' key name is
4668 reserved for delivering a log message.
4671 In addition to the above kv-pairs Squid also understands the following
4672 optional kv-pairs received from URL rewriters:
4674 Associates a TAG with the client TCP connection.
4675 The TAG is treated as a regular annotation but persists across
4676 future requests on the client connection rather than just the
4677 current request. A helper may update the TAG during subsequent
4678 requests be returning a new kv-pair.
4680 When using the concurrency= option the protocol is changed by
4681 introducing a query channel tag in front of the request/response.
4682 The query channel tag is a number between 0 and concurrency-1.
4683 This value must be echoed back unchanged to Squid as the first part
4684 of the response relating to its request.
4686 WARNING: URL re-writing ability should be avoided whenever possible.
4687 Use the URL redirect form of response instead.
4689 Re-write creates a difference in the state held by the client
4690 and server. Possibly causing confusion when the server response
4691 contains snippets of its view state. Embeded URLs, response
4692 and content Location headers, etc. are not re-written by this
4695 By default, a URL rewriter is not used.
4698 NAME: url_rewrite_children redirect_children
4699 TYPE: HelperChildConfig
4700 DEFAULT: 20 startup=0 idle=1 concurrency=0
4701 LOC: Config.redirectChildren
4703 The maximum number of redirector processes to spawn. If you limit
4704 it too few Squid will have to wait for them to process a backlog of
4705 URLs, slowing it down. If you allow too many they will use RAM
4706 and other system resources noticably.
4708 The startup= and idle= options allow some measure of skew in your
4713 Sets a minimum of how many processes are to be spawned when Squid
4714 starts or reconfigures. When set to zero the first request will
4715 cause spawning of the first child process to handle it.
4717 Starting too few will cause an initial slowdown in traffic as Squid
4718 attempts to simultaneously spawn enough processes to cope.
4722 Sets a minimum of how many processes Squid is to try and keep available
4723 at all times. When traffic begins to rise above what the existing
4724 processes can handle this many more will be spawned up to the maximum
4725 configured. A minimum setting of 1 is required.
4729 The number of requests each redirector helper can handle in
4730 parallel. Defaults to 0 which indicates the redirector
4731 is a old-style single threaded redirector.
4733 When this directive is set to a value >= 1 then the protocol
4734 used to communicate with the helper is modified to include
4735 an ID in front of the request/response. The ID from the request
4736 must be echoed back with the response to that request.
4739 NAME: url_rewrite_host_header redirect_rewrites_host_header
4742 LOC: Config.onoff.redir_rewrites_host
4744 To preserve same-origin security policies in browsers and
4745 prevent Host: header forgery by redirectors Squid rewrites
4746 any Host: header in redirected requests.
4748 If you are running an accelerator this may not be a wanted
4749 effect of a redirector. This directive enables you disable
4750 Host: alteration in reverse-proxy traffic.
4752 WARNING: Entries are cached on the result of the URL rewriting
4753 process, so be careful if you have domain-virtual hosts.
4755 WARNING: Squid and other software verifies the URL and Host
4756 are matching, so be careful not to relay through other proxies
4757 or inspecting firewalls with this disabled.
4760 NAME: url_rewrite_access redirector_access
4763 DEFAULT_DOC: Allow, unless rules exist in squid.conf.
4764 LOC: Config.accessList.redirector
4766 If defined, this access list specifies which requests are
4767 sent to the redirector processes.
4769 This clause supports both fast and slow acl types.
4770 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
4773 NAME: url_rewrite_bypass redirector_bypass
4775 LOC: Config.onoff.redirector_bypass
4778 When this is 'on', a request will not go through the
4779 redirector if all the helpers are busy. If this is 'off'
4780 and the redirector queue grows too large, Squid will exit
4781 with a FATAL error and ask you to increase the number of
4782 redirectors. You should only enable this if the redirectors
4783 are not critical to your caching system. If you use
4784 redirectors for access control, and you enable this option,
4785 users may have access to pages they should not
4786 be allowed to request.
4789 NAME: url_rewrite_extras
4790 TYPE: TokenOrQuotedString
4791 LOC: Config.redirector_extras
4792 DEFAULT: "%>a/%>A %un %>rm myip=%la myport=%lp"
4794 Specifies a string to be append to request line format for the
4795 rewriter helper. "Quoted" format values may contain spaces and
4796 logformat %macros. In theory, any logformat %macro can be used.
4797 In practice, a %macro expands as a dash (-) if the helper request is
4798 sent before the required macro information is available to Squid.
4802 OPTIONS FOR STORE ID
4803 -----------------------------------------------------------------------------
4806 NAME: store_id_program storeurl_rewrite_program
4808 LOC: Config.Program.store_id
4811 Specify the location of the executable StoreID helper to use.
4812 Since they can perform almost any function there isn't one included.
4814 For each requested URL, the helper will receive one line with the format
4816 [channel-ID <SP>] URL [<SP> extras]<NL>
4819 After processing the request the helper must reply using the following format:
4821 [channel-ID <SP>] result [<SP> kv-pairs]
4823 The result code can be:
4826 Use the StoreID supplied in 'store-id='.
4829 The default is to use HTTP request URL as the store ID.
4832 An internal error occured in the helper, preventing
4833 a result being identified.
4835 In addition to the above kv-pairs Squid also understands the following
4836 optional kv-pairs received from URL rewriters:
4838 Associates a TAG with the client TCP connection.
4839 Please see url_rewrite_program related documentation for this
4842 Helper programs should be prepared to receive and possibly ignore
4843 additional whitespace-separated tokens on each input line.
4845 When using the concurrency= option the protocol is changed by
4846 introducing a query channel tag in front of the request/response.
4847 The query channel tag is a number between 0 and concurrency-1.
4848 This value must be echoed back unchanged to Squid as the first part
4849 of the response relating to its request.
4851 NOTE: when using StoreID refresh_pattern will apply to the StoreID
4852 returned from the helper and not the URL.
4854 WARNING: Wrong StoreID value returned by a careless helper may result
4855 in the wrong cached response returned to the user.
4857 By default, a StoreID helper is not used.
4860 NAME: store_id_extras
4861 TYPE: TokenOrQuotedString
4862 LOC: Config.storeId_extras
4863 DEFAULT: "%>a/%>A %un %>rm myip=%la myport=%lp"
4865 Specifies a string to be append to request line format for the
4866 StoreId helper. "Quoted" format values may contain spaces and
4867 logformat %macros. In theory, any logformat %macro can be used.
4868 In practice, a %macro expands as a dash (-) if the helper request is
4869 sent before the required macro information is available to Squid.
4872 NAME: store_id_children storeurl_rewrite_children
4873 TYPE: HelperChildConfig
4874 DEFAULT: 20 startup=0 idle=1 concurrency=0
4875 LOC: Config.storeIdChildren
4877 The maximum number of StoreID helper processes to spawn. If you limit
4878 it too few Squid will have to wait for them to process a backlog of
4879 requests, slowing it down. If you allow too many they will use RAM
4880 and other system resources noticably.
4882 The startup= and idle= options allow some measure of skew in your
4887 Sets a minimum of how many processes are to be spawned when Squid
4888 starts or reconfigures. When set to zero the first request will
4889 cause spawning of the first child process to handle it.
4891 Starting too few will cause an initial slowdown in traffic as Squid
4892 attempts to simultaneously spawn enough processes to cope.
4896 Sets a minimum of how many processes Squid is to try and keep available
4897 at all times. When traffic begins to rise above what the existing
4898 processes can handle this many more will be spawned up to the maximum
4899 configured. A minimum setting of 1 is required.
4903 The number of requests each storeID helper can handle in
4904 parallel. Defaults to 0 which indicates the helper
4905 is a old-style single threaded program.
4907 When this directive is set to a value >= 1 then the protocol
4908 used to communicate with the helper is modified to include
4909 an ID in front of the request/response. The ID from the request
4910 must be echoed back with the response to that request.
4913 NAME: store_id_access storeurl_rewrite_access
4916 DEFAULT_DOC: Allow, unless rules exist in squid.conf.
4917 LOC: Config.accessList.store_id
4919 If defined, this access list specifies which requests are
4920 sent to the StoreID processes. By default all requests
4923 This clause supports both fast and slow acl types.
4924 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
4927 NAME: store_id_bypass storeurl_rewrite_bypass
4929 LOC: Config.onoff.store_id_bypass
4932 When this is 'on', a request will not go through the
4933 helper if all helpers are busy. If this is 'off'
4934 and the helper queue grows too large, Squid will exit
4935 with a FATAL error and ask you to increase the number of
4936 helpers. You should only enable this if the helperss
4937 are not critical to your caching system. If you use
4938 helpers for critical caching components, and you enable this
4939 option, users may not get objects from cache.
4943 OPTIONS FOR TUNING THE CACHE
4944 -----------------------------------------------------------------------------
4947 NAME: cache no_cache
4950 DEFAULT_DOC: By default, this directive is unused and has no effect.
4951 LOC: Config.accessList.noCache
4953 Requests denied by this directive will not be served from the cache
4954 and their responses will not be stored in the cache. This directive
4955 has no effect on other transactions and on already cached responses.
4957 This clause supports both fast and slow acl types.
4958 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
4960 This and the two other similar caching directives listed below are
4961 checked at different transaction processing stages, have different
4962 access to response information, affect different cache operations,
4963 and differ in slow ACLs support:
4965 * cache: Checked before Squid makes a hit/miss determination.
4966 No access to reply information!
4967 Denies both serving a hit and storing a miss.
4968 Supports both fast and slow ACLs.
4969 * send_hit: Checked after a hit was detected.
4970 Has access to reply (hit) information.
4971 Denies serving a hit only.
4972 Supports fast ACLs only.
4973 * store_miss: Checked before storing a cachable miss.
4974 Has access to reply (miss) information.
4975 Denies storing a miss only.
4976 Supports fast ACLs only.
4978 If you are not sure which of the three directives to use, apply the
4979 following decision logic:
4981 * If your ACL(s) are of slow type _and_ need response info, redesign.
4982 Squid does not support that particular combination at this time.
4984 * If your directive ACL(s) are of slow type, use "cache"; and/or
4985 * if your directive ACL(s) need no response info, use "cache".
4987 * If you do not want the response cached, use store_miss; and/or
4988 * if you do not want a hit on a cached response, use send_hit.
4994 DEFAULT_DOC: By default, this directive is unused and has no effect.
4995 LOC: Config.accessList.sendHit
4997 Responses denied by this directive will not be served from the cache
4998 (but may still be cached, see store_miss). This directive has no
4999 effect on the responses it allows and on the cached objects.
5001 Please see the "cache" directive for a summary of differences among
5002 store_miss, send_hit, and cache directives.
5004 Unlike the "cache" directive, send_hit only supports fast acl
5005 types. See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
5009 # apply custom Store ID mapping to some URLs
5010 acl MapMe dstdomain .c.example.com
5011 store_id_program ...
5012 store_id_access allow MapMe
5014 # but prevent caching of special responses
5015 # such as 302 redirects that cause StoreID loops
5016 acl Ordinary http_status 200-299
5017 store_miss deny MapMe !Ordinary
5019 # and do not serve any previously stored special responses
5020 # from the cache (in case they were already cached before
5021 # the above store_miss rule was in effect).
5022 send_hit deny MapMe !Ordinary
5028 DEFAULT_DOC: By default, this directive is unused and has no effect.
5029 LOC: Config.accessList.storeMiss
5031 Responses denied by this directive will not be cached (but may still
5032 be served from the cache, see send_hit). This directive has no
5033 effect on the responses it allows and on the already cached responses.
5035 Please see the "cache" directive for a summary of differences among
5036 store_miss, send_hit, and cache directives. See the
5037 send_hit directive for a usage example.
5039 Unlike the "cache" directive, store_miss only supports fast acl
5040 types. See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
5046 LOC: Config.maxStale
5049 This option puts an upper limit on how stale content Squid
5050 will serve from the cache if cache validation fails.
5051 Can be overriden by the refresh_pattern max-stale option.
5054 NAME: refresh_pattern
5055 TYPE: refreshpattern
5059 usage: refresh_pattern [-i] regex min percent max [options]
5061 By default, regular expressions are CASE-SENSITIVE. To make
5062 them case-insensitive, use the -i option.
5064 'Min' is the time (in minutes) an object without an explicit
5065 expiry time should be considered fresh. The recommended
5066 value is 0, any higher values may cause dynamic applications
5067 to be erroneously cached unless the application designer
5068 has taken the appropriate actions.
5070 'Percent' is a percentage of the objects age (time since last
5071 modification age) an object without explicit expiry time
5072 will be considered fresh.
5074 'Max' is an upper limit on how long objects without an explicit
5075 expiry time will be considered fresh.
5077 options: override-expire
5082 ignore-must-revalidate
5089 override-expire enforces min age even if the server
5090 sent an explicit expiry time (e.g., with the
5091 Expires: header or Cache-Control: max-age). Doing this
5092 VIOLATES the HTTP standard. Enabling this feature
5093 could make you liable for problems which it causes.
5095 Note: override-expire does not enforce staleness - it only extends
5096 freshness / min. If the server returns a Expires time which
5097 is longer than your max time, Squid will still consider
5098 the object fresh for that period of time.
5100 override-lastmod enforces min age even on objects
5101 that were modified recently.
5103 reload-into-ims changes a client no-cache or ``reload''
5104 request for a cached entry into a conditional request using
5105 If-Modified-Since and/or If-None-Match headers, provided the
5106 cached entry has a Last-Modified and/or a strong ETag header.
5107 Doing this VIOLATES the HTTP standard. Enabling this feature
5108 could make you liable for problems which it causes.
5110 ignore-reload ignores a client no-cache or ``reload''
5111 header. Doing this VIOLATES the HTTP standard. Enabling
5112 this feature could make you liable for problems which
5115 ignore-no-store ignores any ``Cache-control: no-store''
5116 headers received from a server. Doing this VIOLATES
5117 the HTTP standard. Enabling this feature could make you
5118 liable for problems which it causes.
5120 ignore-must-revalidate ignores any ``Cache-Control: must-revalidate``
5121 headers received from a server. Doing this VIOLATES
5122 the HTTP standard. Enabling this feature could make you
5123 liable for problems which it causes.
5125 ignore-private ignores any ``Cache-control: private''
5126 headers received from a server. Doing this VIOLATES
5127 the HTTP standard. Enabling this feature could make you
5128 liable for problems which it causes.
5130 ignore-auth caches responses to requests with authorization,
5131 as if the originserver had sent ``Cache-control: public''
5132 in the response header. Doing this VIOLATES the HTTP standard.
5133 Enabling this feature could make you liable for problems which
5136 refresh-ims causes squid to contact the origin server
5137 when a client issues an If-Modified-Since request. This
5138 ensures that the client will receive an updated version
5139 if one is available.
5141 store-stale stores responses even if they don't have explicit
5142 freshness or a validator (i.e., Last-Modified or an ETag)
5143 present, or if they're already stale. By default, Squid will
5144 not cache such responses because they usually can't be
5145 reused. Note that such responses will be stale by default.
5147 max-stale=NN provide a maximum staleness factor. Squid won't
5148 serve objects more stale than this even if it failed to
5149 validate the object. Default: use the max_stale global limit.
5151 Basically a cached object is:
5153 FRESH if expires < now, else STALE
5155 FRESH if lm-factor < percent, else STALE
5159 The refresh_pattern lines are checked in the order listed here.
5160 The first entry which matches is used. If none of the entries
5161 match the default will be used.
5163 Note, you must uncomment all the default lines if you want
5164 to change one. The default setting is only active if none is
5170 # Add any of your own refresh_pattern entries above these.
