1 ## Copyright (C) 1996-2025 The Squid Software Foundation and contributors
3 ## Squid software is distributed under GPLv2+ license and includes
4 ## contributions from numerous individuals and organizations.
5 ## Please see the COPYING and CONTRIBUTORS files for details.
10 ----------------------------
12 This is the documentation for the Squid configuration file.
13 This documentation can also be found online at:
14 http://www.squid-cache.org/Doc/config/
16 You may wish to look at the Squid home page and wiki for the
17 FAQ and other documentation:
18 http://www.squid-cache.org/
19 https://wiki.squid-cache.org/SquidFaq
20 https://wiki.squid-cache.org/ConfigExamples
22 This documentation shows what the defaults for various directives
23 happen to be. If you don't need to change the default, you should
24 leave the line out of your squid.conf in most cases.
26 In some cases "none" refers to no default setting at all,
27 while in other cases it refers to the value of the option
28 - the comments for that keyword indicate if this is the case.
33 Configuration options can be included using the "include" directive.
34 Include takes a list of files to include. Quoting and wildcards are
39 include /path/to/included/file/squid.acl.config
41 Includes can be nested up to a hard-coded depth of 16 levels.
42 This arbitrary restriction is to prevent recursive include references
43 from causing Squid entering an infinite loop whilst trying to load
46 Values with byte units
48 Squid accepts size units on some size related directives. All
49 such directives are documented with a default value displaying
52 Units accepted by Squid are:
54 KB - Kilobyte (2^10, 1'024 bytes)
55 MB - Megabyte (2^20, 1'048'576 bytes)
56 GB - Gigabyte (2^30, 1'073'741'824 bytes)
57 Squid does not yet support KiB, MiB, and GiB unit names.
59 Values with time units
61 Time-related directives marked with either "time-units" or
62 "time-units-small" accept a time unit. The supported time units are:
64 nanosecond (time-units-small only)
65 microsecond (time-units-small only)
74 year - 31557790080 milliseconds (just over 365 days)
77 Values with spaces, quotes, and other special characters
79 Squid supports directive parameters with spaces, quotes, and other
80 special characters. Surround such parameters with "double quotes". Use
81 the configuration_includes_quoted_values directive to enable or
84 Squid supports reading configuration option parameters from external
85 files using the syntax:
86 parameters("/path/filename")
88 acl allowlist dstdomain parameters("/etc/squid/allowlist.txt")
90 Conditional configuration
92 If-statements can be used to make configuration directives
96 ... regular configuration directives ...
98 ... regular configuration directives ...]
101 The else part is optional. The keywords "if", "else", and "endif"
102 must be typed on their own lines, as if they were regular
103 configuration directives.
105 NOTE: An else-if condition is not supported.
107 These individual conditions types are supported:
110 Always evaluates to true.
112 Always evaluates to false.
113 <integer> = <integer>
114 Equality comparison of two integer numbers.
119 The following SMP-related preprocessor macros can be used.
121 ${process_name} expands to the current Squid process "name"
122 (e.g., squid1, squid2, or cache1).
124 ${process_number} expands to the current Squid process
125 identifier, which is an integer number (e.g., 1, 2, 3) unique
126 across all Squid processes of the current service instance.
128 ${service_name} expands into the current Squid service instance
129 name identifier which is provided by -n on the command line.
133 Logformat macros can be used in many places outside of the logformat
134 directive. In theory, all of the logformat codes can be used as %macros,
135 where they are supported. In practice, a %macro expands as a dash (-) when
136 the transaction does not yet have enough information and a value is needed.
138 There is no definitive list of what tokens are available at the various
139 stages of the transaction.
141 And some information may already be available to Squid but not yet
142 committed where the macro expansion code can access it (report
143 such instances!). The macro will be expanded into a single dash
144 ('-') in such cases. Not all macros have been tested.
148 # options still not yet ported from 2.7 to 3.x
149 NAME: broken_vary_encoding
152 This option is not yet supported by Squid-3.
158 This option is not yet supported by Squid-3.
164 This option is not yet supported by Squid-3.
167 NAME: external_refresh_check
170 This option is not yet supported by Squid-3.
173 NAME: location_rewrite_program location_rewrite_access location_rewrite_children location_rewrite_concurrency
176 This option is not yet supported by Squid-3.
179 NAME: refresh_stale_hit
182 This option is not yet supported by Squid-3.
185 # Options removed in 7.x
189 Remove this line. Squid no longer supports this feature.
193 # Options removed in 6.x
197 Remove this line. Squid no longer supports this feature.
203 Remove this line. Squid no longer supports this feature.
206 NAME: announce_period
209 Remove this line. Squid no longer supports this feature.
215 Remove this line. Squid no longer supports this feature.
218 NAME: request_entities
221 Remove this line. Squid now accepts HTTP/1.1 requests with bodies.
222 To simplify UI and code, Squid rejects certain HTTP/1.0 requests with bodies.
225 # Options removed in 5.x
229 Remove this line. Squid no longer supports preferential treatment of DNS A records.
232 # Options removed in 4.x
233 NAME: cache_peer_domain cache_host_domain
236 Replace with dstdomain ACLs and cache_peer_access.
242 Remove this line. The behaviour enabled by this is no longer needed.
245 NAME: sslproxy_cafile
248 Remove this line. Use tls_outgoing_options cafile= instead.
251 NAME: sslproxy_capath
254 Remove this line. Use tls_outgoing_options capath= instead.
257 NAME: sslproxy_cipher
260 Remove this line. Use tls_outgoing_options cipher= instead.
263 NAME: sslproxy_client_certificate
266 Remove this line. Use tls_outgoing_options cert= instead.
269 NAME: sslproxy_client_key
272 Remove this line. Use tls_outgoing_options key= instead.
278 Remove this line. Use tls_outgoing_options flags= instead.
281 NAME: sslproxy_options
284 Remove this line. Use tls_outgoing_options options= instead.
287 NAME: sslproxy_version
290 Remove this line. Use tls_outgoing_options options= instead.
293 # Options removed in 3.5
294 NAME: hierarchy_stoplist
297 Remove this line. Use always_direct or cache_peer_access ACLs instead if you need to prevent cache_peer use.
300 # Options removed in 3.4
304 Remove this line. Use acls with access_log directives to control access logging
310 Remove this line. Use acls with icap_log directives to control icap logging
313 # Options Removed in 3.3
314 NAME: ignore_ims_on_miss
317 Remove this line. The HTTP/1.1 feature is now configured by 'cache_miss_revalidate'.
320 # Options Removed in 3.2
321 NAME: balance_on_multiple_ip
324 Remove this line. Squid performs a 'Happy Eyeballs' algorithm, this multiple-IP algorithm is not longer relevant.
327 NAME: chunked_request_body_max_size
330 Remove this line. Squid is now HTTP/1.1 compliant.
333 NAME: dns_v4_fallback
336 Remove this line. Squid performs a 'Happy Eyeballs' algorithm, the 'fallback' algorithm is no longer relevant.
339 NAME: emulate_httpd_log
342 Replace this with an access_log directive using the format 'common' or 'combined'.
348 Use a regular access.log with ACL limiting it to MISS events.
354 Remove this line. Configure FTP page display using the CSS controls in errorpages.css instead.
357 NAME: ignore_expect_100
360 Remove this line. The HTTP/1.1 feature is now fully supported by default.
366 Remove this option from your config. To log FQDN use %>A in the log format.
369 NAME: log_ip_on_direct
372 Remove this option from your config. To log server or peer names use %<A in the log format.
375 NAME: maximum_single_addr_tries
378 Replaced by connect_retries. The behaviour has changed, please read the documentation before altering.
381 NAME: referer_log referrer_log
384 Replace this with an access_log directive using the format 'referrer'.
390 Remove this line. The feature is supported by default in storage types where update is implemented.
393 NAME: url_rewrite_concurrency
396 Remove this line. Set the 'concurrency=' option of url_rewrite_children instead.
402 Replace this with an access_log directive using the format 'useragent'.
405 # Options Removed in 3.1
409 Remove this line. DNS is no longer tested on startup.
412 NAME: extension_methods
415 Remove this line. All valid methods for HTTP are accepted by default.
418 # 2.7 Options Removed/Replaced in 3.2
423 # 2.7 Options Removed/Replaced in 3.1
431 Remove this line. HTTP/1.1 is supported by default.
434 NAME: upgrade_http0.9
437 Remove this line. ICY/1.0 streaming protocol is supported by default.
440 NAME: zph_local zph_mode zph_option zph_parent zph_sibling
443 Alter these entries. Use the qos_flows directive instead.
446 # Options Removed in 3.0
450 Since squid-3.0 replace with request_header_access or reply_header_access
451 depending on whether you wish to match client requests or server replies.
454 NAME: httpd_accel_no_pmtu_disc
457 Since squid-3.0 use the 'disable-pmtu-discovery' flag on http_port instead.
460 NAME: wais_relay_host
463 Replace this line with 'cache_peer' configuration.
466 NAME: wais_relay_port
469 Replace this line with 'cache_peer' configuration.
474 -----------------------------------------------------------------------------
481 DEFAULT_DOC: SMP support disabled.
483 Number of main Squid processes or "workers" to fork and maintain.
484 0: "no daemon" mode, like running "squid -N ..."