5172 refresh_pattern ^ftp: 1440 20% 10080
5173 refresh_pattern ^gopher: 1440 0% 1440
5174 refresh_pattern -i (/cgi-bin/|\?) 0 0% 0
5175 refresh_pattern . 0 20% 4320
5179 NAME: quick_abort_min
5183 LOC: Config.quickAbort.min
5186 NAME: quick_abort_max
5190 LOC: Config.quickAbort.max
5193 NAME: quick_abort_pct
5197 LOC: Config.quickAbort.pct
5199 The cache by default continues downloading aborted requests
5200 which are almost completed (less than 16 KB remaining). This
5201 may be undesirable on slow (e.g. SLIP) links and/or very busy
5202 caches. Impatient users may tie up file descriptors and
5203 bandwidth by repeatedly requesting and immediately aborting
5206 When the user aborts a request, Squid will check the
5207 quick_abort values to the amount of data transferred until
5210 If the transfer has less than 'quick_abort_min' KB remaining,
5211 it will finish the retrieval.
5213 If the transfer has more than 'quick_abort_max' KB remaining,
5214 it will abort the retrieval.
5216 If more than 'quick_abort_pct' of the transfer has completed,
5217 it will finish the retrieval.
5219 If you do not want any retrieval to continue after the client
5220 has aborted, set both 'quick_abort_min' and 'quick_abort_max'
5223 If you want retrievals to always continue if they are being
5224 cached set 'quick_abort_min' to '-1 KB'.
5227 NAME: read_ahead_gap
5228 COMMENT: buffer-size
5230 LOC: Config.readAheadGap
5233 The amount of data the cache will buffer ahead of what has been
5234 sent to the client when retrieving an object from another server.
5238 IFDEF: USE_HTTP_VIOLATIONS
5241 LOC: Config.negativeTtl
5244 Set the Default Time-to-Live (TTL) for failed requests.
5245 Certain types of failures (such as "connection refused" and
5246 "404 Not Found") are able to be negatively-cached for a short time.
5247 Modern web servers should provide Expires: header, however if they
5248 do not this can provide a minimum TTL.
5249 The default is not to cache errors with unknown expiry details.
5251 Note that this is different from negative caching of DNS lookups.
5253 WARNING: Doing this VIOLATES the HTTP standard. Enabling
5254 this feature could make you liable for problems which it
5258 NAME: positive_dns_ttl
5261 LOC: Config.positiveDnsTtl
5264 Upper limit on how long Squid will cache positive DNS responses.
5265 Default is 6 hours (360 minutes). This directive must be set
5266 larger than negative_dns_ttl.
5269 NAME: negative_dns_ttl
5272 LOC: Config.negativeDnsTtl
5275 Time-to-Live (TTL) for negative caching of failed DNS lookups.
5276 This also sets the lower cache limit on positive lookups.
5277 Minimum value is 1 second, and it is not recommendable to go
5278 much below 10 seconds.
5281 NAME: range_offset_limit
5282 COMMENT: size [acl acl...]
5284 LOC: Config.rangeOffsetLimit
5287 usage: (size) [units] [[!]aclname]
5289 Sets an upper limit on how far (number of bytes) into the file
5290 a Range request may be to cause Squid to prefetch the whole file.
5291 If beyond this limit, Squid forwards the Range request as it is and
5292 the result is NOT cached.
5294 This is to stop a far ahead range request (lets say start at 17MB)
5295 from making Squid fetch the whole object up to that point before
5296 sending anything to the client.
5298 Multiple range_offset_limit lines may be specified, and they will
5299 be searched from top to bottom on each request until a match is found.
5300 The first match found will be used. If no line matches a request, the
5301 default limit of 0 bytes will be used.
5303 'size' is the limit specified as a number of units.
5305 'units' specifies whether to use bytes, KB, MB, etc.
5306 If no units are specified bytes are assumed.
5308 A size of 0 causes Squid to never fetch more than the
5309 client requested. (default)
5311 A size of 'none' causes Squid to always fetch the object from the
5312 beginning so it may cache the result. (2.0 style)
5314 'aclname' is the name of a defined ACL.
5316 NP: Using 'none' as the byte value here will override any quick_abort settings
5317 that may otherwise apply to the range request. The range request will
5318 be fully fetched from start to finish regardless of the client
5319 actions. This affects bandwidth usage.
5322 NAME: minimum_expiry_time
5325 LOC: Config.minimum_expiry_time
5328 The minimum caching time according to (Expires - Date)
5329 headers Squid honors if the object can't be revalidated.
5330 The default is 60 seconds.
5332 In reverse proxy environments it might be desirable to honor
5333 shorter object lifetimes. It is most likely better to make
5334 your server return a meaningful Last-Modified header however.
5336 In ESI environments where page fragments often have short
5337 lifetimes, this will often be best set to 0.
5340 NAME: store_avg_object_size
5344 LOC: Config.Store.avgObjectSize
5346 Average object size, used to estimate number of objects your
5347 cache can hold. The default is 13 KB.
5349 This is used to pre-seed the cache index memory allocation to
5350 reduce expensive reallocate operations while handling clients
5351 traffic. Too-large values may result in memory allocation during
5352 peak traffic, too-small values will result in wasted memory.
5354 Check the cache manager 'info' report metrics for the real
5355 object sizes seen by your Squid before tuning this.
5358 NAME: store_objects_per_bucket
5361 LOC: Config.Store.objectsPerBucket
5363 Target number of objects per bucket in the store hash table.
5364 Lowering this value increases the total number of buckets and
5365 also the storage maintenance rate. The default is 20.
5370 -----------------------------------------------------------------------------
5373 NAME: request_header_max_size
5377 LOC: Config.maxRequestHeaderSize
5379 This specifies the maximum size for HTTP headers in a request.
5380 Request headers are usually relatively small (about 512 bytes).
5381 Placing a limit on the request header size will catch certain
5382 bugs (for example with persistent connections) and possibly
5383 buffer-overflow or denial-of-service attacks.
5386 NAME: reply_header_max_size
5390 LOC: Config.maxReplyHeaderSize
5392 This specifies the maximum size for HTTP headers in a reply.
5393 Reply headers are usually relatively small (about 512 bytes).
5394 Placing a limit on the reply header size will catch certain
5395 bugs (for example with persistent connections) and possibly
5396 buffer-overflow or denial-of-service attacks.
5399 NAME: request_body_max_size
5403 DEFAULT_DOC: No limit.
5404 LOC: Config.maxRequestBodySize
5406 This specifies the maximum size for an HTTP request body.
5407 In other words, the maximum size of a PUT/POST request.
5408 A user who attempts to send a request with a body larger
5409 than this limit receives an "Invalid Request" error message.
5410 If you set this parameter to a zero (the default), there will
5411 be no limit imposed.
5413 See also client_request_buffer_max_size for an alternative
5414 limitation on client uploads which can be configured.
5417 NAME: client_request_buffer_max_size
5421 LOC: Config.maxRequestBufferSize
5423 This specifies the maximum buffer size of a client request.
5424 It prevents squid eating too much memory when somebody uploads
5428 NAME: chunked_request_body_max_size
5432 LOC: Config.maxChunkedRequestBodySize
5434 A broken or confused HTTP/1.1 client may send a chunked HTTP
5435 request to Squid. Squid does not have full support for that
5436 feature yet. To cope with such requests, Squid buffers the
5437 entire request and then dechunks request body to create a
5438 plain HTTP/1.0 request with a known content length. The plain
5439 request is then used by the rest of Squid code as usual.
5441 The option value specifies the maximum size of the buffer used
5442 to hold the request before the conversion. If the chunked
5443 request size exceeds the specified limit, the conversion
5444 fails, and the client receives an "unsupported request" error,
5445 as if dechunking was disabled.
5447 Dechunking is enabled by default. To disable conversion of
5448 chunked requests, set the maximum to zero.
5450 Request dechunking feature and this option in particular are a
5451 temporary hack. When chunking requests and responses are fully
5452 supported, there will be no need to buffer a chunked request.
5456 IFDEF: USE_HTTP_VIOLATIONS
5459 DEFAULT_DOC: Obey RFC 2616.
5460 LOC: Config.accessList.brokenPosts
5462 A list of ACL elements which, if matched, causes Squid to send
5463 an extra CRLF pair after the body of a PUT/POST request.
5465 Some HTTP servers has broken implementations of PUT/POST,
5466 and rely on an extra CRLF pair sent by some WWW clients.
5468 Quote from RFC2616 section 4.1 on this matter:
5470 Note: certain buggy HTTP/1.0 client implementations generate an
5471 extra CRLF's after a POST request. To restate what is explicitly
5472 forbidden by the BNF, an HTTP/1.1 client must not preface or follow
5473 a request with an extra CRLF.
5475 This clause only supports fast acl types.
5476 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
5479 acl buggy_server url_regex ^http://....
5480 broken_posts allow buggy_server
5483 NAME: adaptation_uses_indirect_client icap_uses_indirect_client
5486 IFDEF: FOLLOW_X_FORWARDED_FOR&&USE_ADAPTATION
5488 LOC: Adaptation::Config::use_indirect_client
5490 Controls whether the indirect client IP address (instead of the direct
5491 client IP address) is passed to adaptation services.
5493 See also: follow_x_forwarded_for adaptation_send_client_ip
5497 IFDEF: USE_HTTP_VIOLATIONS
5501 LOC: Config.onoff.via
5503 If set (default), Squid will include a Via header in requests and
5504 replies as required by RFC2616.
5510 LOC: Config.onoff.ie_refresh
5513 Microsoft Internet Explorer up until version 5.5 Service
5514 Pack 1 has an issue with transparent proxies, wherein it
5515 is impossible to force a refresh. Turning this on provides
5516 a partial fix to the problem, by causing all IMS-REFRESH
5517 requests from older IE versions to check the origin server
5518 for fresh content. This reduces hit ratio by some amount
5519 (~10% in my experience), but allows users to actually get
5520 fresh content when they want it. Note because Squid
5521 cannot tell if the user is using 5.5 or 5.5SP1, the behavior
5522 of 5.5 is unchanged from old versions of Squid (i.e. a
5523 forced refresh is impossible). Newer versions of IE will,
5524 hopefully, continue to have the new behavior and will be
5525 handled based on that assumption. This option defaults to
5526 the old Squid behavior, which is better for hit ratios but
5527 worse for clients using IE, if they need to be able to
5528 force fresh content.
5531 NAME: vary_ignore_expire
5534 LOC: Config.onoff.vary_ignore_expire
5537 Many HTTP servers supporting Vary gives such objects
5538 immediate expiry time with no cache-control header
5539 when requested by a HTTP/1.0 client. This option
5540 enables Squid to ignore such expiry times until
5541 HTTP/1.1 is fully implemented.
5543 WARNING: If turned on this may eventually cause some
5544 varying objects not intended for caching to get cached.
5547 NAME: request_entities
5549 LOC: Config.onoff.request_entities
5552 Squid defaults to deny GET and HEAD requests with request entities,
5553 as the meaning of such requests are undefined in the HTTP standard
5554 even if not explicitly forbidden.
5556 Set this directive to on if you have clients which insists
5557 on sending request entities in GET or HEAD requests. But be warned
5558 that there is server software (both proxies and web servers) which
5559 can fail to properly process this kind of request which may make you
5560 vulnerable to cache pollution attacks if enabled.
5563 NAME: request_header_access
5564 IFDEF: USE_HTTP_VIOLATIONS
5565 TYPE: http_header_access
5566 LOC: Config.request_header_access
5568 DEFAULT_DOC: No limits.
5570 Usage: request_header_access header_name allow|deny [!]aclname ...
5572 WARNING: Doing this VIOLATES the HTTP standard. Enabling
5573 this feature could make you liable for problems which it
5576 This option replaces the old 'anonymize_headers' and the
5577 older 'http_anonymizer' option with something that is much
5578 more configurable. A list of ACLs for each header name allows
5579 removal of specific header fields under specific conditions.
5581 This option only applies to outgoing HTTP request headers (i.e.,
5582 headers sent by Squid to the next HTTP hop such as a cache peer
5583 or an origin server). The option has no effect during cache hit
5584 detection. The equivalent adaptation vectoring point in ICAP
5585 terminology is post-cache REQMOD.
5587 The option is applied to individual outgoing request header
5588 fields. For each request header field F, Squid uses the first
5589 qualifying sets of request_header_access rules:
5591 1. Rules with header_name equal to F's name.
5592 2. Rules with header_name 'Other', provided F's name is not
5593 on the hard-coded list of commonly used HTTP header names.
5594 3. Rules with header_name 'All'.
5596 Within that qualifying rule set, rule ACLs are checked as usual.
5597 If ACLs of an "allow" rule match, the header field is allowed to
5598 go through as is. If ACLs of a "deny" rule match, the header is
5599 removed and request_header_replace is then checked to identify
5600 if the removed header has a replacement. If no rules within the
5601 set have matching ACLs, the header field is left as is.
5603 For example, to achieve the same behavior as the old
5604 'http_anonymizer standard' option, you should use:
5606 request_header_access From deny all
5607 request_header_access Referer deny all
5608 request_header_access User-Agent deny all
5610 Or, to reproduce the old 'http_anonymizer paranoid' feature
5613 request_header_access Authorization allow all
5614 request_header_access Proxy-Authorization allow all
5615 request_header_access Cache-Control allow all
5616 request_header_access Content-Length allow all
5617 request_header_access Content-Type allow all
5618 request_header_access Date allow all
5619 request_header_access Host allow all
5620 request_header_access If-Modified-Since allow all
5621 request_header_access Pragma allow all
5622 request_header_access Accept allow all
5623 request_header_access Accept-Charset allow all
5624 request_header_access Accept-Encoding allow all
5625 request_header_access Accept-Language allow all
5626 request_header_access Connection allow all
5627 request_header_access All deny all
5629 HTTP reply headers are controlled with the reply_header_access directive.
5631 By default, all headers are allowed (no anonymizing is performed).
5634 NAME: reply_header_access
5635 IFDEF: USE_HTTP_VIOLATIONS
5636 TYPE: http_header_access
5637 LOC: Config.reply_header_access
5639 DEFAULT_DOC: No limits.
5641 Usage: reply_header_access header_name allow|deny [!]aclname ...
5643 WARNING: Doing this VIOLATES the HTTP standard. Enabling
5644 this feature could make you liable for problems which it
5647 This option only applies to reply headers, i.e., from the
5648 server to the client.
5650 This is the same as request_header_access, but in the other
5651 direction. Please see request_header_access for detailed
5654 For example, to achieve the same behavior as the old
5655 'http_anonymizer standard' option, you should use:
5657 reply_header_access Server deny all
5658 reply_header_access WWW-Authenticate deny all
5659 reply_header_access Link deny all
5661 Or, to reproduce the old 'http_anonymizer paranoid' feature
5664 reply_header_access Allow allow all
5665 reply_header_access WWW-Authenticate allow all
5666 reply_header_access Proxy-Authenticate allow all
5667 reply_header_access Cache-Control allow all
5668 reply_header_access Content-Encoding allow all
5669 reply_header_access Content-Length allow all
5670 reply_header_access Content-Type allow all
5671 reply_header_access Date allow all
5672 reply_header_access Expires allow all
5673 reply_header_access Last-Modified allow all
5674 reply_header_access Location allow all
5675 reply_header_access Pragma allow all
5676 reply_header_access Content-Language allow all
5677 reply_header_access Retry-After allow all
5678 reply_header_access Title allow all
5679 reply_header_access Content-Disposition allow all
5680 reply_header_access Connection allow all
5681 reply_header_access All deny all
5683 HTTP request headers are controlled with the request_header_access directive.
5685 By default, all headers are allowed (no anonymizing is
5689 NAME: request_header_replace header_replace
5690 IFDEF: USE_HTTP_VIOLATIONS
5691 TYPE: http_header_replace
5692 LOC: Config.request_header_access
5695 Usage: request_header_replace header_name message
5696 Example: request_header_replace User-Agent Nutscrape/1.0 (CP/M; 8-bit)
5698 This option allows you to change the contents of headers
5699 denied with request_header_access above, by replacing them
5700 with some fixed string.