485 1: "no SMP" mode, start one main Squid process daemon (default)
486 N: start N main Squid process daemons (i.e., SMP mode)
488 In SMP mode, each worker does nearly all what a single Squid daemon
489 does (e.g., listen on http_port and forward HTTP requests).
491 Changing the number of workers requires a restart: Squid warns about but
492 otherwise ignores attempts to change this setting via reconfiguration.
495 NAME: cpu_affinity_map
497 LOC: Config.cpuAffinityMap
499 DEFAULT_DOC: Let operating system decide.
501 Usage: cpu_affinity_map process_numbers=P1,P2,... cores=C1,C2,...
503 Sets 1:1 mapping between Squid processes and CPU cores. For example,
505 cpu_affinity_map process_numbers=1,2,3,4 cores=1,3,5,7
507 affects processes 1 through 4 only and places them on the first
508 four even cores, starting with core #1.
510 CPU cores are numbered starting from 1. Requires support for
511 sched_getaffinity(2) and sched_setaffinity(2) system calls.
513 Multiple cpu_affinity_map options are merged.
518 NAME: shared_memory_locking
521 LOC: Config.shmLocking
524 Whether to ensure that all required shared memory is available by
525 "locking" that shared memory into RAM when Squid starts. The
526 alternative is faster startup time followed by slightly slower
527 performance and, if not enough RAM is actually available during
528 runtime, mysterious crashes.
530 SMP Squid uses many shared memory segments. These segments are
531 brought into Squid memory space using an mmap(2) system call. During
532 Squid startup, the mmap() call often succeeds regardless of whether
533 the system has enough RAM. In general, Squid cannot tell whether the
534 kernel applies this "optimistic" memory allocation policy (but
535 popular modern kernels usually use it).
537 Later, if Squid attempts to actually access the mapped memory
538 regions beyond what the kernel is willing to allocate, the
539 "optimistic" kernel simply kills Squid kid with a SIGBUS signal.
540 Some of the memory limits enforced by the kernel are currently
541 poorly understood: We do not know how to detect and check them. This
542 option ensures that the mapped memory will be available.
544 This option may have a positive performance side-effect: Locking
545 memory at start avoids runtime paging I/O. Paging slows Squid down.
547 Locking memory may require a large enough RLIMIT_MEMLOCK OS limit,
548 CAP_IPC_LOCK capability, or equivalent.
551 NAME: hopeless_kid_revival_delay
554 LOC: Config.hopelessKidRevivalDelay
557 Normally, when a kid process dies, Squid immediately restarts the
558 kid. A kid experiencing frequent deaths is marked as "hopeless" for
559 the duration specified by this directive. Hopeless kids are not
560 automatically restarted.
562 Currently, zero values are not supported because they result in
563 misconfigured SMP Squid instances running forever, endlessly
564 restarting each dying kid. To effectively disable hopeless kids
565 revival, set the delay to a huge value (e.g., 1 year).
567 Reconfiguration also clears all hopeless kids designations, allowing
568 for manual revival of hopeless kids.
572 OPTIONS FOR AUTHENTICATION
573 -----------------------------------------------------------------------------
579 LOC: Auth::TheConfig.schemes
582 This is used to define parameters for the various authentication
583 schemes supported by Squid.
585 format: auth_param scheme parameter [setting]
587 The order in which authentication schemes are presented to the client is
588 dependent on the order the scheme first appears in config file. IE
589 has a bug (it's not RFC 2617 compliant) in that it will use the basic
590 scheme if basic is the first entry presented, even if more secure
591 schemes are presented. For now use the order in the recommended
592 settings section below. If other browsers have difficulties (don't
593 recognize the schemes offered even if you are using basic) either
594 put basic first, or disable the other schemes (by commenting out their
597 Once an authentication scheme is fully configured, it can only be
598 shutdown by shutting squid down and restarting. Changes can be made on
599 the fly and activated with a reconfigure. I.E. You can change to a
600 different helper, but not unconfigure the helper completely.
602 Please note that while this directive defines how Squid processes
603 authentication it does not automatically activate authentication. For a
604 given transaction, (re)authentication is requested in two primary cases
605 detailed below: initial authentication and re-authentication.
607 A client without credentials is requested to authenticate if one of the
608 following ACLs is evaluated by an http_access or adapted_http_access rule:
611 * proxy_auth_regex ACL
613 * external ACL with %ul logformat %code used in FORMAT parameters
614 * external ACL with %LOGIN macro used in FORMAT parameters;
615 this legacy macro currently behaves the same as %ul logformat %code
617 A client with credentials is requested to re-authenticate if http_access
618 or adapted_http_access denies its request _and_ the last evaluated ACL was
619 either proxy_auth, proxy_auth_regex, or an external
620 ACL with %ul or %LOGIN parameter (regardless of whether that last
621 evaluated ACL matched the denied request). Note that a max_user_ip ACL
622 does not have this effect: Requests denied after evaluating max_user_ip
623 trigger an HTTP 403 (Forbidden) response rather than re-authentication.
625 In both initial authentication and re-authentication cases, client access
626 is denied, typically with an HTTP 407 (Proxy Authentication Required) or
627 an HTTP 401 (Unauthorized) response.
629 WARNING: authentication can't be used in a transparently intercepting
630 proxy as the client then thinks it is talking to an origin server and
631 not the proxy. This is a limitation of bending the TCP/IP protocol to
632 transparently intercepting port 80, not a limitation in Squid.
633 Ports flagged 'transparent', 'intercept', or 'tproxy' have
634 authentication disabled.
636 === Parameters common to all schemes. ===
639 Specifies the command for the external authenticator.
641 By default, each authentication scheme is not used unless a
642 program is specified.
644 See https://wiki.squid-cache.org/Features/AddonHelpers for
645 more details on helper operations and creating your own.
648 Specifies a string to be append to request line format for
649 the authentication helper. "Quoted" format values may contain
650 spaces and logformat %macros. In theory, any logformat %macro
651 can be used. In practice, a %macro expands as a dash (-) if
652 the helper request is sent before the required macro
653 information is available to Squid.
655 By default, Squid uses request formats provided in
656 scheme-specific examples below (search for %credentials).
658 The expanded key_extras value is added to the Squid credentials
659 cache and, hence, will affect authentication. It can be used to
660 authenticate different users with identical user names (e.g.,
661 when user authentication depends on http_port).
663 Avoid adding frequently changing information to key_extras. For
664 example, if you add user source IP, and it changes frequently
665 in your environment, then max_user_ip ACL is going to treat
666 every user+IP combination as a unique "user", breaking the ACL
667 and wasting a lot of memory on those user records. It will also
668 force users to authenticate from scratch whenever their IP
672 Specifies the protection scope (aka realm name) which is to be
673 reported to the client for the authentication scheme. It is
674 commonly part of the text the user will see when prompted for
675 their username and password.
677 For Basic the default is "Squid proxy-caching web server".
678 For Digest there is no default, this parameter is mandatory.
679 For NTLM and Negotiate this parameter is ignored.
681 "children" numberofchildren [startup=N] [idle=N] [concurrency=N]
682 [queue-size=N] [on-persistent-overload=action]
683 [reservation-timeout=seconds]
685 The maximum number of authenticator processes to spawn. If
686 you start too few Squid will have to wait for them to process
687 a backlog of credential verifications, slowing it down. When
688 password verifications are done via a (slow) network you are
689 likely to need lots of authenticator processes.
691 The startup= and idle= options permit some skew in the exact
692 amount run. A minimum of startup=N will begin during startup
693 and reconfigure. Squid will start more in groups of up to
694 idle=N in an attempt to meet traffic needs and to keep idle=N
695 free above those traffic needs up to the maximum.
697 The concurrency= option sets the number of concurrent requests
698 the helper can process. The default of 0 is used for helpers
699 who only supports one request at a time. Setting this to a
700 number greater than 0 changes the protocol used to include a
701 channel ID field first on the request/response line, allowing
702 multiple requests to be sent to the same helper in parallel
703 without waiting for the response.
705 Concurrency must not be set unless it's known the helper
706 supports the input format with channel-ID fields.
708 The queue-size option sets the maximum number of queued
709 requests. A request is queued when no existing child can
710 accept it due to concurrency limit and no new child can be
711 started due to numberofchildren limit. The default maximum is
712 2*numberofchildren. Squid is allowed to temporarily exceed the
713 configured maximum, marking the affected helper as
714 "overloaded". If the helper overload lasts more than 3
715 minutes, the action prescribed by the on-persistent-overload
718 The on-persistent-overload=action option specifies Squid
719 reaction to a new helper request arriving when the helper
720 has been overloaded for more that 3 minutes already. The number
721 of queued requests determines whether the helper is overloaded
722 (see the queue-size option).
724 Two actions are supported:
726 die Squid worker quits. This is the default behavior.
728 ERR Squid treats the helper request as if it was
729 immediately submitted, and the helper immediately
730 replied with an ERR response. This action has no effect
731 on the already queued and in-progress helper requests.
733 NOTE: NTLM and Negotiate schemes do not support concurrency
734 in the Squid code module even though some helpers can.