5702 This only applies to request headers, not reply headers.
5704 By default, headers are removed if denied.
5707 NAME: reply_header_replace
5708 IFDEF: USE_HTTP_VIOLATIONS
5709 TYPE: http_header_replace
5710 LOC: Config.reply_header_access
5713 Usage: reply_header_replace header_name message
5714 Example: reply_header_replace Server Foo/1.0
5716 This option allows you to change the contents of headers
5717 denied with reply_header_access above, by replacing them
5718 with some fixed string.
5720 This only applies to reply headers, not request headers.
5722 By default, headers are removed if denied.
5725 NAME: request_header_add
5726 TYPE: HeaderWithAclList
5727 LOC: Config.request_header_add
5730 Usage: request_header_add field-name field-value acl1 [acl2] ...
5731 Example: request_header_add X-Client-CA "CA=%ssl::>cert_issuer" all
5733 This option adds header fields to outgoing HTTP requests (i.e.,
5734 request headers sent by Squid to the next HTTP hop such as a
5735 cache peer or an origin server). The option has no effect during
5736 cache hit detection. The equivalent adaptation vectoring point
5737 in ICAP terminology is post-cache REQMOD.
5739 Field-name is a token specifying an HTTP header name. If a
5740 standard HTTP header name is used, Squid does not check whether
5741 the new header conflicts with any existing headers or violates
5742 HTTP rules. If the request to be modified already contains a
5743 field with the same name, the old field is preserved but the
5744 header field values are not merged.
5746 Field-value is either a token or a quoted string. If quoted
5747 string format is used, then the surrounding quotes are removed
5748 while escape sequences and %macros are processed.
5750 In theory, all of the logformat codes can be used as %macros.
5751 However, unlike logging (which happens at the very end of
5752 transaction lifetime), the transaction may not yet have enough
5753 information to expand a macro when the new header value is needed.
5754 And some information may already be available to Squid but not yet
5755 committed where the macro expansion code can access it (report
5756 such instances!). The macro will be expanded into a single dash
5757 ('-') in such cases. Not all macros have been tested.
5759 One or more Squid ACLs may be specified to restrict header
5760 injection to matching requests. As always in squid.conf, all
5761 ACLs in an option ACL list must be satisfied for the insertion
5762 to happen. The request_header_add option supports fast ACLs
5771 This option used to log custom information about the master
5772 transaction. For example, an admin may configure Squid to log
5773 which "user group" the transaction belongs to, where "user group"
5774 will be determined based on a set of ACLs and not [just]
5775 authentication information.
5776 Values of key/value pairs can be logged using %{key}note macros:
5778 note key value acl ...
5779 logformat myFormat ... %{key}note ...
5782 NAME: relaxed_header_parser
5783 COMMENT: on|off|warn
5785 LOC: Config.onoff.relaxed_header_parser
5788 In the default "on" setting Squid accepts certain forms
5789 of non-compliant HTTP messages where it is unambiguous
5790 what the sending application intended even if the message
5791 is not correctly formatted. The messages is then normalized
5792 to the correct form when forwarded by Squid.
5794 If set to "warn" then a warning will be emitted in cache.log
5795 each time such HTTP error is encountered.
5797 If set to "off" then such HTTP errors will cause the request
5798 or response to be rejected.
5801 NAME: collapsed_forwarding
5804 LOC: Config.onoff.collapsed_forwarding
5807 This option controls whether Squid is allowed to merge multiple
5808 potentially cachable requests for the same URI before Squid knows
5809 whether the response is going to be cachable.
5811 This feature is disabled by default: Enabling collapsed forwarding
5812 needlessly delays forwarding requests that look cachable (when they are
5813 collapsed) but then need to be forwarded individually anyway because
5814 they end up being for uncachable content. However, in some cases, such
5815 as accelleration of highly cachable content with periodic or groupped
5816 expiration times, the gains from collapsing [large volumes of
5817 simultenous refresh requests] outweigh losses from such delays.
5822 -----------------------------------------------------------------------------
5825 NAME: forward_timeout
5828 LOC: Config.Timeout.forward
5831 This parameter specifies how long Squid should at most attempt in
5832 finding a forwarding path for the request before giving up.
5835 NAME: connect_timeout
5838 LOC: Config.Timeout.connect
5841 This parameter specifies how long to wait for the TCP connect to
5842 the requested server or peer to complete before Squid should
5843 attempt to find another path where to forward the request.
5846 NAME: peer_connect_timeout
5849 LOC: Config.Timeout.peer_connect
5852 This parameter specifies how long to wait for a pending TCP
5853 connection to a peer cache. The default is 30 seconds. You
5854 may also set different timeout values for individual neighbors
5855 with the 'connect-timeout' option on a 'cache_peer' line.
5861 LOC: Config.Timeout.read
5864 The read_timeout is applied on server-side connections. After
5865 each successful read(), the timeout will be extended by this
5866 amount. If no data is read again after this amount of time,
5867 the request is aborted and logged with ERR_READ_TIMEOUT. The
5868 default is 15 minutes.
5874 LOC: Config.Timeout.write
5877 This timeout is tracked for all connections that have data
5878 available for writing and are waiting for the socket to become
5879 ready. After each successful write, the timeout is extended by
5880 the configured amount. If Squid has data to write but the
5881 connection is not ready for the configured duration, the
5882 transaction associated with the connection is terminated. The
5883 default is 15 minutes.
5886 NAME: request_timeout
5888 LOC: Config.Timeout.request
5891 How long to wait for complete HTTP request headers after initial
5892 connection establishment.
5895 NAME: client_idle_pconn_timeout persistent_request_timeout
5897 LOC: Config.Timeout.clientIdlePconn
5900 How long to wait for the next HTTP request on a persistent
5901 client connection after the previous request completes.
5904 NAME: client_lifetime
5907 LOC: Config.Timeout.lifetime
5910 The maximum amount of time a client (browser) is allowed to
5911 remain connected to the cache process. This protects the Cache
5912 from having a lot of sockets (and hence file descriptors) tied up
5913 in a CLOSE_WAIT state from remote clients that go away without
5914 properly shutting down (either because of a network failure or
5915 because of a poor client implementation). The default is one
5918 NOTE: The default value is intended to be much larger than any
5919 client would ever need to be connected to your cache. You
5920 should probably change client_lifetime only as a last resort.
5921 If you seem to have many client connections tying up
5922 filedescriptors, we recommend first tuning the read_timeout,
5923 request_timeout, persistent_request_timeout and quick_abort values.
5926 NAME: half_closed_clients
5928 LOC: Config.onoff.half_closed_clients
5931 Some clients may shutdown the sending side of their TCP
5932 connections, while leaving their receiving sides open. Sometimes,
5933 Squid can not tell the difference between a half-closed and a
5934 fully-closed TCP connection.
5936 By default, Squid will immediately close client connections when
5937 read(2) returns "no more data to read."
5939 Change this option to 'on' and Squid will keep open connections
5940 until a read(2) or write(2) on the socket returns an error.
5941 This may show some benefits for reverse proxies. But if not
5942 it is recommended to leave OFF.
5945 NAME: server_idle_pconn_timeout pconn_timeout
5947 LOC: Config.Timeout.serverIdlePconn
5950 Timeout for idle persistent connections to servers and other
5957 LOC: Ident::TheConfig.timeout
5960 Maximum time to wait for IDENT lookups to complete.
5962 If this is too high, and you enabled IDENT lookups from untrusted
5963 users, you might be susceptible to denial-of-service by having
5964 many ident requests going at once.
5967 NAME: shutdown_lifetime
5970 LOC: Config.shutdownLifetime
5973 When SIGTERM or SIGHUP is received, the cache is put into
5974 "shutdown pending" mode until all active sockets are closed.
5975 This value is the lifetime to set for all open descriptors
5976 during shutdown mode. Any active clients after this many
5977 seconds will receive a 'timeout' message.
5981 ADMINISTRATIVE PARAMETERS
5982 -----------------------------------------------------------------------------
5988 LOC: Config.adminEmail
5990 Email-address of local cache manager who will receive
5991 mail if the cache dies. The default is "webmaster".
5997 LOC: Config.EmailFrom
5999 From: email-address for mail sent when the cache dies.
6000 The default is to use 'squid@unique_hostname'.
6002 See also: unique_hostname directive.
6008 LOC: Config.EmailProgram
6010 Email program used to send mail if the cache dies.
6011 The default is "mail". The specified program must comply
6012 with the standard Unix mail syntax:
6013 mail-program recipient < mailfile
6015 Optional command line options can be specified.
6018 NAME: cache_effective_user
6020 DEFAULT: @DEFAULT_CACHE_EFFECTIVE_USER@
6021 LOC: Config.effectiveUser
6023 If you start Squid as root, it will change its effective/real
6024 UID/GID to the user specified below. The default is to change
6025 to UID of @DEFAULT_CACHE_EFFECTIVE_USER@.
6026 see also; cache_effective_group
6029 NAME: cache_effective_group
6032 DEFAULT_DOC: Use system group memberships of the cache_effective_user account
6033 LOC: Config.effectiveGroup
6035 Squid sets the GID to the effective user's default group ID
6036 (taken from the password file) and supplementary group list
6037 from the groups membership.
6039 If you want Squid to run with a specific GID regardless of
6040 the group memberships of the effective user then set this
6041 to the group (or GID) you want Squid to run as. When set
6042 all other group privileges of the effective user are ignored
6043 and only this GID is effective. If Squid is not started as
6044 root the user starting Squid MUST be member of the specified
6047 This option is not recommended by the Squid Team.
6048 Our preference is for administrators to configure a secure
6049 user account for squid with UID/GID matching system policies.
6052 NAME: httpd_suppress_version_string
6056 LOC: Config.onoff.httpd_suppress_version_string
6058 Suppress Squid version string info in HTTP headers and HTML error pages.
6061 NAME: visible_hostname
6063 LOC: Config.visibleHostname
6065 DEFAULT_DOC: Automatically detect the system host name
6067 If you want to present a special hostname in error messages, etc,
6068 define this. Otherwise, the return value of gethostname()
6069 will be used. If you have multiple caches in a cluster and
6070 get errors about IP-forwarding you must set them to have individual
6071 names with this setting.
6074 NAME: unique_hostname
6076 LOC: Config.uniqueHostname
6078 DEFAULT_DOC: Copy the value from visible_hostname
6080 If you want to have multiple machines with the same
6081 'visible_hostname' you must give each machine a different
6082 'unique_hostname' so forwarding loops can be detected.
6085 NAME: hostname_aliases
6087 LOC: Config.hostnameAliases
6090 A list of other DNS names your cache has.
6098 Minimum umask which should be enforced while the proxy
6099 is running, in addition to the umask set at startup.
6101 For a traditional octal representation of umasks, start
6106 OPTIONS FOR THE CACHE REGISTRATION SERVICE
6107 -----------------------------------------------------------------------------
6109 This section contains parameters for the (optional) cache
6110 announcement service. This service is provided to help
6111 cache administrators locate one another in order to join or
6112 create cache hierarchies.
6114 An 'announcement' message is sent (via UDP) to the registration
6115 service by Squid. By default, the announcement message is NOT
6116 SENT unless you enable it with 'announce_period' below.
6118 The announcement message includes your hostname, plus the
6119 following information from this configuration file:
6125 All current information is processed regularly and made
6126 available on the Web at http://www.ircache.net/Cache/Tracker/.
6129 NAME: announce_period
6131 LOC: Config.Announce.period
6133 DEFAULT_DOC: Announcement messages disabled.
6135 This is how frequently to send cache announcements.
6137 To enable announcing your cache, just set an announce period.
6140 announce_period 1 day
6145 DEFAULT: tracker.ircache.net
6146 LOC: Config.Announce.host
6148 Set the hostname where announce registration messages will be sent.
6150 See also announce_port and announce_file
6156 LOC: Config.Announce.file
6158 The contents of this file will be included in the announce
6159 registration messages.
6165 LOC: Config.Announce.port
6167 Set the port where announce registration messages will be sent.
6169 See also announce_host and announce_file
6173 HTTPD-ACCELERATOR OPTIONS
6174 -----------------------------------------------------------------------------
6177 NAME: httpd_accel_surrogate_id
6180 DEFAULT_DOC: visible_hostname is used if no specific ID is set.
6181 LOC: Config.Accel.surrogate_id
6183 Surrogates (http://www.esi.org/architecture_spec_1.0.html)
6184 need an identification token to allow control targeting. Because
6185 a farm of surrogates may all perform the same tasks, they may share
6186 an identification token.
6189 NAME: http_accel_surrogate_remote
6193 LOC: Config.onoff.surrogate_is_remote
6195 Remote surrogates (such as those in a CDN) honour the header
6196 "Surrogate-Control: no-store-remote".
6198 Set this to on to have squid behave as a remote surrogate.
6202 IFDEF: USE_SQUID_ESI
6203 COMMENT: libxml2|expat|custom
6205 LOC: ESIParser::Type
6208 ESI markup is not strictly XML compatible. The custom ESI parser
6209 will give higher performance, but cannot handle non ASCII character
6214 DELAY POOL PARAMETERS
6215 -----------------------------------------------------------------------------
6219 TYPE: delay_pool_count
6221 IFDEF: USE_DELAY_POOLS
6224 This represents the number of delay pools to be used. For example,
6225 if you have one class 2 delay pool and one class 3 delays pool, you
6226 have a total of 2 delay pools.
6228 See also delay_parameters, delay_class, delay_access for pool
6229 configuration details.
6233 TYPE: delay_pool_class
6235 IFDEF: USE_DELAY_POOLS
6238 This defines the class of each delay pool. There must be exactly one
6239 delay_class line for each delay pool. For example, to define two
6240 delay pools, one of class 2 and one of class 3, the settings above
6244 delay_pools 4 # 4 delay pools
6245 delay_class 1 2 # pool 1 is a class 2 pool
6246 delay_class 2 3 # pool 2 is a class 3 pool
6247 delay_class 3 4 # pool 3 is a class 4 pool
6248 delay_class 4 5 # pool 4 is a class 5 pool
6250 The delay pool classes are:
6252 class 1 Everything is limited by a single aggregate
6255 class 2 Everything is limited by a single aggregate
6256 bucket as well as an "individual" bucket chosen
6257 from bits 25 through 32 of the IPv4 address.
6259 class 3 Everything is limited by a single aggregate
6260 bucket as well as a "network" bucket chosen
6261 from bits 17 through 24 of the IP address and a
6262 "individual" bucket chosen from bits 17 through
6263 32 of the IPv4 address.
6265 class 4 Everything in a class 3 delay pool, with an
6266 additional limit on a per user basis. This
6267 only takes effect if the username is established
6268 in advance - by forcing authentication in your
6271 class 5 Requests are grouped according their tag (see
6272 external_acl's tag= reply).
6275 Each pool also requires a delay_parameters directive to configure the pool size
6276 and speed limits used whenever the pool is applied to a request. Along with
6277 a set of delay_access directives to determine when it is used.
6279 NOTE: If an IP address is a.b.c.d
6280 -> bits 25 through 32 are "d"
6281 -> bits 17 through 24 are "c"
6282 -> bits 17 through 32 are "c * 256 + d"
6284 NOTE-2: Due to the use of bitmasks in class 2,3,4 pools they only apply to
6285 IPv4 traffic. Class 1 and 5 pools may be used with IPv6 traffic.
6287 This clause only supports fast acl types.
6288 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
6290 See also delay_parameters and delay_access.
6294 TYPE: delay_pool_access
6296 DEFAULT_DOC: Deny using the pool, unless allow rules exist in squid.conf for the pool.
6297 IFDEF: USE_DELAY_POOLS
6300 This is used to determine which delay pool a request falls into.
6302 delay_access is sorted per pool and the matching starts with pool 1,
6303 then pool 2, ..., and finally pool N. The first delay pool where the
6304 request is allowed is selected for the request. If it does not allow
6305 the request to any pool then the request is not delayed (default).
6307 For example, if you want some_big_clients in delay
6308 pool 1 and lotsa_little_clients in delay pool 2:
6310 delay_access 1 allow some_big_clients
6311 delay_access 1 deny all
6312 delay_access 2 allow lotsa_little_clients
6313 delay_access 2 deny all
6314 delay_access 3 allow authenticated_clients
6316 See also delay_parameters and delay_class.
6320 NAME: delay_parameters
6321 TYPE: delay_pool_rates
6323 IFDEF: USE_DELAY_POOLS
6326 This defines the parameters for a delay pool. Each delay pool has
6327 a number of "buckets" associated with it, as explained in the
6328 description of delay_class.
6330 For a class 1 delay pool, the syntax is:
6332 delay_parameters pool aggregate
6334 For a class 2 delay pool:
6336 delay_parameters pool aggregate individual
6338 For a class 3 delay pool:
6340 delay_parameters pool aggregate network individual
6342 For a class 4 delay pool:
6344 delay_parameters pool aggregate network individual user
6346 For a class 5 delay pool:
6348 delay_parameters pool tagrate
6350 The option variables are:
6352 pool a pool number - ie, a number between 1 and the
6353 number specified in delay_pools as used in
6356 aggregate the speed limit parameters for the aggregate bucket
6359 individual the speed limit parameters for the individual
6360 buckets (class 2, 3).
6362 network the speed limit parameters for the network buckets
6365 user the speed limit parameters for the user buckets
6368 tagrate the speed limit parameters for the tag buckets
6371 A pair of delay parameters is written restore/maximum, where restore is
6372 the number of bytes (not bits - modem and network speeds are usually
6373 quoted in bits) per second placed into the bucket, and maximum is the
6374 maximum number of bytes which can be in the bucket at any time.
6376 There must be one delay_parameters line for each delay pool.
6379 For example, if delay pool number 1 is a class 2 delay pool as in the
6380 above example, and is being used to strictly limit each host to 64Kbit/sec
6381 (plus overheads), with no overall limit, the line is:
6383 delay_parameters 1 -1/-1 8000/8000
6385 Note that 8 x 8000 KByte/sec -> 64Kbit/sec.
6387 Note that the figure -1 is used to represent "unlimited".
6390 And, if delay pool number 2 is a class 3 delay pool as in the above
6391 example, and you want to limit it to a total of 256Kbit/sec (strict limit)
6392 with each 8-bit network permitted 64Kbit/sec (strict limit) and each
6393 individual host permitted 4800bit/sec with a bucket maximum size of 64Kbits
6394 to permit a decent web page to be downloaded at a decent speed
6395 (if the network is not being limited due to overuse) but slow down
6396 large downloads more significantly:
6398 delay_parameters 2 32000/32000 8000/8000 600/8000
6400 Note that 8 x 32000 KByte/sec -> 256Kbit/sec.
6401 8 x 8000 KByte/sec -> 64Kbit/sec.
6402 8 x 600 Byte/sec -> 4800bit/sec.
6405 Finally, for a class 4 delay pool as in the example - each user will
6406 be limited to 128Kbits/sec no matter how many workstations they are logged into.:
6408 delay_parameters 4 32000/32000 8000/8000 600/64000 16000/16000
6411 See also delay_class and delay_access.
6415 NAME: delay_initial_bucket_level
6416 COMMENT: (percent, 0-100)
6419 IFDEF: USE_DELAY_POOLS
6420 LOC: Config.Delay.initial
6422 The initial bucket percentage is used to determine how much is put
6423 in each bucket when squid starts, is reconfigured, or first notices
6424 a host accessing it (in class 2 and class 3, individual hosts and
6425 networks only have buckets associated with them once they have been
6430 CLIENT DELAY POOL PARAMETERS
6431 -----------------------------------------------------------------------------
6434 NAME: client_delay_pools
6435 TYPE: client_delay_pool_count
6437 IFDEF: USE_DELAY_POOLS
6438 LOC: Config.ClientDelay
6440 This option specifies the number of client delay pools used. It must
6441 preceed other client_delay_* options.
6444 client_delay_pools 2
6446 See also client_delay_parameters and client_delay_access.
6449 NAME: client_delay_initial_bucket_level
6450 COMMENT: (percent, 0-no_limit)
6453 IFDEF: USE_DELAY_POOLS
6454 LOC: Config.ClientDelay.initial
6456 This option determines the initial bucket size as a percentage of
6457 max_bucket_size from client_delay_parameters. Buckets are created
6458 at the time of the "first" connection from the matching IP. Idle
6459 buckets are periodically deleted up.
6461 You can specify more than 100 percent but note that such "oversized"
6462 buckets are not refilled until their size goes down to max_bucket_size
6463 from client_delay_parameters.
6466 client_delay_initial_bucket_level 50
6469 NAME: client_delay_parameters
6470 TYPE: client_delay_pool_rates
6472 IFDEF: USE_DELAY_POOLS
6473 LOC: Config.ClientDelay
6476 This option configures client-side bandwidth limits using the
6479 client_delay_parameters pool speed_limit max_bucket_size
6481 pool is an integer ID used for client_delay_access matching.
6483 speed_limit is bytes added to the bucket per second.
6485 max_bucket_size is the maximum size of a bucket, enforced after any
6486 speed_limit additions.
6488 Please see the delay_parameters option for more information and
6492 client_delay_parameters 1 1024 2048
6493 client_delay_parameters 2 51200 16384
6495 See also client_delay_access.
6499 NAME: client_delay_access
6500 TYPE: client_delay_pool_access
6502 DEFAULT_DOC: Deny use of the pool, unless allow rules exist in squid.conf for the pool.
6503 IFDEF: USE_DELAY_POOLS
6504 LOC: Config.ClientDelay
6506 This option determines the client-side delay pool for the
6509 client_delay_access pool_ID allow|deny acl_name
6511 All client_delay_access options are checked in their pool ID
6512 order, starting with pool 1. The first checked pool with allowed
6513 request is selected for the request. If no ACL matches or there
6514 are no client_delay_access options, the request bandwidth is not
6517 The ACL-selected pool is then used to find the
6518 client_delay_parameters for the request. Client-side pools are
6519 not used to aggregate clients. Clients are always aggregated
6520 based on their source IP addresses (one bucket per source IP).
6522 This clause only supports fast acl types.
6523 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
6524 Additionally, only the client TCP connection details are available.
6525 ACLs testing HTTP properties will not work.
6527 Please see delay_access for more examples.
6530 client_delay_access 1 allow low_rate_network
6531 client_delay_access 2 allow vips_network
6534 See also client_delay_parameters and client_delay_pools.
6538 WCCPv1 AND WCCPv2 CONFIGURATION OPTIONS
6539 -----------------------------------------------------------------------------
6544 LOC: Config.Wccp.router
6546 DEFAULT_DOC: WCCP disabled.
6549 Use this option to define your WCCP ``home'' router for
6552 wccp_router supports a single WCCP(v1) router
6554 wccp2_router supports multiple WCCPv2 routers
6556 only one of the two may be used at the same time and defines
6557 which version of WCCP to use.
6561 TYPE: IpAddress_list
6562 LOC: Config.Wccp2.router
6564 DEFAULT_DOC: WCCPv2 disabled.
6567 Use this option to define your WCCP ``home'' router for
6570 wccp_router supports a single WCCP(v1) router
6572 wccp2_router supports multiple WCCPv2 routers
6574 only one of the two may be used at the same time and defines
6575 which version of WCCP to use.
6580 LOC: Config.Wccp.version
6584 This directive is only relevant if you need to set up WCCP(v1)
6585 to some very old and end-of-life Cisco routers. In all other
6586 setups it must be left unset or at the default setting.
6587 It defines an internal version in the WCCP(v1) protocol,
6588 with version 4 being the officially documented protocol.
6590 According to some users, Cisco IOS 11.2 and earlier only
6591 support WCCP version 3. If you're using that or an earlier
6592 version of IOS, you may need to change this value to 3, otherwise
6593 do not specify this parameter.
6596 NAME: wccp2_rebuild_wait
6598 LOC: Config.Wccp2.rebuildwait
6602 If this is enabled Squid will wait for the cache dir rebuild to finish
6603 before sending the first wccp2 HereIAm packet
6606 NAME: wccp2_forwarding_method
6608 LOC: Config.Wccp2.forwarding_method
6612 WCCP2 allows the setting of forwarding methods between the
6613 router/switch and the cache. Valid values are as follows:
6615 gre - GRE encapsulation (forward the packet in a GRE/WCCP tunnel)
6616 l2 - L2 redirect (forward the packet using Layer 2/MAC rewriting)
6618 Currently (as of IOS 12.4) cisco routers only support GRE.
6619 Cisco switches only support the L2 redirect assignment method.
6622 NAME: wccp2_return_method
6624 LOC: Config.Wccp2.return_method
6628 WCCP2 allows the setting of return methods between the
6629 router/switch and the cache for packets that the cache
6630 decides not to handle. Valid values are as follows:
6632 gre - GRE encapsulation (forward the packet in a GRE/WCCP tunnel)
6633 l2 - L2 redirect (forward the packet using Layer 2/MAC rewriting)
6635 Currently (as of IOS 12.4) cisco routers only support GRE.
6636 Cisco switches only support the L2 redirect assignment.
6638 If the "ip wccp redirect exclude in" command has been
6639 enabled on the cache interface, then it is still safe for
6640 the proxy server to use a l2 redirect method even if this
6641 option is set to GRE.
6644 NAME: wccp2_assignment_method
6646 LOC: Config.Wccp2.assignment_method
6650 WCCP2 allows the setting of methods to assign the WCCP hash
6651 Valid values are as follows:
6653 hash - Hash assignment
6654 mask - Mask assignment
6656 As a general rule, cisco routers support the hash assignment method
6657 and cisco switches support the mask assignment method.
6662 LOC: Config.Wccp2.info
6663 DEFAULT_IF_NONE: standard 0
6664 DEFAULT_DOC: Use the 'web-cache' standard service.
6667 WCCP2 allows for multiple traffic services. There are two
6668 types: "standard" and "dynamic". The standard type defines
6669 one service id - http (id 0). The dynamic service ids can be from
6670 51 to 255 inclusive. In order to use a dynamic service id
6671 one must define the type of traffic to be redirected; this is done
6672 using the wccp2_service_info option.
6674 The "standard" type does not require a wccp2_service_info option,
6675 just specifying the service id will suffice.
6677 MD5 service authentication can be enabled by adding
6678 "password=<password>" to the end of this service declaration.
6682 wccp2_service standard 0 # for the 'web-cache' standard service
6683 wccp2_service dynamic 80 # a dynamic service type which will be
6684 # fleshed out with subsequent options.
6685 wccp2_service standard 0 password=foo
6688 NAME: wccp2_service_info
6689 TYPE: wccp2_service_info
6690 LOC: Config.Wccp2.info
6694 Dynamic WCCPv2 services require further information to define the
6695 traffic you wish to have diverted.
6699 wccp2_service_info <id> protocol=<protocol> flags=<flag>,<flag>..
6700 priority=<priority> ports=<port>,<port>..
6702 The relevant WCCPv2 flags:
6703 + src_ip_hash, dst_ip_hash
6704 + source_port_hash, dst_port_hash
6705 + src_ip_alt_hash, dst_ip_alt_hash
6706 + src_port_alt_hash, dst_port_alt_hash
6709 The port list can be one to eight entries.
6713 wccp2_service_info 80 protocol=tcp flags=src_ip_hash,ports_source
6714 priority=240 ports=80
6716 Note: the service id must have been defined by a previous
6717 'wccp2_service dynamic <id>' entry.
6722 LOC: Config.Wccp2.weight
6726 Each cache server gets assigned a set of the destination
6727 hash proportional to their weight.
6732 LOC: Config.Wccp.address
6734 DEFAULT_DOC: Address selected by the operating system.
6737 Use this option if you require WCCPv2 to use a specific
6740 The default behavior is to not bind to any specific address.
6745 LOC: Config.Wccp2.address
6747 DEFAULT_DOC: Address selected by the operating system.
6750 Use this option if you require WCCP to use a specific
6753 The default behavior is to not bind to any specific address.
6757 PERSISTENT CONNECTION HANDLING
6758 -----------------------------------------------------------------------------
6760 Also see "pconn_timeout" in the TIMEOUTS section
6763 NAME: client_persistent_connections
6765 LOC: Config.onoff.client_pconns
6768 Persistent connection support for clients.
6769 Squid uses persistent connections (when allowed). You can use
6770 this option to disable persistent connections with clients.
6773 NAME: server_persistent_connections
6775 LOC: Config.onoff.server_pconns
6778 Persistent connection support for servers.
6779 Squid uses persistent connections (when allowed). You can use
6780 this option to disable persistent connections with servers.
6783 NAME: persistent_connection_after_error
6785 LOC: Config.onoff.error_pconns
6788 With this directive the use of persistent connections after
6789 HTTP errors can be disabled. Useful if you have clients
6790 who fail to handle errors on persistent connections proper.
6793 NAME: detect_broken_pconn
6795 LOC: Config.onoff.detect_broken_server_pconns
6798 Some servers have been found to incorrectly signal the use
6799 of HTTP/1.0 persistent connections even on replies not
6800 compatible, causing significant delays. This server problem
6801 has mostly been seen on redirects.
6803 By enabling this directive Squid attempts to detect such
6804 broken replies and automatically assume the reply is finished
6805 after 10 seconds timeout.
6809 CACHE DIGEST OPTIONS
6810 -----------------------------------------------------------------------------
6813 NAME: digest_generation
6814 IFDEF: USE_CACHE_DIGESTS
6816 LOC: Config.onoff.digest_generation
6819 This controls whether the server will generate a Cache Digest
6820 of its contents. By default, Cache Digest generation is
6821 enabled if Squid is compiled with --enable-cache-digests defined.
6824 NAME: digest_bits_per_entry
6825 IFDEF: USE_CACHE_DIGESTS
6827 LOC: Config.digest.bits_per_entry
6830 This is the number of bits of the server's Cache Digest which
6831 will be associated with the Digest entry for a given HTTP
6832 Method and URL (public key) combination. The default is 5.
6835 NAME: digest_rebuild_period
6836 IFDEF: USE_CACHE_DIGESTS
6839 LOC: Config.digest.rebuild_period
6842 This is the wait time between Cache Digest rebuilds.
6845 NAME: digest_rewrite_period
6847 IFDEF: USE_CACHE_DIGESTS
6849 LOC: Config.digest.rewrite_period
6852 This is the wait time between Cache Digest writes to
6856 NAME: digest_swapout_chunk_size
6859 IFDEF: USE_CACHE_DIGESTS
6860 LOC: Config.digest.swapout_chunk_size
6863 This is the number of bytes of the Cache Digest to write to
6864 disk at a time. It defaults to 4096 bytes (4KB), the Squid
6868 NAME: digest_rebuild_chunk_percentage
6869 COMMENT: (percent, 0-100)
6870 IFDEF: USE_CACHE_DIGESTS
6872 LOC: Config.digest.rebuild_chunk_percentage
6875 This is the percentage of the Cache Digest to be scanned at a
6876 time. By default it is set to 10% of the Cache Digest.