736 The reservation-timeout=seconds option allows NTLM and Negotiate
737 helpers to forget about clients that abandon their in-progress
738 connection authentication without closing the connection. The
739 timeout is measured since the last helper response received by
740 Squid for the client. Fractional seconds are not supported.
742 After the timeout, the helper will be used for other clients if
743 there are no unreserved helpers available. In the latter case,
744 the old client attempt to resume authentication will not be
745 forwarded to the helper (and the client should open a new HTTP
746 connection and retry authentication from scratch).
748 By default, reservations do not expire and clients that keep
749 their connections open without completing authentication may
750 exhaust all NTLM and Negotiate helpers.
753 If you experience problems with PUT/POST requests when using
754 the NTLM or Negotiate schemes then you can try setting this
755 to off. This will cause Squid to forcibly close the connection
756 on the initial request where the browser asks which schemes
757 are supported by the proxy.
759 For Basic and Digest this parameter is ignored.
762 Useful for sending credentials to authentication backends that
763 expect UTF-8 encoding (e.g., LDAP).
765 When this option is enabled, Squid uses HTTP Accept-Language
766 request header to guess the received credentials encoding
767 (ISO-Latin-1, CP1251, or UTF-8) and then converts the first
768 two encodings into UTF-8.
770 When this option is disabled and by default, Squid sends
771 credentials in their original (i.e. received) encoding.
773 This parameter is only honored for Basic and Digest schemes.
774 For Basic, the entire username:password credentials are
775 checked and, if necessary, re-encoded. For Digest -- just the
776 username component. For NTLM and Negotiate schemes, this
777 parameter is ignored.
779 IF HAVE_AUTH_MODULE_BASIC
780 === Basic authentication parameters ===
782 "credentialsttl" timetolive
783 Specifies how long squid assumes an externally validated
784 username:password pair is valid for - in other words how
785 often the helper program is called for that user. Set this
786 low to force revalidation with short lived passwords.
788 NOTE: setting this high does not impact your susceptibility
789 to replay attacks unless you are using an one-time password
790 system (such as SecureID). If you are using such a system,
791 you will be vulnerable to replay attacks unless you also
792 use the max_user_ip ACL in an http_access rule.
794 "casesensitive" on|off
795 Specifies whether upper case letters in client-sent usernames are
796 preserved. By default and when explicitly set to "off", a username
797 extracted from Proxy-Authorization or Authorization request header is
798 forced to lower case before user credentials are checked or stored.
800 Most user databases are case insensitive, allowing the same username to be
801 spelled using both lower and upper case letters. For such databases,
802 either setting should work, but forcing usernames to lower case may
803 still make a big difference for Squid internal caches like those used by
804 an external ACL with %un logformat code in FORMAT and a user_max_ip ACL.
806 When working with a case sensitive database, set casesensitive to "on".
808 Squid ACLs like proxy_auth are case-sensitive by default. An ACL using
809 upper case letters in user names (e.g., `acl badGuys proxy_auth Bob`)
810 will not match any user with Basic Authentication credentials unless
811 casesensitive is explicitly turned "on" (to preserve "Bob" username
812 instead of converting it to "bob" before the ACL is checked).
815 IF HAVE_AUTH_MODULE_DIGEST
816 === Digest authentication parameters ===
818 "nonce_garbage_interval" timeinterval
819 Specifies the interval that nonces that have been issued
820 to client_agent's are checked for validity.
822 "nonce_max_duration" timeinterval
823 Specifies the maximum length of time a given nonce will be
826 "nonce_max_count" number
827 Specifies the maximum number of times a given nonce can be
830 "nonce_strictness" on|off
831 Determines if squid requires strict increment-by-1 behavior
832 for nonce counts, or just incrementing (off - for use when
833 user agents generate nonce counts that occasionally miss 1
834 (ie, 1,2,4,6)). Default off.
836 "check_nonce_count" on|off
837 This directive if set to off can disable the nonce count check
838 completely to work around buggy digest qop implementations in
839 certain mainstream browser versions. Default on to check the
840 nonce count to protect from authentication replay attacks.
842 "post_workaround" on|off
843 This is a workaround to certain buggy browsers who send an
844 incorrect request digest in POST requests when reusing the
845 same nonce as acquired earlier on a GET request.
849 === Example Configuration ===
851 This configuration displays the recommended authentication scheme
852 order from most to least secure with recommended minimum configuration
853 settings for each scheme:
855 #auth_param negotiate program <uncomment and complete this line to activate>
856 #auth_param negotiate children 20 startup=0 idle=1
858 #auth_param digest program <uncomment and complete this line to activate>
859 #auth_param digest children 20 startup=0 idle=1
860 #auth_param digest realm Squid proxy-caching web server
861 #auth_param digest nonce_garbage_interval 5 minutes
862 #auth_param digest nonce_max_duration 30 minutes
863 #auth_param digest nonce_max_count 50
865 #auth_param ntlm program <uncomment and complete this line to activate>
866 #auth_param ntlm children 20 startup=0 idle=1
868 #auth_param basic program <uncomment and complete this line>
869 #auth_param basic children 5 startup=5 idle=1
870 #auth_param basic credentialsttl 2 hours
873 NAME: authenticate_cache_garbage_interval
877 LOC: Auth::TheConfig.garbageCollectInterval
879 The time period between garbage collection across the username cache.
880 This is a trade-off between memory utilization (long intervals - say
881 2 days) and CPU (short intervals - say 1 minute). Only change if you
885 NAME: authenticate_ttl
889 LOC: Auth::TheConfig.credentialsTtl
891 The time a user & their credentials stay in the logged in
892 user cache since their last request. When the garbage
893 interval passes, all user credentials that have passed their
894 TTL are removed from memory.
897 NAME: authenticate_ip_ttl
900 LOC: Auth::TheConfig.ipTtl
903 If you use proxy authentication and the 'max_user_ip' ACL,
904 this directive controls how long Squid remembers the IP
905 addresses associated with each user. Use a small value
906 (e.g., 60 seconds) if your users might change addresses
907 quickly, as is the case with dialup. You might be safe
908 using a larger value (e.g., 2 hours) in a corporate LAN
909 environment with relatively static address assignments.
914 -----------------------------------------------------------------------------
917 NAME: external_acl_type
918 TYPE: externalAclHelper
919 LOC: Config.externalAclHelperList
922 This option defines external acl classes using a helper program
923 to look up the status
925 external_acl_type name [options] FORMAT /path/to/helper [helper arguments]
929 ttl=n TTL in seconds for cached results (defaults to 3600
933 TTL for cached negative lookups (default same
936 grace=n Percentage remaining of TTL where a refresh of a
937 cached entry should be initiated without needing to
938 wait for a new reply. (default is for no grace period)
940 cache=n The maximum number of entries in the result cache. The
941 default limit is 262144 entries. Each cache entry usually
942 consumes at least 256 bytes. Squid currently does not remove
943 expired cache entries until the limit is reached, so a proxy
944 will sooner or later reach the limit. The expanded FORMAT
945 value is used as the cache key, so if the details in FORMAT
946 are highly variable, a larger cache may be needed to produce
947 reduction in helper load.
950 Maximum number of acl helper processes spawned to service
951 external acl lookups of this type. (default 5)
954 Minimum number of acl helper processes to spawn during
955 startup and reconfigure to service external acl lookups
956 of this type. (default 0)
959 Number of acl helper processes to keep ahead of traffic
960 loads. Squid will spawn this many at once whenever load
961 rises above the capabilities of existing processes.
962 Up to the value of children-max. (default 1)
964 concurrency=n concurrency level per process. Only used with helpers
965 capable of processing more than one query at a time.
967 queue-size=N The queue-size option sets the maximum number of
968 queued requests. A request is queued when no existing
969 helper can accept it due to concurrency limit and no
970 new helper can be started due to children-max limit.
971 If the queued requests exceed queue size, the acl is
972 ignored. The default value is set to 2*children-max.
974 protocol=2.5 Compatibility mode for Squid-2.5 external acl helpers.
976 ipv4 / ipv6 IP protocol used to communicate with this helper.
977 The default is to auto-detect IPv6 and use it when available.
980 FORMAT is a series of %macro codes. See logformat directive for a full list
981 of the accepted codes. Although note that at the time of any external ACL
982 being tested data may not be available and thus some %macro expand to '-'.
984 In addition to the logformat codes; when processing external ACLs these
985 additional macros are made available:
987 %ACL The name of the ACL being tested.
989 %DATA The ACL arguments specified in the referencing config
990 'acl ... external' line, separated by spaces (an
991 "argument string"). see acl external.
993 If there are no ACL arguments %DATA expands to '-'.
995 If you do not specify a DATA macro inside FORMAT,
996 Squid automatically appends %DATA to your FORMAT.
997 Note that Squid-3.x may expand %DATA to whitespace
998 or nothing in this case.
1000 By default, Squid applies URL-encoding to each ACL
1001 argument inside the argument string. If an explicit
1002 encoding modifier is used (e.g., %#DATA), then Squid
1003 encodes the whole argument string as a single token
1004 (e.g., with %#DATA, spaces between arguments become
1007 If SSL is enabled, the following formatting codes become available:
1009 %USER_CERT SSL User certificate in PEM format
1010 %USER_CERTCHAIN SSL User certificate chain in PEM format
1011 %USER_CERT_xx SSL User certificate subject attribute xx
1012 %USER_CA_CERT_xx SSL User certificate issuer attribute xx
1015 NOTE: all other format codes accepted by older Squid versions
1019 General request syntax:
1021 [channel-ID] FORMAT-values
1024 FORMAT-values consists of transaction details expanded with
1025 whitespace separation per the config file FORMAT specification
1026 using the FORMAT macros listed above.
1028 Request values sent to the helper are URL escaped to protect
1029 each value in requests against whitespaces.
1031 If using protocol=2.5 then the request sent to the helper is not
1032 URL escaped to protect against whitespace.
1034 NOTE: protocol=3.0 is deprecated as no longer necessary.
1036 When using the concurrency= option the protocol is changed by
1037 introducing a query channel tag in front of the request/response.
1038 The query channel tag is a number between 0 and concurrency-1.
1039 This value must be echoed back unchanged to Squid as the first part
1040 of the response relating to its request.
1043 The helper receives lines expanded per the above format specification
1044 and for each input line returns 1 line starting with OK/ERR/BH result
1045 code and optionally followed by additional keywords with more details.
1048 General result syntax:
1050 [channel-ID] result keyword=value ...
1052 Result consists of one of the codes:
1055 the ACL test produced a match.
1058 the ACL test does not produce a match.
1061 An internal error occurred in the helper, preventing
1062 a result being identified.
1064 The meaning of 'a match' is determined by your squid.conf
1065 access control configuration. See the Squid wiki for details.
1069 user= The users name (login)
1071 password= The users password (for login= cache_peer option)
1073 message= Message describing the reason for this response.
1074 Available as %o in error pages.
1075 Useful on (ERR and BH results).
1077 tag= Apply a tag to a request. Only sets a tag once,
1078 does not alter existing tags.
1080 log= String to be logged in access.log. Available as
1081 %ea in logformat specifications.
1083 clt_conn_tag= Associates a TAG with the client TCP connection.
1084 Please see url_rewrite_program related documentation
1087 Any keywords may be sent on any response whether OK, ERR or BH.
1089 All response keyword values need to be a single token with URL
1090 escaping, or enclosed in double quotes (") and escaped using \ on
1091 any double quotes or \ characters within the value. The wrapping
1092 double quotes are removed before the value is interpreted by Squid.
1093 \r and \n are also replace by CR and LF.
1095 Some example key values:
1099 user="J. \"Bob\" Smith"
1104 LOC: Config.namedAcls
1106 DEFAULT: ssl::certHasExpired ssl_error X509_V_ERR_CERT_HAS_EXPIRED
1107 DEFAULT: ssl::certNotYetValid ssl_error X509_V_ERR_CERT_NOT_YET_VALID
1108 DEFAULT: ssl::certDomainMismatch ssl_error SQUID_X509_V_ERR_DOMAIN_MISMATCH
1109 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
1110 DEFAULT: ssl::certSelfSigned ssl_error X509_V_ERR_DEPTH_ZERO_SELF_SIGNED_CERT
1112 DEFAULT: all src all
1113 DEFAULT: manager url_regex +i ^[^:]+://[^/]+/squid-internal-mgr/
1114 DEFAULT: localhost src 127.0.0.1/32 ::1
1115 DEFAULT: to_localhost dst 127.0.0.0/8 0.0.0.0/32 ::1/128 ::/128
1116 DEFAULT: to_linklocal dst 169.254.0.0/16 fe80::/10
1117 DEFAULT: CONNECT method CONNECT
1118 DEFAULT_DOC: ACLs all, manager, localhost, to_localhost, to_linklocal, and CONNECT are predefined.
1120 Defining an Access List
1122 Every access list definition must begin with an aclname and acltype,
1123 followed by either type-specific arguments or a quoted filename that
1126 acl aclname acltype argument ...
1127 acl aclname acltype "file" ...
1129 When using "file", the file should contain one item per line.
1134 Some acl types supports options which changes their default behaviour:
1136 -i,+i By default, regular expressions are CASE-SENSITIVE. To make them
1137 case-insensitive, use the -i option. To return case-sensitive
1138 use the +i option between patterns, or make a new ACL line
1141 -n Disable lookups and address type conversions. If lookup or
1142 conversion is required because the parameter type (IP or
1143 domain name) does not match the message address type (domain
1144 name or IP), then the ACL would immediately declare a mismatch
1145 without any warnings or lookups.
1148 Perform a list membership test, interpreting values as
1149 comma-separated token lists and matching against individual
1150 tokens instead of whole values.
1151 The optional "delimiters" parameter specifies one or more
1152 alternative non-alphanumeric delimiter characters.
1153 non-alphanumeric delimiter characters.
1155 -- Used to stop processing all options, in the case the first acl
1156 value has '-' character as first character (for example the '-'
1157 is a valid domain name)
1159 Some acl types require suspending the current request in order
1160 to access some external data source.
1161 Those which do are marked with the tag [slow], those which
1162 don't are marked as [fast].
1163 See https://wiki.squid-cache.org/SquidFaq/SquidAcl
1164 for further information
1166 ***** ACL TYPES AVAILABLE *****
1168 acl aclname src ip-address/mask ... # clients IP address [fast]
1169 acl aclname src addr1-addr2/mask ... # range of addresses [fast]
1170 acl aclname dst [-n] ip-address/mask ... # URL host's IP address [slow]
1171 acl aclname localip ip-address/mask ... # IP address the client connected to [fast]
1174 acl aclname arp mac-address ...
1175 acl aclname eui64 eui64-address ...
1177 # MAC (EUI-48) and EUI-64 addresses use xx:xx:xx:xx:xx:xx notation.
1179 # The 'arp' ACL code is not portable to all operating systems.
1180 # It works on Linux, Solaris, Windows, FreeBSD, and some other
1183 # The eui_lookup directive is required to be 'on' (the default)
1184 # and Squid built with --enable-eui for MAC/EUI addresses to be
1185 # available for this ACL.
1187 # Squid can only determine the MAC/EUI address for IPv4
1188 # clients that are on the same subnet. If the client is on a
1189 # different subnet, then Squid cannot find out its address.
1191 # IPv6 protocol does not contain ARP. MAC/EUI is either
1192 # encoded directly in the IPv6 address or not available.
1194 acl aclname clientside_mark mark[/mask] ...
1195 # matches CONNMARK of an accepted connection [fast]
1196 # DEPRECATED. Use the 'client_connection_mark' instead.
1198 acl aclname client_connection_mark mark[/mask] ...
1199 # matches CONNMARK of an accepted connection [fast]
1201 # mark and mask are unsigned integers (hex, octal, or decimal).
1202 # If multiple marks are given, then the ACL matches if at least
1205 # Uses netfilter-conntrack library.
1206 # Requires building Squid with --enable-linux-netfilter.
1208 # The client, various intermediaries, and Squid itself may set
1209 # CONNMARK at various times. The last CONNMARK set wins. This ACL
1210 # checks the mark present on an accepted connection or set by
1211 # Squid afterwards, depending on the ACL check timing. This ACL
1212 # effectively ignores any mark set by other agents after Squid has
1213 # accepted the connection.
1215 acl aclname srcdomain .foo.com ...
1216 # reverse lookup, from client IP [slow]
1217 acl aclname dstdomain [-n] .foo.com ...
1218 # Destination server from URL [fast]
1219 acl aclname srcdom_regex [-i] \.foo\.com ...
1220 # POSIX extended regex matching client name [slow]
1221 acl aclname dstdom_regex [-n] [-i] \.foo\.com ...
1222 # POSIX extended regex matching server [fast]
1224 # For dstdomain and dstdom_regex a reverse lookup is tried if a IP
1225 # based URL is used and no match is found. The name "none" is used
1226 # if the reverse lookup fails.
1228 acl aclname peername myPeer ...
1229 acl aclname peername_regex [-i] regex-pattern ...
1231 # match against a named cache_peer entry
1232 # set unique name= on cache_peer lines for reliable use.
1234 acl aclname time [day-abbrevs] [h1:m1-h2:m2]
1244 # h1:m1 must be less than h2:m2
1246 acl aclname url_regex [-i] ^http:// ...
1247 # POSIX extended regex matching on whole URL [fast]
1249 # If request URL contains only valid pct-encoded triplets (RFC 3986),
1250 # all of them are decoded before matching (e.g., `%25` triplet is
1251 # replaced with a single `%` character). If request URL contains at
1252 # least one `%` character that does not start a valid pct-encoded
1253 # triplet (e.g., `%%`, `%X`, or `%2Y`), then the URL is not decoded at
1254 # all (i.e. the raw request URL is used for matching).
1256 # If a request URL is decoded as described above, then all request URL
1257 # characters starting with the decoded `%00` pct-encoded triplet (if
1258 # any) are ignored during matching. There is currently no way to match
1259 # that triplet itself in a correctly percent-encoded URL.
1261 # ACL parameters are not decoded.
1263 acl aclname urllogin [-i] [^a-zA-Z0-9] ...