6881 -----------------------------------------------------------------------------
6886 LOC: Config.Port.snmp
6888 DEFAULT_DOC: SNMP disabled.
6891 The port number where Squid listens for SNMP requests. To enable
6892 SNMP support set this to a suitable port number. Port number
6893 3401 is often used for the Squid SNMP agent. By default it's
6894 set to "0" (disabled)
6902 LOC: Config.accessList.snmp
6904 DEFAULT_DOC: Deny, unless rules exist in squid.conf.
6907 Allowing or denying access to the SNMP port.
6909 All access to the agent is denied by default.
6912 snmp_access allow|deny [!]aclname ...
6914 This clause only supports fast acl types.
6915 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
6918 snmp_access allow snmppublic localhost
6919 snmp_access deny all
6922 NAME: snmp_incoming_address
6924 LOC: Config.Addrs.snmp_incoming
6926 DEFAULT_DOC: Accept SNMP packets from all machine interfaces.
6929 Just like 'udp_incoming_address', but for the SNMP port.
6931 snmp_incoming_address is used for the SNMP socket receiving
6932 messages from SNMP agents.
6934 The default snmp_incoming_address is to listen on all
6935 available network interfaces.
6938 NAME: snmp_outgoing_address
6940 LOC: Config.Addrs.snmp_outgoing
6942 DEFAULT_DOC: Use snmp_incoming_address or an address selected by the operating system.
6945 Just like 'udp_outgoing_address', but for the SNMP port.
6947 snmp_outgoing_address is used for SNMP packets returned to SNMP
6950 If snmp_outgoing_address is not set it will use the same socket
6951 as snmp_incoming_address. Only change this if you want to have
6952 SNMP replies sent using another address than where this Squid
6953 listens for SNMP queries.
6955 NOTE, snmp_incoming_address and snmp_outgoing_address can not have
6956 the same value since they both use the same port.
6961 -----------------------------------------------------------------------------
6964 NAME: icp_port udp_port
6967 DEFAULT_DOC: ICP disabled.
6968 LOC: Config.Port.icp
6970 The port number where Squid sends and receives ICP queries to
6971 and from neighbor caches. The standard UDP port for ICP is 3130.
6974 icp_port @DEFAULT_ICP_PORT@
6981 DEFAULT_DOC: HTCP disabled.
6982 LOC: Config.Port.htcp
6984 The port number where Squid sends and receives HTCP queries to
6985 and from neighbor caches. To turn it on you want to set it to
6992 NAME: log_icp_queries
6996 LOC: Config.onoff.log_udp
6998 If set, ICP queries are logged to access.log. You may wish
6999 do disable this if your ICP load is VERY high to speed things
7000 up or to simplify log analysis.
7003 NAME: udp_incoming_address
7005 LOC:Config.Addrs.udp_incoming
7007 DEFAULT_DOC: Accept packets from all machine interfaces.
7009 udp_incoming_address is used for UDP packets received from other
7012 The default behavior is to not bind to any specific address.
7014 Only change this if you want to have all UDP queries received on
7015 a specific interface/address.
7017 NOTE: udp_incoming_address is used by the ICP, HTCP, and DNS
7018 modules. Altering it will affect all of them in the same manner.
7020 see also; udp_outgoing_address
7022 NOTE, udp_incoming_address and udp_outgoing_address can not
7023 have the same value since they both use the same port.
7026 NAME: udp_outgoing_address
7028 LOC: Config.Addrs.udp_outgoing
7030 DEFAULT_DOC: Use udp_incoming_address or an address selected by the operating system.
7032 udp_outgoing_address is used for UDP packets sent out to other
7035 The default behavior is to not bind to any specific address.
7037 Instead it will use the same socket as udp_incoming_address.
7038 Only change this if you want to have UDP queries sent using another
7039 address than where this Squid listens for UDP queries from other
7042 NOTE: udp_outgoing_address is used by the ICP, HTCP, and DNS
7043 modules. Altering it will affect all of them in the same manner.
7045 see also; udp_incoming_address
7047 NOTE, udp_incoming_address and udp_outgoing_address can not
7048 have the same value since they both use the same port.
7055 LOC: Config.onoff.icp_hit_stale
7057 If you want to return ICP_HIT for stale cache objects, set this
7058 option to 'on'. If you have sibling relationships with caches
7059 in other administrative domains, this should be 'off'. If you only
7060 have sibling relationships with caches under your control,
7061 it is probably okay to set this to 'on'.
7062 If set to 'on', your siblings should use the option "allow-miss"
7063 on their cache_peer lines for connecting to you.
7066 NAME: minimum_direct_hops
7069 LOC: Config.minDirectHops
7071 If using the ICMP pinging stuff, do direct fetches for sites
7072 which are no more than this many hops away.
7075 NAME: minimum_direct_rtt
7079 LOC: Config.minDirectRtt
7081 If using the ICMP pinging stuff, do direct fetches for sites
7082 which are no more than this many rtt milliseconds away.
7088 LOC: Config.Netdb.low
7090 The low water mark for the ICMP measurement database.
7092 Note: high watermark controlled by netdb_high directive.
7094 These watermarks are counts, not percents. The defaults are
7095 (low) 900 and (high) 1000. When the high water mark is
7096 reached, database entries will be deleted until the low
7103 LOC: Config.Netdb.high
7105 The high water mark for the ICMP measurement database.
7107 Note: low watermark controlled by netdb_low directive.
7109 These watermarks are counts, not percents. The defaults are
7110 (low) 900 and (high) 1000. When the high water mark is
7111 reached, database entries will be deleted until the low
7115 NAME: netdb_ping_period
7117 LOC: Config.Netdb.period
7120 The minimum period for measuring a site. There will be at
7121 least this much delay between successive pings to the same
7122 network. The default is five minutes.
7129 LOC: Config.onoff.query_icmp
7131 If you want to ask your peers to include ICMP data in their ICP
7132 replies, enable this option.
7134 If your peer has configured Squid (during compilation) with
7135 '--enable-icmp' that peer will send ICMP pings to origin server
7136 sites of the URLs it receives. If you enable this option the
7137 ICP replies from that peer will include the ICMP data (if available).
7138 Then, when choosing a parent cache, Squid will choose the parent with
7139 the minimal RTT to the origin server. When this happens, the
7140 hierarchy field of the access.log will be
7141 "CLOSEST_PARENT_MISS". This option is off by default.
7144 NAME: test_reachability
7148 LOC: Config.onoff.test_reachability
7150 When this is 'on', ICP MISS replies will be ICP_MISS_NOFETCH
7151 instead of ICP_MISS if the target host is NOT in the ICMP
7152 database, or has a zero RTT.
7155 NAME: icp_query_timeout
7158 DEFAULT_DOC: Dynamic detection.
7160 LOC: Config.Timeout.icp_query
7162 Normally Squid will automatically determine an optimal ICP
7163 query timeout value based on the round-trip-time of recent ICP
7164 queries. If you want to override the value determined by
7165 Squid, set this 'icp_query_timeout' to a non-zero value. This
7166 value is specified in MILLISECONDS, so, to use a 2-second
7167 timeout (the old default), you would write:
7169 icp_query_timeout 2000
7172 NAME: maximum_icp_query_timeout
7176 LOC: Config.Timeout.icp_query_max
7178 Normally the ICP query timeout is determined dynamically. But
7179 sometimes it can lead to very large values (say 5 seconds).
7180 Use this option to put an upper limit on the dynamic timeout
7181 value. Do NOT use this option to always use a fixed (instead
7182 of a dynamic) timeout value. To set a fixed timeout see the
7183 'icp_query_timeout' directive.
7186 NAME: minimum_icp_query_timeout
7190 LOC: Config.Timeout.icp_query_min
7192 Normally the ICP query timeout is determined dynamically. But
7193 sometimes it can lead to very small timeouts, even lower than
7194 the normal latency variance on your link due to traffic.
7195 Use this option to put an lower limit on the dynamic timeout
7196 value. Do NOT use this option to always use a fixed (instead
7197 of a dynamic) timeout value. To set a fixed timeout see the
7198 'icp_query_timeout' directive.
7201 NAME: background_ping_rate
7205 LOC: Config.backgroundPingRate
7207 Controls how often the ICP pings are sent to siblings that
7208 have background-ping set.
7212 MULTICAST ICP OPTIONS
7213 -----------------------------------------------------------------------------
7218 LOC: Config.mcast_group_list
7221 This tag specifies a list of multicast groups which your server
7222 should join to receive multicasted ICP queries.
7224 NOTE! Be very careful what you put here! Be sure you
7225 understand the difference between an ICP _query_ and an ICP
7226 _reply_. This option is to be set only if you want to RECEIVE
7227 multicast queries. Do NOT set this option to SEND multicast
7228 ICP (use cache_peer for that). ICP replies are always sent via
7229 unicast, so this option does not affect whether or not you will
7230 receive replies from multicast group members.
7232 You must be very careful to NOT use a multicast address which
7233 is already in use by another group of caches.
7235 If you are unsure about multicast, please read the Multicast
7236 chapter in the Squid FAQ (http://www.squid-cache.org/FAQ/).
7238 Usage: mcast_groups 239.128.16.128 224.0.1.20
7240 By default, Squid doesn't listen on any multicast groups.
7243 NAME: mcast_miss_addr
7244 IFDEF: MULTICAST_MISS_STREAM
7246 LOC: Config.mcast_miss.addr
7248 DEFAULT_DOC: disabled.
7250 If you enable this option, every "cache miss" URL will
7251 be sent out on the specified multicast address.
7253 Do not enable this option unless you are are absolutely
7254 certain you understand what you are doing.
7257 NAME: mcast_miss_ttl
7258 IFDEF: MULTICAST_MISS_STREAM
7260 LOC: Config.mcast_miss.ttl
7263 This is the time-to-live value for packets multicasted
7264 when multicasting off cache miss URLs is enabled. By
7265 default this is set to 'site scope', i.e. 16.
7268 NAME: mcast_miss_port
7269 IFDEF: MULTICAST_MISS_STREAM
7271 LOC: Config.mcast_miss.port
7274 This is the port number to be used in conjunction with
7278 NAME: mcast_miss_encode_key
7279 IFDEF: MULTICAST_MISS_STREAM
7281 LOC: Config.mcast_miss.encode_key
7282 DEFAULT: XXXXXXXXXXXXXXXX
7284 The URLs that are sent in the multicast miss stream are
7285 encrypted. This is the encryption key.
7288 NAME: mcast_icp_query_timeout
7292 LOC: Config.Timeout.mcast_icp_query
7294 For multicast peers, Squid regularly sends out ICP "probes" to
7295 count how many other peers are listening on the given multicast
7296 address. This value specifies how long Squid should wait to
7297 count all the replies. The default is 2000 msec, or 2
7302 INTERNAL ICON OPTIONS
7303 -----------------------------------------------------------------------------
7306 NAME: icon_directory
7308 LOC: Config.icons.directory
7309 DEFAULT: @DEFAULT_ICON_DIR@
7311 Where the icons are stored. These are normally kept in
7315 NAME: global_internal_static
7317 LOC: Config.onoff.global_internal_static
7320 This directive controls is Squid should intercept all requests for
7321 /squid-internal-static/ no matter which host the URL is requesting
7322 (default on setting), or if nothing special should be done for
7323 such URLs (off setting). The purpose of this directive is to make
7324 icons etc work better in complex cache hierarchies where it may
7325 not always be possible for all corners in the cache mesh to reach
7326 the server generating a directory listing.
7329 NAME: short_icon_urls
7331 LOC: Config.icons.use_short_names
7334 If this is enabled Squid will use short URLs for icons.
7335 If disabled it will revert to the old behavior of including
7336 it's own name and port in the URL.
7338 If you run a complex cache hierarchy with a mix of Squid and
7339 other proxies you may need to disable this directive.
7344 -----------------------------------------------------------------------------
7347 NAME: error_directory
7349 LOC: Config.errorDirectory
7351 DEFAULT_DOC: Send error pages in the clients preferred language
7353 If you wish to create your own versions of the default
7354 error files to customize them to suit your company copy
7355 the error/template files to another directory and point
7358 WARNING: This option will disable multi-language support
7359 on error pages if used.
7361 The squid developers are interested in making squid available in
7362 a wide variety of languages. If you are making translations for a
7363 language that Squid does not currently provide please consider
7364 contributing your translation back to the project.
7365 http://wiki.squid-cache.org/Translations
7367 The squid developers working on translations are happy to supply drop-in
7368 translated error files in exchange for any new language contributions.
7371 NAME: error_default_language
7372 IFDEF: USE_ERR_LOCALES
7374 LOC: Config.errorDefaultLanguage
7376 DEFAULT_DOC: Generate English language pages.
7378 Set the default language which squid will send error pages in
7379 if no existing translation matches the clients language
7382 If unset (default) generic English will be used.
7384 The squid developers are interested in making squid available in
7385 a wide variety of languages. If you are interested in making
7386 translations for any language see the squid wiki for details.
7387 http://wiki.squid-cache.org/Translations
7390 NAME: error_log_languages
7391 IFDEF: USE_ERR_LOCALES
7393 LOC: Config.errorLogMissingLanguages
7396 Log to cache.log what languages users are attempting to
7397 auto-negotiate for translations.
7399 Successful negotiations are not logged. Only failures
7400 have meaning to indicate that Squid may need an upgrade
7401 of its error page translations.
7404 NAME: err_page_stylesheet
7406 LOC: Config.errorStylesheet
7407 DEFAULT: @DEFAULT_CONFIG_DIR@/errorpage.css
7409 CSS Stylesheet to pattern the display of Squid default error pages.
7411 For information on CSS see http://www.w3.org/Style/CSS/
7416 LOC: Config.errHtmlText
7419 HTML text to include in error messages. Make this a "mailto"
7420 URL to your admin address, or maybe just a link to your
7421 organizations Web page.
7423 To include this in your error messages, you must rewrite
7424 the error template files (found in the "errors" directory).
7425 Wherever you want the 'err_html_text' line to appear,
7426 insert a %L tag in the error template file.
7429 NAME: email_err_data
7432 LOC: Config.onoff.emailErrData
7435 If enabled, information about the occurred error will be
7436 included in the mailto links of the ERR pages (if %W is set)
7437 so that the email body contains the data.
7438 Syntax is <A HREF="mailto:%w%W">%w</A>
7443 LOC: Config.denyInfoList
7446 Usage: deny_info err_page_name acl
7447 or deny_info http://... acl
7448 or deny_info TCP_RESET acl
7450 This can be used to return a ERR_ page for requests which
7451 do not pass the 'http_access' rules. Squid remembers the last
7452 acl it evaluated in http_access, and if a 'deny_info' line exists
7453 for that ACL Squid returns a corresponding error page.
7455 The acl is typically the last acl on the http_access deny line which
7456 denied access. The exceptions to this rule are:
7457 - When Squid needs to request authentication credentials. It's then
7458 the first authentication related acl encountered
7459 - When none of the http_access lines matches. It's then the last
7460 acl processed on the last http_access line.
7461 - When the decision to deny access was made by an adaptation service,
7462 the acl name is the corresponding eCAP or ICAP service_name.
7464 NP: If providing your own custom error pages with error_directory
7465 you may also specify them by your custom file name:
7466 Example: deny_info ERR_CUSTOM_ACCESS_DENIED bad_guys
7468 By defaut Squid will send "403 Forbidden". A different 4xx or 5xx
7469 may be specified by prefixing the file name with the code and a colon.
7470 e.g. 404:ERR_CUSTOM_ACCESS_DENIED
7472 Alternatively you can tell Squid to reset the TCP connection
7473 by specifying TCP_RESET.