1264 # POSIX extended regex matching on URL login field [fast]
1266 # This ACL does not match requests with a URL that lacks a login field.
1268 # This ACL handles RFC 3986 pct-encoded triplets in the login field as
1269 # url_regex ACL handles those triplets in the entire request URL.
1271 acl aclname urlpath_regex [-i] \.gif$ ...
1272 # POSIX extended regex matching on URL path [fast]
1274 acl aclname port 80 70 21 0-1024 ...
1275 # destination TCP port (or port range) of the request [fast]
1277 # Port 0 matches requests that have no explicit and no default destination
1278 # ports (e.g., HTTP requests with URN targets)
1280 acl aclname localport 3128 ... # TCP port the client connected to [fast]
1281 # NP: for interception mode this is usually '80'
1283 acl aclname myportname 3128 ... # *_port name [fast]
1285 acl aclname proto HTTP FTP ... # request protocol [fast]
1287 acl aclname method GET POST ... # HTTP request method [fast]
1289 acl aclname http_status 200 301 500- 400-403 ...
1290 # status code in reply [fast]
1292 acl aclname browser [-i] regex ...
1293 # POSIX extended regex match on User-Agent header
1294 # (see also req_header below) [fast]
1296 acl aclname referer_regex [-i] regex ...
1297 # POSIX extended regex match on Referer header [fast]
1298 # Referer is highly unreliable, so use with care
1300 acl aclname proxy_auth [-i] username ...
1301 # perform http authentication challenge to the client and match against
1302 # supplied credentials [slow]
1304 # takes a list of allowed usernames.
1305 # use REQUIRED to accept any valid username.
1307 # See proxy_auth_regex for more information. The two ACLs differ only in
1308 # their parameter syntax and username matching algorithm.
1310 acl aclname proxy_auth_regex [-i] username_pattern ...
1311 # perform http authentication challenge to the client and
1312 # POSIX extended regex match on supplied username [slow]
1314 # Will use proxy authentication in forward-proxy scenarios, and plain
1315 # http authentication in reverse-proxy scenarios
1317 # NOTE: when a Proxy-Authentication header is sent but it is not
1318 # needed during ACL checking the username is NOT logged
1321 # NOTE: proxy_auth requires a EXTERNAL authentication program
1322 # to check username/password combinations (see
1323 # auth_param directive).
1325 # NOTE: proxy_auth can't be used in a transparent/intercepting proxy
1326 # as the browser needs to be configured for using a proxy in order
1327 # to respond to proxy authentication.
1329 acl aclname snmp_community string ...
1330 # A community string to limit access to your SNMP Agent [fast]
1333 # acl snmppublic snmp_community public
1335 acl aclname maxconn number
1336 # This will be matched when the client's IP address has
1337 # more than <number> TCP connections established. [fast]
1338 # NOTE: This only measures direct TCP links so X-Forwarded-For
1339 # indirect clients are not counted.
1341 acl aclname max_user_ip [-s] number
1342 # This will be matched when the user attempts to log in from more
1343 # than <number> different ip addresses. The authenticate_ip_ttl
1344 # parameter controls the timeout on the ip entries. [fast]
1345 # If -s is specified the limit is strict, denying browsing
1346 # from any further IP addresses until the ttl has expired. Without
1347 # -s Squid will just annoy the user by "randomly" denying requests.
1348 # (the counter is reset each time the limit is reached and a
1349 # request is denied)
1350 # NOTE: in acceleration mode or where there is mesh of child proxies,
1351 # clients may appear to come from multiple addresses if they are
1352 # going through proxy farms, so a limit of 1 may cause user problems.
1354 acl aclname random probability
1355 # Pseudo-randomly match requests. Based on the probability given.
1356 # Probability may be written as a decimal (0.333), fraction (1/3)
1357 # or ratio of matches:non-matches (3:5).
1359 acl aclname req_mime_type [-i] mime-type ...
1360 # POSIX extended regex match against the mime type of the request generated
1361 # by the client. Can be used to detect file upload or some
1362 # types HTTP tunneling requests [fast]
1363 # NOTE: This does NOT match the reply. You cannot use this
1364 # to match the returned file type.
1366 acl aclname req_header header-name [-i] any\.regex\.here
1367 # POSIX extended regex match against any of the known request headers. May be
1368 # thought of as a superset of "browser", "referer" and "mime-type"
1371 acl aclname rep_mime_type [-i] mime-type ...
1372 # POSIX extended regex match against the mime type of the reply received by
1373 # squid. Can be used to detect file download or some
1374 # types HTTP tunneling requests. [fast]
1375 # NOTE: This has no effect in http_access rules. It only has
1376 # effect in rules that affect the reply data stream such as
1377 # http_reply_access.
1379 acl aclname rep_header header-name [-i] any\.regex\.here
1380 # POSIX extended regex match against any of the known reply headers. May be
1381 # thought of as a superset of "browser", "referer" and "mime-type"
1384 acl aclname external class_name [arguments...]
1385 # external ACL lookup via a helper class defined by the
1386 # external_acl_type directive [slow]
1388 acl aclname user_cert attribute values...
1389 # match against attributes in a user SSL certificate
1390 # attribute is one of DN/C/O/CN/L/ST or a numerical OID [fast]
1392 acl aclname ca_cert attribute values...
1393 # match against attributes a users issuing CA SSL certificate
1394 # attribute is one of DN/C/O/CN/L/ST or a numerical OID [fast]
1396 acl aclname ext_user [-i] username ...
1397 # string match on username returned by external acl helper [slow]
1398 # use REQUIRED to accept any non-null user name.
1400 # See also: ext_user_regex. The two ACLs differ only in their parameter
1401 # syntax and username matching algorithm.
1403 acl aclname ext_user_regex [-i] username_pattern ...
1404 # POSIX extended regex match on username returned by external acl helper [slow]
1406 acl aclname tag tagvalue ...
1407 # string match on tag returned by external acl helper [fast]
1408 # DEPRECATED. Only the first tag will match with this ACL.
1409 # Use the 'note' ACL instead for handling multiple tag values.
1411 acl aclname hier_code codename ...
1412 # string match against squid hierarchy code(s); [fast]
1413 # e.g., DIRECT, PARENT_HIT, NONE, etc.
1415 # NOTE: This has no effect in http_access rules. It only has
1416 # effect in rules that affect the reply data stream such as
1417 # http_reply_access.
1419 acl aclname note [-m[=delimiters]] name [value ...]
1420 # match transaction annotation [fast]
1421 # Without values, matches any annotation with a given name.
1422 # With value(s), matches any annotation with a given name that
1423 # also has one of the given values.
1424 # If the -m flag is used, then the value of the named
1425 # annotation is interpreted as a list of tokens, and the ACL
1426 # matches individual name=token pairs rather than whole
1427 # name=value pairs. See "ACL Options" above for more info.
1428 # Annotation sources include note and adaptation_meta directives
1429 # as well as helper and eCAP responses.
1431 acl aclname annotate_transaction [-m[=delimiters]] key=value ...
1432 acl aclname annotate_transaction [-m[=delimiters]] key+=value ...
1433 # Always matches. [fast]
1434 # Used for its side effect: This ACL immediately adds a
1435 # key=value annotation to the current master transaction.
1436 # The added annotation can then be tested using note ACL and
1437 # logged (or sent to helpers) using %note format code.
1439 # Annotations can be specified using replacement and addition
1440 # formats. The key=value form replaces old same-key annotation
1441 # value(s). The key+=value form appends a new value to the old
1442 # same-key annotation. Both forms create a new key=value
1443 # annotation if no same-key annotation exists already. If
1444 # -m flag is used, then the value is interpreted as a list
1445 # and the annotation will contain key=token pair(s) instead of the
1446 # whole key=value pair.
1448 # This ACL is especially useful for recording complex multi-step
1449 # ACL-driven decisions. For example, the following configuration
1450 # avoids logging transactions accepted after aclX matched:
1452 # # First, mark transactions accepted after aclX matched
1453 # acl markSpecial annotate_transaction special=true
1454 # http_access allow acl001
1456 # http_access deny acl100
1457 # http_access allow aclX markSpecial
1459 # # Second, do not log marked transactions:
1460 # acl markedSpecial note special true
1461 # access_log ... deny markedSpecial
1463 # # Note that the following would not have worked because aclX
1464 # # alone does not determine whether the transaction was allowed:
1465 # access_log ... deny aclX # Wrong!
1467 # Warning: This ACL annotates the transaction even when negated
1468 # and even if subsequent ACLs fail to match. For example, the
1469 # following three rules will have exactly the same effect as far
1470 # as annotations set by the "mark" ACL are concerned:
1472 # some_directive acl1 ... mark # rule matches if mark is reached
1473 # some_directive acl1 ... !mark # rule never matches
1474 # some_directive acl1 ... mark !all # rule never matches
1476 acl aclname annotate_client [-m[=delimiters]] key=value ...
1477 acl aclname annotate_client [-m[=delimiters]] key+=value ...
1479 # Always matches. [fast]
1480 # Used for its side effect: This ACL immediately adds a
1481 # key=value annotation to the current client-to-Squid
1482 # connection. Connection annotations are propagated to the current
1483 # and all future master transactions on the annotated connection.
1484 # See the annotate_transaction ACL for details.
1486 # For example, the following configuration avoids rewriting URLs
1487 # of transactions bumped by SslBump:
1489 # # First, mark bumped connections:
1490 # acl markBumped annotate_client bumped=true
1491 # ssl_bump peek acl1
1492 # ssl_bump stare acl2
1493 # ssl_bump bump acl3 markBumped
1494 # ssl_bump splice all
1496 # # Second, do not send marked transactions to the redirector:
1497 # acl markedBumped note bumped true
1498 # url_rewrite_access deny markedBumped
1500 # # Note that the following would not have worked because acl3 alone
1501 # # does not determine whether the connection is going to be bumped:
1502 # url_rewrite_access deny acl3 # Wrong!
1504 acl aclname adaptation_service service ...
1505 # Matches the name of any icap_service, ecap_service,
1506 # adaptation_service_set, or adaptation_service_chain that Squid
1507 # has used (or attempted to use) for the master transaction.
1508 # This ACL must be defined after the corresponding adaptation
1509 # service is named in squid.conf. This ACL is usable with
1510 # adaptation_meta because it starts matching immediately after
1511 # the service has been selected for adaptation.
1513 acl aclname transaction_initiator initiator ...
1514 # Matches transaction's initiator [fast]
1516 # Supported initiators are:
1517 # certificate-fetching: matches transactions fetching
1518 # a missing intermediate TLS certificate
1519 # cache-digest: matches transactions fetching Cache Digests
1521 # htcp: matches HTCP requests from peers
1522 # icp: matches ICP requests to peers
1523 # icmp: matches ICMP RTT database (NetDB) requests to peers
1524 # internal: matches any of the above
1525 # client: matches transactions containing an HTTP or FTP
1526 # client request received at a Squid *_port
1527 # all: matches any transaction, including internal transactions
1528 # without a configurable initiator and hopefully rare
1529 # transactions without a known-to-Squid initiator
1531 # Multiple initiators are ORed.
1533 acl aclname has component
1534 # matches a transaction "component" [fast]
1536 # Supported transaction components are:
1537 # request: transaction has a request header (at least)
1538 # response: transaction has a response header (at least)
1539 # ALE: transaction has an internally-generated Access Log Entry
1540 # structure; bugs notwithstanding, all transaction have it
1542 # For example, the following configuration helps when dealing with HTTP
1543 # clients that close connections without sending a request header:
1545 # acl hasRequest has request
1546 # acl logMe note important_transaction
1547 # # avoid "logMe ACL is used in context without an HTTP request" warnings
1548 # access_log ... logformat=detailed hasRequest logMe
1549 # # log request-less transactions, instead of ignoring them
1550 # access_log ... logformat=brief !hasRequest
1552 # Multiple components are not supported for one "acl" rule, but
1553 # can be specified (and are ORed) using multiple same-name rules:
1555 # # OK, this strange logging daemon needs request or response,
1556 # # but can work without either a request or a response:
1557 # acl hasWhatMyLoggingDaemonNeeds has request
1558 # acl hasWhatMyLoggingDaemonNeeds has response
1560 acl aclname at_step step
1561 # match against the current request processing step [fast]
1563 # GeneratingCONNECT: Generating HTTP CONNECT request headers
1565 # The following ssl_bump processing steps are recognized:
1566 # SslBump1: After getting TCP-level and HTTP CONNECT info.
1567 # SslBump2: After getting SSL Client Hello info.
1568 # SslBump3: After getting SSL Server Hello info.
1572 acl aclname ssl_error errorname
1573 # match against SSL certificate validation error [fast]
1575 # When used with sslproxy_cert_error, this ACL tests a single
1576 # certificate validation error currently being evaluated by that
1577 # directive. When used with slproxy_cert_sign or sslproxy_cert_adapt,
1578 # the ACL tests all past certificate validation errors associated with
1579 # the current Squid-to-server connection (attempt). This ACL is not yet
1580 # supported for use with other directives.
1582 # For valid error names see in @DEFAULT_ERROR_DIR@/templates/error-details.txt
1585 # The following can be used as shortcuts for certificate properties:
1586 # [ssl::]certHasExpired: the "not after" field is in the past
1587 # [ssl::]certNotYetValid: the "not before" field is in the future
1588 # [ssl::]certUntrusted: The certificate issuer is not to be trusted.
1589 # [ssl::]certSelfSigned: The certificate is self signed.
1590 # [ssl::]certDomainMismatch: The certificate CN domain does not
1591 # match the name the name of the host we are connecting to.
1593 # The ssl::certHasExpired, ssl::certNotYetValid, ssl::certDomainMismatch,
1594 # ssl::certUntrusted, and ssl::certSelfSigned can also be used as
1595 # predefined ACLs, just like the 'all' ACL.
1597 acl aclname server_cert_fingerprint fingerprint
1598 # match against server SSL certificate fingerprint [fast]
1600 # The fingerprint is the digest of the DER encoded version
1601 # of the whole certificate. The user should use the form: XX:XX:...
1602 # The SHA1 digest algorithm is the default and is currently
1603 # the only algorithm supported.
1605 acl aclname ssl::server_name [option] .foo.com ...
1606 # matches server name obtained from various sources [fast]
1608 # The ACL computes server name(s) using such information sources as
1609 # CONNECT request URI, TLS client SNI, and TLS server certificate
1610 # subject (CN and SubjectAltName). The computed server name(s) usually
1611 # change with each SslBump step, as more info becomes available:
1612 # * SNI is used as the server name instead of the request URI,
1613 # * subject name(s) from the server certificate (CN and
1614 # SubjectAltName) are used as the server names instead of SNI.
1616 # When the ACL computes multiple server names, matching any single
1617 # computed name is sufficient for the ACL to match.
1619 # The "none" name can be used to match transactions where the ACL
1620 # could not compute the server name using any information source
1621 # that was both available and allowed to be used by the ACL options at
1622 # the ACL evaluation time.
1624 # Unlike dstdomain, this ACL does not perform DNS lookups.
1626 # A server name may be an IP address. For example, subject alternative
1627 # names (a.k.a. SANs) in some real server certificates include IPv4 and
1628 # IPv6 entries. Internally, Squid uses inet_ntop(3) to prep IP names for
1629 # matching. When using IPv6 names, use "::" notation (if applicable).
1630 # Do not use brackets. For example: 1080::8:800:200c:417a.
1632 # An ACL option below may be used to restrict what information
1633 # sources are used to extract the server names from:
1635 # --client-requested
1636 # The server name is SNI regardless of what the server says.
1638 # The server name(s) are the certificate subject name(s), regardless
1639 # of what the client has requested. If the server certificate is
1640 # unavailable, then the name is "none".
1642 # The server name is either SNI (if SNI matches at least one of the
1643 # certificate subject names) or "none" (otherwise). When the server
1644 # certificate is unavailable, the consensus server name is SNI.
1646 # Combining multiple options in one ACL is a fatal configuration
1649 # For all options: If no SNI is available, then the CONNECT request
1650 # target (a.k.a. URI) is used instead of SNI (for an intercepted
1651 # connection, this target is the destination IP address).
1653 acl aclname ssl::server_name_regex [-i] \.foo\.com ...
1654 # POSIX extended regex matches server name obtained from various sources [fast]
1656 # See ssl::server_name for details, including IPv6 address formatting
1657 # caveats. Use case-insensitive matching (i.e. -i option) to reduce
1658 # dependency on how Squid formats or sanitizes server names.
1660 acl aclname connections_encrypted
1661 # matches transactions with all HTTP messages received over TLS
1662 # transport connections. [fast]
1664 # The master transaction deals with HTTP messages received from
1665 # various sources. All sources used by the master transaction in the
1666 # past are considered by the ACL. The following rules define whether
1667 # a given message source taints the entire master transaction,
1668 # resulting in ACL mismatches:
1670 # * The HTTP client transport connection is not TLS.
1671 # * An adaptation service connection-encryption flag is off.
1672 # * The peer or origin server transport connection is not TLS.
1674 # Caching currently does not affect these rules. This cache ignorance
1675 # implies that only the current HTTP client transport and REQMOD
1676 # services status determine whether this ACL matches a from-cache
1677 # transaction. The source of the cached response does not have any
1678 # effect on future transaction that use the cached response without
1679 # revalidation. This may change.
1681 # DNS, ICP, and HTCP exchanges during the master transaction do not
1682 # affect these rules.
1684 acl aclname any-of acl1 acl2 ...
1685 # match any one of the acls [fast or slow]
1686 # The first matching ACL stops further ACL evaluation.
1688 # ACLs from multiple any-of lines with the same name are ORed.
1689 # For example, A = (a1 or a2) or (a3 or a4) can be written as
1690 # acl A any-of a1 a2
1691 # acl A any-of a3 a4
1693 # This group ACL is fast if all evaluated ACLs in the group are fast
1694 # and slow otherwise.
1696 acl aclname all-of acl1 acl2 ...
1697 # match all of the acls [fast or slow]
1698 # The first mismatching ACL stops further ACL evaluation.
1700 # ACLs from multiple all-of lines with the same name are ORed.