7475 Or you can specify an error URL or URL pattern. The browsers will
7476 get redirected to the specified URL after formatting tags have
7477 been replaced. Redirect will be done with 302 or 307 according to
7478 HTTP/1.1 specs. A different 3xx code may be specified by prefixing
7479 the URL. e.g. 303:http://example.com/
7482 %a - username (if available. Password NOT included)
7485 %E - Error description
7487 %H - Request domain name
7488 %i - Client IP Address
7490 %o - Message result from external ACL helper
7491 %p - Request Port number
7492 %P - Request Protocol name
7493 %R - Request URL path
7494 %T - Timestamp in RFC 1123 format
7495 %U - Full canonical URL from client
7496 (HTTPS URLs terminate with *)
7497 %u - Full canonical URL from client
7498 %w - Admin email from squid.conf
7500 %% - Literal percent (%) code
7505 OPTIONS INFLUENCING REQUEST FORWARDING
7506 -----------------------------------------------------------------------------
7509 NAME: nonhierarchical_direct
7511 LOC: Config.onoff.nonhierarchical_direct
7514 By default, Squid will send any non-hierarchical requests
7515 (not cacheable request type) direct to origin servers.
7517 When this is set to "off", Squid will prefer to send these
7518 requests to parents.
7520 Note that in most configurations, by turning this off you will only
7521 add latency to these request without any improvement in global hit
7524 This option only sets a preference. If the parent is unavailable a
7525 direct connection to the origin server may still be attempted. To
7526 completely prevent direct connections use never_direct.
7531 LOC: Config.onoff.prefer_direct
7534 Normally Squid tries to use parents for most requests. If you for some
7535 reason like it to first try going direct and only use a parent if
7536 going direct fails set this to on.
7538 By combining nonhierarchical_direct off and prefer_direct on you
7539 can set up Squid to use a parent as a backup path if going direct
7542 Note: If you want Squid to use parents for all requests see
7543 the never_direct directive. prefer_direct only modifies how Squid
7544 acts on cacheable requests.
7547 NAME: cache_miss_revalidate
7551 LOC: Config.onoff.cache_miss_revalidate
7553 RFC 7232 defines a conditional request mechanism to prevent
7554 response objects being unnecessarily transferred over the network.
7555 If that mechanism is used by the client and a cache MISS occurs
7556 it can prevent new cache entries being created.
7558 This option determines whether Squid on cache MISS will pass the
7559 client revalidation request to the server or tries to fetch new
7560 content for caching. It can be useful while the cache is mostly
7561 empty to more quickly have the cache populated by generating
7562 non-conditional GETs.
7564 When set to 'on' (default), Squid will pass all client If-* headers
7565 to the server. This permits server responses without a cacheable
7566 payload to be delivered and on MISS no new cache entry is created.
7568 When set to 'off' and if the request is cacheable, Squid will
7569 remove the clients If-Modified-Since and If-None-Match headers from
7570 the request sent to the server. This requests a 200 status response
7571 from the server to create a new cache entry with.
7576 LOC: Config.accessList.AlwaysDirect
7578 DEFAULT_DOC: Prevent any cache_peer being used for this request.
7580 Usage: always_direct allow|deny [!]aclname ...
7582 Here you can use ACL elements to specify requests which should
7583 ALWAYS be forwarded by Squid to the origin servers without using
7584 any peers. For example, to always directly forward requests for
7585 local servers ignoring any parents or siblings you may have use
7588 acl local-servers dstdomain my.domain.net
7589 always_direct allow local-servers
7591 To always forward FTP requests directly, use
7594 always_direct allow FTP
7596 NOTE: There is a similar, but opposite option named
7597 'never_direct'. You need to be aware that "always_direct deny
7598 foo" is NOT the same thing as "never_direct allow foo". You
7599 may need to use a deny rule to exclude a more-specific case of
7600 some other rule. Example:
7602 acl local-external dstdomain external.foo.net
7603 acl local-servers dstdomain .foo.net
7604 always_direct deny local-external
7605 always_direct allow local-servers
7607 NOTE: If your goal is to make the client forward the request
7608 directly to the origin server bypassing Squid then this needs
7609 to be done in the client configuration. Squid configuration
7610 can only tell Squid how Squid should fetch the object.
7612 NOTE: This directive is not related to caching. The replies
7613 is cached as usual even if you use always_direct. To not cache
7614 the replies see the 'cache' directive.
7616 This clause supports both fast and slow acl types.
7617 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
7622 LOC: Config.accessList.NeverDirect
7624 DEFAULT_DOC: Allow DNS results to be used for this request.
7626 Usage: never_direct allow|deny [!]aclname ...
7628 never_direct is the opposite of always_direct. Please read
7629 the description for always_direct if you have not already.
7631 With 'never_direct' you can use ACL elements to specify
7632 requests which should NEVER be forwarded directly to origin
7633 servers. For example, to force the use of a proxy for all
7634 requests, except those in your local domain use something like:
7636 acl local-servers dstdomain .foo.net
7637 never_direct deny local-servers
7638 never_direct allow all
7640 or if Squid is inside a firewall and there are local intranet
7641 servers inside the firewall use something like:
7643 acl local-intranet dstdomain .foo.net
7644 acl local-external dstdomain external.foo.net
7645 always_direct deny local-external
7646 always_direct allow local-intranet
7647 never_direct allow all
7649 This clause supports both fast and slow acl types.
7650 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
7654 ADVANCED NETWORKING OPTIONS
7655 -----------------------------------------------------------------------------
7658 NAME: incoming_udp_average incoming_icp_average
7661 LOC: Config.comm_incoming.udp.average
7663 Heavy voodoo here. I can't even believe you are reading this.
7664 Are you crazy? Don't even think about adjusting these unless
7665 you understand the algorithms in comm_select.c first!
7668 NAME: incoming_tcp_average incoming_http_average
7671 LOC: Config.comm_incoming.tcp.average
7673 Heavy voodoo here. I can't even believe you are reading this.
7674 Are you crazy? Don't even think about adjusting these unless
7675 you understand the algorithms in comm_select.c first!
7678 NAME: incoming_dns_average
7681 LOC: Config.comm_incoming.dns.average
7683 Heavy voodoo here. I can't even believe you are reading this.
7684 Are you crazy? Don't even think about adjusting these unless
7685 you understand the algorithms in comm_select.c first!
7688 NAME: min_udp_poll_cnt min_icp_poll_cnt
7691 LOC: Config.comm_incoming.udp.min_poll
7693 Heavy voodoo here. I can't even believe you are reading this.
7694 Are you crazy? Don't even think about adjusting these unless
7695 you understand the algorithms in comm_select.c first!
7698 NAME: min_dns_poll_cnt
7701 LOC: Config.comm_incoming.dns.min_poll
7703 Heavy voodoo here. I can't even believe you are reading this.
7704 Are you crazy? Don't even think about adjusting these unless
7705 you understand the algorithms in comm_select.c first!
7708 NAME: min_tcp_poll_cnt min_http_poll_cnt
7711 LOC: Config.comm_incoming.tcp.min_poll
7713 Heavy voodoo here. I can't even believe you are reading this.
7714 Are you crazy? Don't even think about adjusting these unless
7715 you understand the algorithms in comm_select.c first!
7721 LOC: Config.accept_filter
7725 The name of an accept(2) filter to install on Squid's
7726 listen socket(s). This feature is perhaps specific to
7727 FreeBSD and requires support in the kernel.
7729 The 'httpready' filter delays delivering new connections
7730 to Squid until a full HTTP request has been received.
7731 See the accf_http(9) man page for details.
7733 The 'dataready' filter delays delivering new connections
7734 to Squid until there is some data to process.
7735 See the accf_dataready(9) man page for details.
7739 The 'data' filter delays delivering of new connections
7740 to Squid until there is some data to process by TCP_ACCEPT_DEFER.
7741 You may optionally specify a number of seconds to wait by
7742 'data=N' where N is the number of seconds. Defaults to 30
7743 if not specified. See the tcp(7) man page for details.
7746 accept_filter httpready
7751 NAME: client_ip_max_connections
7753 LOC: Config.client_ip_max_connections
7755 DEFAULT_DOC: No limit.
7757 Set an absolute limit on the number of connections a single
7758 client IP can use. Any more than this and Squid will begin to drop
7759 new connections from the client until it closes some links.
7761 Note that this is a global limit. It affects all HTTP, HTCP, Gopher and FTP
7762 connections from the client. For finer control use the ACL access controls.
7764 Requires client_db to be enabled (the default).
7766 WARNING: This may noticably slow down traffic received via external proxies
7767 or NAT devices and cause them to rebound error messages back to their clients.
7770 NAME: tcp_recv_bufsize
7774 DEFAULT_DOC: Use operating system TCP defaults.
7775 LOC: Config.tcpRcvBufsz
7777 Size of receive buffer to set for TCP sockets. Probably just
7778 as easy to change your kernel's default.
7779 Omit from squid.conf to use the default buffer size.
7784 -----------------------------------------------------------------------------
7791 LOC: Adaptation::Icap::TheConfig.onoff
7794 If you want to enable the ICAP module support, set this to on.
7797 NAME: icap_connect_timeout
7800 LOC: Adaptation::Icap::TheConfig.connect_timeout_raw
7803 This parameter specifies how long to wait for the TCP connect to
7804 the requested ICAP server to complete before giving up and either
7805 terminating the HTTP transaction or bypassing the failure.
7807 The default for optional services is peer_connect_timeout.
7808 The default for essential services is connect_timeout.
7809 If this option is explicitly set, its value applies to all services.
7812 NAME: icap_io_timeout
7816 DEFAULT_DOC: Use read_timeout.
7817 LOC: Adaptation::Icap::TheConfig.io_timeout_raw
7820 This parameter specifies how long to wait for an I/O activity on
7821 an established, active ICAP connection before giving up and
7822 either terminating the HTTP transaction or bypassing the
7826 NAME: icap_service_failure_limit
7827 COMMENT: limit [in memory-depth time-units]
7828 TYPE: icap_service_failure_limit
7830 LOC: Adaptation::Icap::TheConfig
7833 The limit specifies the number of failures that Squid tolerates
7834 when establishing a new TCP connection with an ICAP service. If
7835 the number of failures exceeds the limit, the ICAP service is
7836 not used for new ICAP requests until it is time to refresh its
7839 A negative value disables the limit. Without the limit, an ICAP
7840 service will not be considered down due to connectivity failures
7841 between ICAP OPTIONS requests.
7843 Squid forgets ICAP service failures older than the specified
7844 value of memory-depth. The memory fading algorithm
7845 is approximate because Squid does not remember individual
7846 errors but groups them instead, splitting the option
7847 value into ten time slots of equal length.
7849 When memory-depth is 0 and by default this option has no
7850 effect on service failure expiration.
7852 Squid always forgets failures when updating service settings
7853 using an ICAP OPTIONS transaction, regardless of this option
7857 # suspend service usage after 10 failures in 5 seconds:
7858 icap_service_failure_limit 10 in 5 seconds
7861 NAME: icap_service_revival_delay
7864 LOC: Adaptation::Icap::TheConfig.service_revival_delay
7867 The delay specifies the number of seconds to wait after an ICAP
7868 OPTIONS request failure before requesting the options again. The
7869 failed ICAP service is considered "down" until fresh OPTIONS are
7872 The actual delay cannot be smaller than the hardcoded minimum
7873 delay of 30 seconds.
7876 NAME: icap_preview_enable
7880 LOC: Adaptation::Icap::TheConfig.preview_enable
7883 The ICAP Preview feature allows the ICAP server to handle the
7884 HTTP message by looking only at the beginning of the message body
7885 or even without receiving the body at all. In some environments,
7886 previews greatly speedup ICAP processing.
7888 During an ICAP OPTIONS transaction, the server may tell Squid what
7889 HTTP messages should be previewed and how big the preview should be.
7890 Squid will not use Preview if the server did not request one.
7892 To disable ICAP Preview for all ICAP services, regardless of
7893 individual ICAP server OPTIONS responses, set this option to "off".
7895 icap_preview_enable off
7898 NAME: icap_preview_size
7901 LOC: Adaptation::Icap::TheConfig.preview_size
7903 DEFAULT_DOC: No preview sent.
7905 The default size of preview data to be sent to the ICAP server.
7906 This value might be overwritten on a per server basis by OPTIONS requests.
7909 NAME: icap_206_enable
7913 LOC: Adaptation::Icap::TheConfig.allow206_enable
7916 206 (Partial Content) responses is an ICAP extension that allows the
7917 ICAP agents to optionally combine adapted and original HTTP message
7918 content. The decision to combine is postponed until the end of the
7919 ICAP response. Squid supports Partial Content extension by default.
7921 Activation of the Partial Content extension is negotiated with each
7922 ICAP service during OPTIONS exchange. Most ICAP servers should handle
7923 negotation correctly even if they do not support the extension, but
7924 some might fail. To disable Partial Content support for all ICAP
7925 services and to avoid any negotiation, set this option to "off".
7931 NAME: icap_default_options_ttl
7934 LOC: Adaptation::Icap::TheConfig.default_options_ttl
7937 The default TTL value for ICAP OPTIONS responses that don't have
7938 an Options-TTL header.
7941 NAME: icap_persistent_connections
7945 LOC: Adaptation::Icap::TheConfig.reuse_connections
7948 Whether or not Squid should use persistent connections to
7952 NAME: adaptation_send_client_ip icap_send_client_ip
7954 IFDEF: USE_ADAPTATION
7956 LOC: Adaptation::Config::send_client_ip
7959 If enabled, Squid shares HTTP client IP information with adaptation
7960 services. For ICAP, Squid adds the X-Client-IP header to ICAP requests.
7961 For eCAP, Squid sets the libecap::metaClientIp transaction option.
7963 See also: adaptation_uses_indirect_client
7966 NAME: adaptation_send_username icap_send_client_username
7968 IFDEF: USE_ADAPTATION
7970 LOC: Adaptation::Config::send_username
7973 This sends authenticated HTTP client username (if available) to
7974 the adaptation service.
7976 For ICAP, the username value is encoded based on the
7977 icap_client_username_encode option and is sent using the header
7978 specified by the icap_client_username_header option.
7981 NAME: icap_client_username_header
7984 LOC: Adaptation::Icap::TheConfig.client_username_header
7985 DEFAULT: X-Client-Username
7987 ICAP request header name to use for adaptation_send_username.
7990 NAME: icap_client_username_encode
7994 LOC: Adaptation::Icap::TheConfig.client_username_encode
7997 Whether to base64 encode the authenticated client username.
8001 TYPE: icap_service_type
8003 LOC: Adaptation::Icap::TheConfig
8006 Defines a single ICAP service using the following format:
8008 icap_service id vectoring_point uri [option ...]
8011 an opaque identifier or name which is used to direct traffic to
8012 this specific service. Must be unique among all adaptation
8013 services in squid.conf.
8015 vectoring_point: reqmod_precache|reqmod_postcache|respmod_precache|respmod_postcache
8016 This specifies at which point of transaction processing the
8017 ICAP service should be activated. *_postcache vectoring points
8018 are not yet supported.
8020 uri: icap://servername:port/servicepath
8021 ICAP server and service location.
8023 ICAP does not allow a single service to handle both REQMOD and RESPMOD
8024 transactions. Squid does not enforce that requirement. You can specify
8025 services with the same service_url and different vectoring_points. You
8026 can even specify multiple identical services as long as their
8027 service_names differ.
8029 To activate a service, use the adaptation_access directive. To group
8030 services, use adaptation_service_chain and adaptation_service_set.
8032 Service options are separated by white space. ICAP services support
8033 the following name=value options:
8036 If set to 'on' or '1', the ICAP service is treated as
8037 optional. If the service cannot be reached or malfunctions,
8038 Squid will try to ignore any errors and process the message as
8039 if the service was not enabled. No all ICAP errors can be
8040 bypassed. If set to 0, the ICAP service is treated as
8041 essential and all ICAP errors will result in an error page
8042 returned to the HTTP client.