1701 # For example, B = (b1 and b2) or (b3 and b4) can be written as
1702 # acl B all-of b1 b2
1703 # acl B all-of b3 b4
1705 # This group ACL is fast if all evaluated ACLs in the group are fast
1706 # and slow otherwise.
1709 acl macaddress arp 09:00:2b:23:45:67
1710 acl password proxy_auth REQUIRED
1711 acl fileupload req_mime_type -i ^multipart/form-data$
1712 acl javascript rep_mime_type -i ^application/x-javascript$
1716 # Recommended minimum configuration:
1719 # Example rule allowing access from your local networks.
1720 # Adapt to list your (internal) IP networks from where browsing
1722 acl localnet src 0.0.0.1-0.255.255.255 # RFC 1122 "this" network (LAN)
1723 acl localnet src 10.0.0.0/8 # RFC 1918 local private network (LAN)
1724 acl localnet src 100.64.0.0/10 # RFC 6598 shared address space (CGN)
1725 acl localnet src 169.254.0.0/16 # RFC 3927 link-local (directly plugged) machines
1726 acl localnet src 172.16.0.0/12 # RFC 1918 local private network (LAN)
1727 acl localnet src 192.168.0.0/16 # RFC 1918 local private network (LAN)
1728 acl localnet src fc00::/7 # RFC 4193 local private network range
1729 acl localnet src fe80::/10 # RFC 4291 link-local (directly plugged) machines
1731 acl SSL_ports port 443
1732 acl Safe_ports port 80 # http
1733 acl Safe_ports port 21 # ftp
1734 acl Safe_ports port 443 # https
1735 acl Safe_ports port 70 # gopher
1736 acl Safe_ports port 210 # wais
1737 acl Safe_ports port 1025-65535 # unregistered ports
1738 acl Safe_ports port 280 # http-mgmt
1739 acl Safe_ports port 488 # gss-http
1740 acl Safe_ports port 591 # filemaker
1741 acl Safe_ports port 777 # multiling http
1745 NAME: proxy_protocol_access
1747 LOC: Config.accessList.proxyProtocol
1749 DEFAULT_DOC: all TCP connections to ports with require-proxy-header will be denied
1751 Determine which client proxies can be trusted to provide correct
1752 information regarding real client IP address using PROXY protocol.
1754 Requests may pass through a chain of several other proxies
1755 before reaching us. The original source details may by sent in:
1756 * HTTP message Forwarded header, or
1757 * HTTP message X-Forwarded-For header, or
1758 * PROXY protocol connection header.
1760 This directive is solely for validating new PROXY protocol
1761 connections received from a port flagged with require-proxy-header.
1762 It is checked only once after TCP connection setup.
1764 A deny match results in TCP connection closure.
1766 An allow match is required for Squid to permit the corresponding
1767 TCP connection, before Squid even looks for HTTP request headers.
1768 If there is an allow match, Squid starts using PROXY header information
1769 to determine the source address of the connection for all future ACL
1770 checks, logging, etc.
1772 SECURITY CONSIDERATIONS:
1774 Any host from which we accept client IP details can place
1775 incorrect information in the relevant header, and Squid
1776 will use the incorrect information as if it were the
1777 source address of the request. This may enable remote
1778 hosts to bypass any access control restrictions that are
1779 based on the client's source addresses.
1781 This clause only supports fast acl types.
1782 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1785 NAME: follow_x_forwarded_for
1787 IFDEF: FOLLOW_X_FORWARDED_FOR
1788 LOC: Config.accessList.followXFF
1789 DEFAULT_IF_NONE: deny all
1790 DEFAULT_DOC: X-Forwarded-For header will be ignored.
1792 Determine which client proxies can be trusted to provide correct
1793 information regarding real client IP address.
1795 Requests may pass through a chain of several other proxies
1796 before reaching us. The original source details may by sent in:
1797 * HTTP message Forwarded header, or
1798 * HTTP message X-Forwarded-For header, or
1799 * PROXY protocol connection header.
1801 PROXY protocol connections are controlled by the proxy_protocol_access
1802 directive which is checked before this.
1804 If a request reaches us from a source that is allowed by this
1805 directive, then we trust the information it provides regarding
1806 the IP of the client it received from (if any).
1808 For the purpose of ACLs used in this directive the src ACL type always
1809 matches the address we are testing and srcdomain matches its rDNS.
1811 On each HTTP request Squid checks for X-Forwarded-For header fields.
1812 If found the header values are iterated in reverse order and an allow
1813 match is required for Squid to continue on to the next value.
1814 The verification ends when a value receives a deny match, cannot be
1815 tested, or there are no more values to test.
1816 NOTE: Squid does not yet follow the Forwarded HTTP header.
1818 The end result of this process is an IP address that we will
1819 refer to as the indirect client address. This address may
1820 be treated as the client address for access control, ICAP, delay
1821 pools and logging, depending on the acl_uses_indirect_client,
1822 icap_uses_indirect_client, delay_pool_uses_indirect_client,
1823 log_uses_indirect_client and tproxy_uses_indirect_client options.
1825 This clause only supports fast acl types.
1826 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1828 SECURITY CONSIDERATIONS:
1830 Any host from which we accept client IP details can place
1831 incorrect information in the relevant header, and Squid
1832 will use the incorrect information as if it were the
1833 source address of the request. This may enable remote
1834 hosts to bypass any access control restrictions that are
1835 based on the client's source addresses.
1839 acl localhost src 127.0.0.1
1840 acl my_other_proxy srcdomain .proxy.example.com
1841 follow_x_forwarded_for allow localhost
1842 follow_x_forwarded_for allow my_other_proxy
1845 NAME: acl_uses_indirect_client
1848 IFDEF: FOLLOW_X_FORWARDED_FOR
1850 LOC: Config.onoff.acl_uses_indirect_client
1852 Controls whether the indirect client address
1853 (see follow_x_forwarded_for) is used instead of the
1854 direct client address in acl matching.
1856 NOTE: maxconn ACL considers direct TCP links and indirect
1857 clients will always have zero. So no match.
1860 NAME: delay_pool_uses_indirect_client
1863 IFDEF: FOLLOW_X_FORWARDED_FOR&&USE_DELAY_POOLS
1865 LOC: Config.onoff.delay_pool_uses_indirect_client
1867 Controls whether the indirect client address
1868 (see follow_x_forwarded_for) is used instead of the
1869 direct client address in delay pools.
1872 NAME: log_uses_indirect_client
1875 IFDEF: FOLLOW_X_FORWARDED_FOR
1877 LOC: Config.onoff.log_uses_indirect_client
1879 Controls whether the indirect client address
1880 (see follow_x_forwarded_for) is used instead of the
1881 direct client address in the access log.
1884 NAME: tproxy_uses_indirect_client
1887 IFDEF: FOLLOW_X_FORWARDED_FOR&&LINUX_NETFILTER
1889 LOC: Config.onoff.tproxy_uses_indirect_client
1891 Controls whether the indirect client address
1892 (see follow_x_forwarded_for) is used instead of the
1893 direct client address when spoofing the outgoing client.
1895 This has no effect on requests arriving in non-tproxy
1898 SECURITY WARNING: Usage of this option is dangerous
1899 and should not be used trivially. Correct configuration
1900 of follow_x_forwarded_for with a limited set of trusted
1901 sources is required to prevent abuse of your proxy.
1904 NAME: spoof_client_ip
1906 LOC: Config.accessList.spoof_client_ip
1908 DEFAULT_DOC: Allow spoofing on all TPROXY traffic.
1910 Control client IP address spoofing of TPROXY traffic based on
1911 defined access lists.
1913 spoof_client_ip allow|deny [!]aclname ...
1915 If there are no "spoof_client_ip" lines present, the default
1916 is to "allow" spoofing of any suitable request.
1918 Note that the cache_peer "no-tproxy" option overrides this ACL.
1920 This clause supports fast acl types.
1921 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1926 LOC: Config.accessList.http
1927 DEFAULT_IF_NONE: deny all
1928 DEFAULT_DOC: Deny, unless rules exist in squid.conf.
1930 Allowing or Denying access based on defined access lists
1932 To allow or deny a message received on an HTTP, HTTPS, or FTP port:
1933 http_access allow|deny [!]aclname ...
1935 NOTE on default values:
1937 If there are no "access" lines present, the default is to deny
1940 If none of the "access" lines cause a match, the default is the
1941 opposite of the last line in the list. If the last line was
1942 deny, the default is allow. Conversely, if the last line
1943 is allow, the default will be deny. For these reasons, it is a
1944 good idea to have an "deny all" entry at the end of your access
1945 lists to avoid potential confusion.
1947 This clause supports both fast and slow acl types.
1948 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1953 # Recommended minimum Access Permission configuration:
1955 # Deny requests to certain unsafe ports
1956 http_access deny !Safe_ports
1958 # Deny CONNECT to other than secure SSL ports
1959 http_access deny CONNECT !SSL_ports
1961 # Only allow cachemgr access from localhost
1962 http_access allow localhost manager
1963 http_access deny manager
1965 # This default configuration only allows localhost requests because a more
1966 # permissive Squid installation could introduce new attack vectors into the
1967 # network by proxying external TCP connections to unprotected services.