8044 Bypass is off by default: services are treated as essential.
8047 If set to 'on' or '1', the ICAP service is allowed to
8048 dynamically change the current message adaptation plan by
8049 returning a chain of services to be used next. The services
8050 are specified using the X-Next-Services ICAP response header
8051 value, formatted as a comma-separated list of service names.
8052 Each named service should be configured in squid.conf. Other
8053 services are ignored. An empty X-Next-Services value results
8054 in an empty plan which ends the current adaptation.
8056 Dynamic adaptation plan may cross or cover multiple supported
8057 vectoring points in their natural processing order.
8059 Routing is not allowed by default: the ICAP X-Next-Services
8060 response header is ignored.
8063 Only has effect on split-stack systems. The default on those systems
8064 is to use IPv4-only connections. When set to 'on' this option will
8065 make Squid use IPv6-only connections to contact this ICAP service.
8067 on-overload=block|bypass|wait|force
8068 If the service Max-Connections limit has been reached, do
8069 one of the following for each new ICAP transaction:
8070 * block: send an HTTP error response to the client
8071 * bypass: ignore the "over-connected" ICAP service
8072 * wait: wait (in a FIFO queue) for an ICAP connection slot
8073 * force: proceed, ignoring the Max-Connections limit
8075 In SMP mode with N workers, each worker assumes the service
8076 connection limit is Max-Connections/N, even though not all
8077 workers may use a given service.
8079 The default value is "bypass" if service is bypassable,
8080 otherwise it is set to "wait".
8084 Use the given number as the Max-Connections limit, regardless
8085 of the Max-Connections value given by the service, if any.
8087 Older icap_service format without optional named parameters is
8088 deprecated but supported for backward compatibility.
8091 icap_service svcBlocker reqmod_precache icap://icap1.mydomain.net:1344/reqmod bypass=0
8092 icap_service svcLogger reqmod_precache icap://icap2.mydomain.net:1344/respmod routing=on
8096 TYPE: icap_class_type
8101 This deprecated option was documented to define an ICAP service
8102 chain, even though it actually defined a set of similar, redundant
8103 services, and the chains were not supported.
8105 To define a set of redundant services, please use the
8106 adaptation_service_set directive. For service chains, use
8107 adaptation_service_chain.
8111 TYPE: icap_access_type
8116 This option is deprecated. Please use adaptation_access, which
8117 has the same ICAP functionality, but comes with better
8118 documentation, and eCAP support.
8123 -----------------------------------------------------------------------------
8130 LOC: Adaptation::Ecap::TheConfig.onoff
8133 Controls whether eCAP support is enabled.
8137 TYPE: ecap_service_type
8139 LOC: Adaptation::Ecap::TheConfig
8142 Defines a single eCAP service
8144 ecap_service id vectoring_point uri [option ...]
8147 an opaque identifier or name which is used to direct traffic to
8148 this specific service. Must be unique among all adaptation
8149 services in squid.conf.
8151 vectoring_point: reqmod_precache|reqmod_postcache|respmod_precache|respmod_postcache
8152 This specifies at which point of transaction processing the
8153 eCAP service should be activated. *_postcache vectoring points
8154 are not yet supported.
8156 uri: ecap://vendor/service_name?custom&cgi=style¶meters=optional
8157 Squid uses the eCAP service URI to match this configuration
8158 line with one of the dynamically loaded services. Each loaded
8159 eCAP service must have a unique URI. Obtain the right URI from
8160 the service provider.
8162 To activate a service, use the adaptation_access directive. To group
8163 services, use adaptation_service_chain and adaptation_service_set.
8165 Service options are separated by white space. eCAP services support
8166 the following name=value options:
8169 If set to 'on' or '1', the eCAP service is treated as optional.
8170 If the service cannot be reached or malfunctions, Squid will try
8171 to ignore any errors and process the message as if the service
8172 was not enabled. No all eCAP errors can be bypassed.
8173 If set to 'off' or '0', the eCAP service is treated as essential
8174 and all eCAP errors will result in an error page returned to the
8177 Bypass is off by default: services are treated as essential.
8180 If set to 'on' or '1', the eCAP service is allowed to
8181 dynamically change the current message adaptation plan by
8182 returning a chain of services to be used next.
8184 Dynamic adaptation plan may cross or cover multiple supported
8185 vectoring points in their natural processing order.
8187 Routing is not allowed by default.
8189 Older ecap_service format without optional named parameters is
8190 deprecated but supported for backward compatibility.
8194 ecap_service s1 reqmod_precache ecap://filters.R.us/leakDetector?on_error=block bypass=off
8195 ecap_service s2 respmod_precache ecap://filters.R.us/virusFilter config=/etc/vf.cfg bypass=on
8198 NAME: loadable_modules
8200 IFDEF: USE_LOADABLE_MODULES
8201 LOC: Config.loadable_module_names
8204 Instructs Squid to load the specified dynamic module(s) or activate
8205 preloaded module(s).
8207 loadable_modules @DEFAULT_PREFIX@/lib/MinimalAdapter.so
8211 MESSAGE ADAPTATION OPTIONS
8212 -----------------------------------------------------------------------------
8215 NAME: adaptation_service_set
8216 TYPE: adaptation_service_set_type
8217 IFDEF: USE_ADAPTATION
8222 Configures an ordered set of similar, redundant services. This is
8223 useful when hot standby or backup adaptation servers are available.
8225 adaptation_service_set set_name service_name1 service_name2 ...
8227 The named services are used in the set declaration order. The first
8228 applicable adaptation service from the set is used first. The next
8229 applicable service is tried if and only if the transaction with the
8230 previous service fails and the message waiting to be adapted is still
8233 When adaptation starts, broken services are ignored as if they were
8234 not a part of the set. A broken service is a down optional service.
8236 The services in a set must be attached to the same vectoring point
8237 (e.g., pre-cache) and use the same adaptation method (e.g., REQMOD).
8239 If all services in a set are optional then adaptation failures are
8240 bypassable. If all services in the set are essential, then a
8241 transaction failure with one service may still be retried using
8242 another service from the set, but when all services fail, the master
8243 transaction fails as well.
8245 A set may contain a mix of optional and essential services, but that
8246 is likely to lead to surprising results because broken services become
8247 ignored (see above), making previously bypassable failures fatal.
8248 Technically, it is the bypassability of the last failed service that
8251 See also: adaptation_access adaptation_service_chain
8254 adaptation_service_set svcBlocker urlFilterPrimary urlFilterBackup
8255 adaptation service_set svcLogger loggerLocal loggerRemote
8258 NAME: adaptation_service_chain
8259 TYPE: adaptation_service_chain_type
8260 IFDEF: USE_ADAPTATION
8265 Configures a list of complementary services that will be applied
8266 one-by-one, forming an adaptation chain or pipeline. This is useful
8267 when Squid must perform different adaptations on the same message.
8269 adaptation_service_chain chain_name service_name1 svc_name2 ...
8271 The named services are used in the chain declaration order. The first
8272 applicable adaptation service from the chain is used first. The next
8273 applicable service is applied to the successful adaptation results of
8274 the previous service in the chain.
8276 When adaptation starts, broken services are ignored as if they were
8277 not a part of the chain. A broken service is a down optional service.
8279 Request satisfaction terminates the adaptation chain because Squid
8280 does not currently allow declaration of RESPMOD services at the
8281 "reqmod_precache" vectoring point (see icap_service or ecap_service).
8283 The services in a chain must be attached to the same vectoring point
8284 (e.g., pre-cache) and use the same adaptation method (e.g., REQMOD).
8286 A chain may contain a mix of optional and essential services. If an
8287 essential adaptation fails (or the failure cannot be bypassed for
8288 other reasons), the master transaction fails. Otherwise, the failure
8289 is bypassed as if the failed adaptation service was not in the chain.
8291 See also: adaptation_access adaptation_service_set
8294 adaptation_service_chain svcRequest requestLogger urlFilter leakDetector
8297 NAME: adaptation_access
8298 TYPE: adaptation_access_type
8299 IFDEF: USE_ADAPTATION
8302 DEFAULT_DOC: Allow, unless rules exist in squid.conf.
8304 Sends an HTTP transaction to an ICAP or eCAP adaptation service.
8306 adaptation_access service_name allow|deny [!]aclname...
8307 adaptation_access set_name allow|deny [!]aclname...
8309 At each supported vectoring point, the adaptation_access
8310 statements are processed in the order they appear in this
8311 configuration file. Statements pointing to the following services
8312 are ignored (i.e., skipped without checking their ACL):
8314 - services serving different vectoring points
8315 - "broken-but-bypassable" services
8316 - "up" services configured to ignore such transactions
8317 (e.g., based on the ICAP Transfer-Ignore header).
8319 When a set_name is used, all services in the set are checked
8320 using the same rules, to find the first applicable one. See
8321 adaptation_service_set for details.
8323 If an access list is checked and there is a match, the
8324 processing stops: For an "allow" rule, the corresponding
8325 adaptation service is used for the transaction. For a "deny"
8326 rule, no adaptation service is activated.
8328 It is currently not possible to apply more than one adaptation
8329 service at the same vectoring point to the same HTTP transaction.
8331 See also: icap_service and ecap_service
8334 adaptation_access service_1 allow all
8337 NAME: adaptation_service_iteration_limit
8339 IFDEF: USE_ADAPTATION
8340 LOC: Adaptation::Config::service_iteration_limit
8343 Limits the number of iterations allowed when applying adaptation
8344 services to a message. If your longest adaptation set or chain
8345 may have more than 16 services, increase the limit beyond its
8346 default value of 16. If detecting infinite iteration loops sooner
8347 is critical, make the iteration limit match the actual number
8348 of services in your longest adaptation set or chain.
8350 Infinite adaptation loops are most likely with routing services.
8352 See also: icap_service routing=1
8355 NAME: adaptation_masterx_shared_names
8357 IFDEF: USE_ADAPTATION
8358 LOC: Adaptation::Config::masterx_shared_name
8361 For each master transaction (i.e., the HTTP request and response
8362 sequence, including all related ICAP and eCAP exchanges), Squid
8363 maintains a table of metadata. The table entries are (name, value)
8364 pairs shared among eCAP and ICAP exchanges. The table is destroyed
8365 with the master transaction.
8367 This option specifies the table entry names that Squid must accept
8368 from and forward to the adaptation transactions.
8370 An ICAP REQMOD or RESPMOD transaction may set an entry in the
8371 shared table by returning an ICAP header field with a name
8372 specified in adaptation_masterx_shared_names.
8374 An eCAP REQMOD or RESPMOD transaction may set an entry in the
8375 shared table by implementing the libecap::visitEachOption() API
8376 to provide an option with a name specified in
8377 adaptation_masterx_shared_names.
8379 Squid will store and forward the set entry to subsequent adaptation
8380 transactions within the same master transaction scope.
8382 Only one shared entry name is supported at this time.
8385 # share authentication information among ICAP services
8386 adaptation_masterx_shared_names X-Subscriber-ID
8389 NAME: adaptation_meta
8391 IFDEF: USE_ADAPTATION
8392 LOC: Adaptation::Config::metaHeaders
8395 This option allows Squid administrator to add custom ICAP request
8396 headers or eCAP options to Squid ICAP requests or eCAP transactions.
8397 Use it to pass custom authentication tokens and other
8398 transaction-state related meta information to an ICAP/eCAP service.
8400 The addition of a meta header is ACL-driven:
8401 adaptation_meta name value [!]aclname ...
8403 Processing for a given header name stops after the first ACL list match.
8404 Thus, it is impossible to add two headers with the same name. If no ACL
8405 lists match for a given header name, no such header is added. For
8408 # do not debug transactions except for those that need debugging
8409 adaptation_meta X-Debug 1 needs_debugging
8411 # log all transactions except for those that must remain secret
8412 adaptation_meta X-Log 1 !keep_secret
8414 # mark transactions from users in the "G 1" group
8415 adaptation_meta X-Authenticated-Groups "G 1" authed_as_G1
8417 The "value" parameter may be a regular squid.conf token or a "double
8418 quoted string". Within the quoted string, use backslash (\) to escape
8419 any character, which is currently only useful for escaping backslashes
8420 and double quotes. For example,
8421 "this string has one backslash (\\) and two \"quotes\""
8423 Used adaptation_meta header values may be logged via %note
8424 logformat code. If multiple adaptation_meta headers with the same name
8425 are used during master transaction lifetime, the header values are
8426 logged in the order they were used and duplicate values are ignored
8427 (only the first repeated value will be logged).
8433 LOC: Adaptation::Icap::TheConfig.repeat
8434 DEFAULT_IF_NONE: deny all
8436 This ACL determines which retriable ICAP transactions are
8437 retried. Transactions that received a complete ICAP response
8438 and did not have to consume or produce HTTP bodies to receive
8439 that response are usually retriable.
8441 icap_retry allow|deny [!]aclname ...
8443 Squid automatically retries some ICAP I/O timeouts and errors
8444 due to persistent connection race conditions.
8446 See also: icap_retry_limit
8449 NAME: icap_retry_limit
8452 LOC: Adaptation::Icap::TheConfig.repeat_limit
8454 DEFAULT_DOC: No retries are allowed.
8456 Limits the number of retries allowed.
8458 Communication errors due to persistent connection race
8459 conditions are unavoidable, automatically retried, and do not
8460 count against this limit.
8462 See also: icap_retry
8468 -----------------------------------------------------------------------------
8471 NAME: check_hostnames
8474 LOC: Config.onoff.check_hostnames
8476 For security and stability reasons Squid can check
8477 hostnames for Internet standard RFC compliance. If you want
8478 Squid to perform these checks turn this directive on.
8481 NAME: allow_underscore
8484 LOC: Config.onoff.allow_underscore
8486 Underscore characters is not strictly allowed in Internet hostnames
8487 but nevertheless used by many sites. Set this to off if you want
8488 Squid to be strict about the standard.
8489 This check is performed only when check_hostnames is set to on.
8492 NAME: dns_retransmit_interval
8495 LOC: Config.Timeout.idns_retransmit
8497 Initial retransmit interval for DNS queries. The interval is
8498 doubled each time all configured DNS servers have been tried.
8504 LOC: Config.Timeout.idns_query
8506 DNS Query timeout. If no response is received to a DNS query
8507 within this time all DNS servers for the queried domain
8508 are assumed to be unavailable.
8511 NAME: dns_packet_max
8513 DEFAULT_DOC: EDNS disabled
8515 LOC: Config.dns.packet_max
8517 Maximum number of bytes packet size to advertise via EDNS.
8518 Set to "none" to disable EDNS large packet support.
8520 For legacy reasons DNS UDP replies will default to 512 bytes which
8521 is too small for many responses. EDNS provides a means for Squid to
8522 negotiate receiving larger responses back immediately without having
8523 to failover with repeat requests. Responses larger than this limit
8524 will retain the old behaviour of failover to TCP DNS.
8526 Squid has no real fixed limit internally, but allowing packet sizes
8527 over 1500 bytes requires network jumbogram support and is usually not
8530 WARNING: The RFC also indicates that some older resolvers will reply
8531 with failure of the whole request if the extension is added. Some
8532 resolvers have already been identified which will reply with mangled
8533 EDNS response on occasion. Usually in response to many-KB jumbogram
8534 sizes being advertised by Squid.
8535 Squid will currently treat these both as an unable-to-resolve domain
8536 even if it would be resolvable without EDNS.
8543 DEFAULT_DOC: Search for single-label domain names is disabled.
8544 LOC: Config.onoff.res_defnames
8546 Normally the RES_DEFNAMES resolver option is disabled
8547 (see res_init(3)). This prevents caches in a hierarchy
8548 from interpreting single-component hostnames locally. To allow
8549 Squid to handle single-component names, enable this option.
8552 NAME: dns_multicast_local
8556 DEFAULT_DOC: Search for .local and .arpa names is disabled.
8557 LOC: Config.onoff.dns_mdns
8559 When set to on, Squid sends multicast DNS lookups on the local
8560 network for domains ending in .local and .arpa.
8561 This enables local servers and devices to be contacted in an
8562 ad-hoc or zero-configuration network environment.
8565 NAME: dns_nameservers
8568 DEFAULT_DOC: Use operating system definitions
8569 LOC: Config.dns_nameservers
8571 Use this if you want to specify a list of DNS name servers
8572 (IP addresses) to use instead of those given in your
8573 /etc/resolv.conf file.
8575 On Windows platforms, if no value is specified here or in
8576 the /etc/resolv.conf file, the list of DNS name servers are
8577 taken from the Windows registry, both static and dynamic DHCP
8578 configurations are supported.
8580 Example: dns_nameservers 10.0.0.1 192.172.0.4
8585 DEFAULT: @DEFAULT_HOSTS@
8586 LOC: Config.etcHostsPath
8588 Location of the host-local IP name-address associations
8589 database. Most Operating Systems have such a file on different
8591 - Un*X & Linux: /etc/hosts
8592 - Windows NT/2000: %SystemRoot%\system32\drivers\etc\hosts
8593 (%SystemRoot% value install default is c:\winnt)
8594 - Windows XP/2003: %SystemRoot%\system32\drivers\etc\hosts
8595 (%SystemRoot% value install default is c:\windows)
8596 - Windows 9x/Me: %windir%\hosts
8597 (%windir% value is usually c:\windows)
8598 - Cygwin: /etc/hosts
8600 The file contains newline-separated definitions, in the
8601 form ip_address_in_dotted_form name [name ...] names are
8602 whitespace-separated. Lines beginning with an hash (#)
8603 character are comments.
8605 The file is checked at startup and upon configuration.
8606 If set to 'none', it won't be checked.
8607 If append_domain is used, that domain will be added to
8608 domain-local (i.e. not containing any dot character) host
8614 LOC: Config.appendDomain
8616 DEFAULT_DOC: Use operating system definitions
8618 Appends local domain name to hostnames without any dots in
8619 them. append_domain must begin with a period.
8621 Be warned there are now Internet names with no dots in
8622 them using only top-domain names, so setting this may
8623 cause some Internet sites to become unavailable.
8626 append_domain .yourdomain.com
8629 NAME: ignore_unknown_nameservers
8631 LOC: Config.onoff.ignore_unknown_nameservers
8634 By default Squid checks that DNS responses are received
8635 from the same IP addresses they are sent to. If they
8636 don't match, Squid ignores the response and writes a warning
8637 message to cache.log. You can allow responses from unknown
8638 nameservers by setting this option to 'off'.
8644 LOC: Config.dns.v4_first
8646 With the IPv6 Internet being as fast or faster than IPv4 Internet
8647 for most networks Squid prefers to contact websites over IPv6.
8649 This option reverses the order of preference to make Squid contact
8650 dual-stack websites over IPv4 first. Squid will still perform both
8651 IPv6 and IPv4 DNS lookups before connecting.
8654 This option will restrict the situations under which IPv6
8655 connectivity is used (and tested). Hiding network problems
8656 which would otherwise be detected and warned about.
8660 COMMENT: (number of entries)
8663 LOC: Config.ipcache.size
8665 Maximum number of DNS IP cache entries.
8672 LOC: Config.ipcache.low
8679 LOC: Config.ipcache.high
8681 The size, low-, and high-water marks for the IP cache.
8684 NAME: fqdncache_size
8685 COMMENT: (number of entries)
8688 LOC: Config.fqdncache.size
8690 Maximum number of FQDN cache entries.
8695 -----------------------------------------------------------------------------
8698 NAME: configuration_includes_quoted_values
8700 TYPE: configuration_includes_quoted_values
8702 LOC: ConfigParser::RecognizeQuotedValues
8704 If set, Squid will recognize each "quoted string" after a configuration
8705 directive as a single parameter. The quotes are stripped before the
8706 parameter value is interpreted or used.
8707 See "Values with spaces, quotes, and other special characters"
8708 section for more details.
8715 LOC: Config.onoff.mem_pools
8717 If set, Squid will keep pools of allocated (but unused) memory
8718 available for future use. If memory is a premium on your
8719 system and you believe your malloc library outperforms Squid
8720 routines, disable this.
8723 NAME: memory_pools_limit
8727 LOC: Config.MemPools.limit
8729 Used only with memory_pools on:
8730 memory_pools_limit 50 MB
8732 If set to a non-zero value, Squid will keep at most the specified
8733 limit of allocated (but unused) memory in memory pools. All free()
8734 requests that exceed this limit will be handled by your malloc
8735 library. Squid does not pre-allocate any memory, just safe-keeps
8736 objects that otherwise would be free()d. Thus, it is safe to set
8737 memory_pools_limit to a reasonably high value even if your
8738 configuration will use less memory.
8740 If set to none, Squid will keep all memory it can. That is, there
8741 will be no limit on the total amount of memory used for safe-keeping.
8743 To disable memory allocation optimization, do not set
8744 memory_pools_limit to 0 or none. Set memory_pools to "off" instead.
8746 An overhead for maintaining memory pools is not taken into account
8747 when the limit is checked. This overhead is close to four bytes per
8748 object kept. However, pools may actually _save_ memory because of
8749 reduced memory thrashing in your malloc library.
8753 COMMENT: on|off|transparent|truncate|delete
8756 LOC: opt_forwarded_for
8758 If set to "on", Squid will append your client's IP address
8759 in the HTTP requests it forwards. By default it looks like:
8761 X-Forwarded-For: 192.1.2.3
8763 If set to "off", it will appear as
8765 X-Forwarded-For: unknown
8767 If set to "transparent", Squid will not alter the
8768 X-Forwarded-For header in any way.
8770 If set to "delete", Squid will delete the entire
8771 X-Forwarded-For header.
8773 If set to "truncate", Squid will remove all existing
8774 X-Forwarded-For entries, and place the client IP as the sole entry.
8777 NAME: cachemgr_passwd
8778 TYPE: cachemgrpasswd
8780 DEFAULT_DOC: No password. Actions which require password are denied.
8781 LOC: Config.passwd_list
8783 Specify passwords for cachemgr operations.
8785 Usage: cachemgr_passwd password action action ...
8787 Some valid actions are (see cache manager menu for a full list):
8827 * Indicates actions which will not be performed without a
8828 valid password, others can be performed if not listed here.
8830 To disable an action, set the password to "disable".
8831 To allow performing an action without a password, set the
8834 Use the keyword "all" to set the same password for all actions.
8837 cachemgr_passwd secret shutdown
8838 cachemgr_passwd lesssssssecret info stats/objects
8839 cachemgr_passwd disable all
8846 LOC: Config.onoff.client_db
8848 If you want to disable collecting per-client statistics,
8849 turn off client_db here.
8852 NAME: refresh_all_ims
8856 LOC: Config.onoff.refresh_all_ims
8858 When you enable this option, squid will always check
8859 the origin server for an update when a client sends an
8860 If-Modified-Since request. Many browsers use IMS
8861 requests when the user requests a reload, and this
8862 ensures those clients receive the latest version.
8864 By default (off), squid may return a Not Modified response
8865 based on the age of the cached version.
8868 NAME: reload_into_ims
8869 IFDEF: USE_HTTP_VIOLATIONS
8873 LOC: Config.onoff.reload_into_ims
8875 When you enable this option, client no-cache or ``reload''
8876 requests will be changed to If-Modified-Since requests.
8877 Doing this VIOLATES the HTTP standard. Enabling this
8878 feature could make you liable for problems which it
8881 see also refresh_pattern for a more selective approach.
8884 NAME: connect_retries
8886 LOC: Config.connect_retries
8888 DEFAULT_DOC: Do not retry failed connections.
8890 This sets the maximum number of connection attempts made for each
8891 TCP connection. The connect_retries attempts must all still
8892 complete within the connection timeout period.
8894 The default is not to re-try if the first connection attempt fails.
8895 The (not recommended) maximum is 10 tries.
8897 A warning message will be generated if it is set to a too-high
8898 value and the configured value will be over-ridden.
8900 Note: These re-tries are in addition to forward_max_tries
8901 which limit how many different addresses may be tried to find
8905 NAME: retry_on_error
8907 LOC: Config.retry.onerror
8910 If set to ON Squid will automatically retry requests when
8911 receiving an error response with status 403 (Forbidden),
8912 500 (Internal Error), 501 or 503 (Service not available).
8913 Status 502 and 504 (Gateway errors) are always retried.
8915 This is mainly useful if you are in a complex cache hierarchy to
8916 work around access control errors.
8918 NOTE: This retry will attempt to find another working destination.
8919 Which is different from the server which just failed.
8922 NAME: as_whois_server
8924 LOC: Config.as_whois_server
8925 DEFAULT: whois.ra.net
8927 WHOIS server to query for AS numbers. NOTE: AS numbers are
8928 queried only when Squid starts up, not for every request.
8933 LOC: Config.onoff.offline
8936 Enable this option and Squid will never try to validate cached
8940 NAME: uri_whitespace
8941 TYPE: uri_whitespace
8942 LOC: Config.uri_whitespace
8945 What to do with requests that have whitespace characters in the
8948 strip: The whitespace characters are stripped out of the URL.
8949 This is the behavior recommended by RFC2396 and RFC3986
8950 for tolerant handling of generic URI.
8951 NOTE: This is one difference between generic URI and HTTP URLs.
8953 deny: The request is denied. The user receives an "Invalid
8955 This is the behaviour recommended by RFC2616 for safe
8956 handling of HTTP request URL.
8958 allow: The request is allowed and the URI is not changed. The
8959 whitespace characters remain in the URI. Note the
8960 whitespace is passed to redirector processes if they
8962 Note this may be considered a violation of RFC2616
8963 request parsing where whitespace is prohibited in the
8966 encode: The request is allowed and the whitespace characters are
8967 encoded according to RFC1738.
8969 chop: The request is allowed and the URI is chopped at the
8973 NOTE the current Squid implementation of encode and chop violates
8974 RFC2616 by not using a 301 redirect after altering the URL.
8979 LOC: Config.chroot_dir
8982 Specifies a directory where Squid should do a chroot() while
8983 initializing. This also causes Squid to fully drop root
8984 privileges after initializing. This means, for example, if you
8985 use a HTTP port less than 1024 and try to reconfigure, you may
8986 get an error saying that Squid can not open the port.
8989 NAME: balance_on_multiple_ip
8991 LOC: Config.onoff.balance_on_multiple_ip
8994 Modern IP resolvers in squid sort lookup results by preferred access.
8995 By default squid will use these IP in order and only rotates to
8996 the next listed when the most preffered fails.
8998 Some load balancing servers based on round robin DNS have been
8999 found not to preserve user session state across requests
9000 to different IP addresses.
9002 Enabling this directive Squid rotates IP's per request.
9005 NAME: pipeline_prefetch
9006 TYPE: pipelinePrefetch
9007 LOC: Config.pipeline_max_prefetch
9009 DEFAULT_DOC: Do not pre-parse pipelined requests.
9011 HTTP clients may send a pipeline of 1+N requests to Squid using a
9012 single connection, without waiting for Squid to respond to the first
9013 of those requests. This option limits the number of concurrent
9014 requests Squid will try to handle in parallel. If set to N, Squid
9015 will try to receive and process up to 1+N requests on the same
9016 connection concurrently.
9018 Defaults to 0 (off) for bandwidth management and access logging
9021 NOTE: pipelining requires persistent connections to clients.
9023 WARNING: pipelining breaks NTLM and Negotiate/Kerberos authentication.
9026 NAME: high_response_time_warning
9029 LOC: Config.warnings.high_rptm
9031 DEFAULT_DOC: disabled.
9033 If the one-minute median response time exceeds this value,
9034 Squid prints a WARNING with debug level 0 to get the
9035 administrators attention. The value is in milliseconds.
9038 NAME: high_page_fault_warning
9040 LOC: Config.warnings.high_pf
9042 DEFAULT_DOC: disabled.
9044 If the one-minute average page fault rate exceeds this
9045 value, Squid prints a WARNING with debug level 0 to get
9046 the administrators attention. The value is in page faults
9050 NAME: high_memory_warning
9052 LOC: Config.warnings.high_memory
9053 IFDEF: HAVE_MSTATS&&HAVE_GNUMALLOC_H
9055 DEFAULT_DOC: disabled.
9057 If the memory usage (as determined by gnumalloc, if available and used)
9058 exceeds this amount, Squid prints a WARNING with debug level 0 to get
9059 the administrators attention.
9061 # TODO: link high_memory_warning to mempools?
9063 NAME: sleep_after_fork
9064 COMMENT: (microseconds)
9066 LOC: Config.sleep_after_fork
9069 When this is set to a non-zero value, the main Squid process
9070 sleeps the specified number of microseconds after a fork()
9071 system call. This sleep may help the situation where your
9072 system reports fork() failures due to lack of (virtual)
9073 memory. Note, however, if you have a lot of child
9074 processes, these sleep delays will add up and your
9075 Squid will not service requests for some amount of time
9076 until all the child processes have been started.
9077 On Windows value less then 1000 (1 milliseconds) are
9081 NAME: windows_ipaddrchangemonitor
9082 IFDEF: _SQUID_WINDOWS_
9086 LOC: Config.onoff.WIN32_IpAddrChangeMonitor
9088 On Windows Squid by default will monitor IP address changes and will
9089 reconfigure itself after any detected event. This is very useful for
9090 proxies connected to internet with dial-up interfaces.
9091 In some cases (a Proxy server acting as VPN gateway is one) it could be
9092 desiderable to disable this behaviour setting this to 'off'.
9093 Note: after changing this, Squid service must be restarted.
9098 IFDEF: USE_SQUID_EUI
9100 LOC: Eui::TheConfig.euiLookup
9102 Whether to lookup the EUI or MAC address of a connected client.
9105 NAME: max_filedescriptors max_filedesc
9108 DEFAULT_DOC: Use operating system limits set by ulimit.
9109 LOC: Config.max_filedescriptors
9111 Reduce the maximum number of filedescriptors supported below
9112 the usual operating system defaults.
9114 Remove from squid.conf to inherit the current ulimit setting.
9116 Note: Changing this requires a restart of Squid. Also
9117 not all I/O types supports large values (eg on Windows).
9124 DEFAULT_DOC: SMP support disabled.
9126 Number of main Squid processes or "workers" to fork and maintain.
9127 0: "no daemon" mode, like running "squid -N ..."
9128 1: "no SMP" mode, start one main Squid process daemon (default)
9129 N: start N main Squid process daemons (i.e., SMP mode)
9131 In SMP mode, each worker does nearly all what a single Squid daemon
9132 does (e.g., listen on http_port and forward HTTP requests).
9135 NAME: cpu_affinity_map
9136 TYPE: CpuAffinityMap
9137 LOC: Config.cpuAffinityMap
9139 DEFAULT_DOC: Let operating system decide.
9141 Usage: cpu_affinity_map process_numbers=P1,P2,... cores=C1,C2,...
9143 Sets 1:1 mapping between Squid processes and CPU cores. For example,
9145 cpu_affinity_map process_numbers=1,2,3,4 cores=1,3,5,7
9147 affects processes 1 through 4 only and places them on the first
9148 four even cores, starting with core #1.
9150 CPU cores are numbered starting from 1. Requires support for
9151 sched_getaffinity(2) and sched_setaffinity(2) system calls.
9153 Multiple cpu_affinity_map options are merged.