1968 http_access allow localhost
1970 # The two deny rules below are unnecessary in this default configuration
1971 # because they are followed by a "deny all" rule. However, they may become
1972 # critically important when you start allowing external requests below them.
1974 # Protect web applications running on the same server as Squid. They often
1975 # assume that only local users can access them at "localhost" ports.
1976 http_access deny to_localhost
1978 # Protect cloud servers that provide local users with sensitive info about
1979 # their server via certain well-known link-local (a.k.a. APIPA) addresses.
1980 http_access deny to_linklocal
1983 # INSERT YOUR OWN RULE(S) HERE TO ALLOW ACCESS FROM YOUR CLIENTS
1986 # For example, to allow access from your local networks, you may uncomment the
1987 # following rule (and/or add rules that match your definition of "local"):
1988 # http_access allow localnet
1990 # And finally deny all other access to this proxy
1991 http_access deny all
1995 NAME: adapted_http_access http_access2
1997 LOC: Config.accessList.adapted_http
1999 DEFAULT_DOC: Allow, unless rules exist in squid.conf.
2001 Allowing or Denying access based on defined access lists
2003 Essentially identical to http_access, but runs after redirectors
2004 and ICAP/eCAP adaptation. Allowing access control based on their
2007 If not set then only http_access is used.
2010 NAME: http_reply_access
2012 LOC: Config.accessList.reply
2014 DEFAULT_DOC: Allow, unless rules exist in squid.conf.
2016 Allow replies to client requests. This is complementary to http_access.
2018 http_reply_access allow|deny [!] aclname ...
2020 NOTE: if there are no access lines present, the default is to allow
2023 If none of the access lines cause a match the opposite of the
2024 last line will apply. Thus it is good practice to end the rules
2025 with an "allow all" or "deny all" entry.
2027 This clause supports both fast and slow acl types.
2028 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
2033 LOC: Config.accessList.icp
2035 DEFAULT_DOC: Deny, unless rules exist in squid.conf.
2037 Allowing or Denying access to the ICP port based on defined
2040 icp_access allow|deny [!]aclname ...
2042 NOTE: The default if no icp_access lines are present is to
2043 deny all traffic. This default may cause problems with peers
2046 This clause only supports fast acl types.
2047 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
2049 # Allow ICP queries from local networks only
2050 #icp_access allow localnet
2051 #icp_access deny all
2057 LOC: Config.accessList.htcp
2059 DEFAULT_DOC: Deny, unless rules exist in squid.conf.
2061 Controls whether HTCP TST requests received on htcp_port are allowed.
2063 htcp_access allow|deny [!]aclname ...
2065 This directive does not control whether HTCP CLR requests are allowed.
2066 Use htcp_clr_access directive for that.
2068 This directive does not control whether HTCP requests with other opcodes
2069 are allowed (e.g., NOP, MON, and SET). Squid ignores those HTCP requests.
2071 NOTE: The default if no htcp_access lines are present is to
2072 deny all HTCP TST traffic. This default may cause problems with peers
2073 using the htcp option.
2075 This clause only supports fast acl types.
2076 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
2078 # Allow HTCP TST queries from local networks only
2079 #htcp_access allow localnet
2080 #htcp_access deny all
2083 NAME: htcp_clr_access
2086 LOC: Config.accessList.htcp_clr
2088 DEFAULT_DOC: Deny, unless rules exist in squid.conf.
2090 Controls whether HTCP CLR requests received on htcp_port are allowed.
2091 See htcp_access for controlling other HTCP messages.
2093 htcp_clr_access allow|deny [!]aclname ...
2095 HTCP CLR requests purge matching cached entries. They may be forwarded to
2096 specially marked cache_peers (see cache_peer HTCP options for details).
2098 This clause only supports fast acl types.
2099 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
2101 # Allow HTCP CLR requests from trusted peers
2102 acl htcp_clr_peer src 192.0.2.2 2001:DB8::2
2103 htcp_clr_access allow htcp_clr_peer
2104 htcp_clr_access deny all
2109 LOC: Config.accessList.miss
2111 DEFAULT_DOC: Allow, unless rules exist in squid.conf.
2113 Determines whether network access is permitted when satisfying a request.
2116 to force your neighbors to use you as a sibling instead of
2119 acl localclients src 192.0.2.0/24 2001:DB8::a:0/64
2120 miss_access deny !localclients
2121 miss_access allow all
2123 This means only your local clients are allowed to fetch relayed/MISS
2124 replies from the network and all other clients can only fetch cached
2127 The default for this setting allows all clients who passed the
2128 http_access rules to relay via this proxy.
2130 This clause only supports fast acl types.
2131 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
2134 NAME: reply_body_max_size
2135 COMMENT: size [acl acl...]
2138 DEFAULT_DOC: No limit is applied.
2139 LOC: Config.ReplyBodySize
2141 This option specifies the maximum size of a reply body. It can be
2142 used to prevent users from downloading very large files, such as
2143 MP3's and movies. When the reply headers are received, the
2144 reply_body_max_size lines are processed, and the first line where
2145 all (if any) listed ACLs are true is used as the maximum body size
2148 This size is checked twice. First when we get the reply headers,
2149 we check the content-length value. If the content length value exists
2150 and is larger than the allowed size, the request is denied and the
2151 user receives an error message that says "the request or reply
2152 is too large." If there is no content-length, and the reply
2153 size exceeds this limit, the client's connection is just closed
2154 and they will receive a partial reply.
2156 WARNING: downstream caches probably can not detect a partial reply
2157 if there is no content-length header, so they will cache
2158 partial responses and give them out as hits. You should NOT
2159 use this option if you have downstream caches.
2161 WARNING: A maximum size smaller than the size of squid's error messages
2162 will cause an infinite loop and crash squid. Ensure that the smallest
2163 non-zero value you use is greater that the maximum header size plus
2164 the size of your largest error page.
2166 If you set this parameter none (the default), there will be
2169 Configuration Format is:
2170 reply_body_max_size SIZE UNITS [acl ...]
2172 reply_body_max_size 10 MB
2176 NAME: on_unsupported_protocol
2177 TYPE: on_unsupported_protocol
2178 LOC: Config.accessList.on_unsupported_protocol
2180 DEFAULT_DOC: Respond with an error message to unidentifiable traffic
2182 Determines Squid behavior when encountering strange requests at the
2183 beginning of an accepted TCP connection or the beginning of a bumped
2184 CONNECT tunnel. Controlling Squid reaction to unexpected traffic is
2185 especially useful in interception environments where Squid is likely
2186 to see connections for unsupported protocols that Squid should either
2187 terminate or tunnel at TCP level.
2189 on_unsupported_protocol <action> [!]acl ...
2191 The first matching action wins. Only fast ACLs are supported.
2193 Supported actions are:
2195 tunnel: Establish a TCP connection with the intended server and
2196 blindly shovel TCP packets between the client and server.
2198 respond: Respond with an error message, using the transfer protocol
2199 for the Squid port that received the request (e.g., HTTP
2200 for connections intercepted at the http_port). This is the
2203 Squid expects the following traffic patterns:
2205 http_port: a plain HTTP request
2206 https_port: SSL/TLS handshake followed by an [encrypted] HTTP request
2207 ftp_port: a plain FTP command (no on_unsupported_protocol support yet!)
2208 CONNECT tunnel on http_port: same as https_port
2209 CONNECT tunnel on https_port: same as https_port
2211 Currently, this directive has effect on intercepted connections and
2212 bumped tunnels only. Other cases are not supported because Squid
2213 cannot know the intended destination of other traffic.
2216 # define what Squid errors indicate receiving non-HTTP traffic:
2217 acl foreignProtocol squid_error ERR_PROTOCOL_UNKNOWN ERR_TOO_BIG
2218 # define what Squid errors indicate receiving nothing:
2219 acl serverTalksFirstProtocol squid_error ERR_REQUEST_START_TIMEOUT
2220 # tunnel everything that does not look like HTTP:
2221 on_unsupported_protocol tunnel foreignProtocol
2222 # tunnel if we think the client waits for the server to talk first:
2223 on_unsupported_protocol tunnel serverTalksFirstProtocol
2224 # in all other error cases, just send an HTTP "error page" response:
2225 on_unsupported_protocol respond all
2227 See also: squid_error ACL
2233 LOC: Auth::TheConfig.schemeAccess
2235 DEFAULT_DOC: use all auth_param schemes in their configuration order
2237 Use this directive to customize authentication schemes presence and
2238 order in Squid's Unauthorized and Authentication Required responses.
2240 auth_schemes scheme1,scheme2,... [!]aclname ...
2242 where schemeN is the name of one of the authentication schemes
2243 configured using auth_param directives. At least one scheme name is
2244 required. Multiple scheme names are separated by commas. Either
2245 avoid whitespace or quote the entire schemes list.
2247 A special "ALL" scheme name expands to all auth_param-configured
2248 schemes in their configuration order. This directive cannot be used
2249 to configure Squid to offer no authentication schemes at all.
2251 The first matching auth_schemes rule determines the schemes order
2252 for the current Authentication Required transaction. Note that the
2253 future response is not yet available during auth_schemes evaluation.
2255 If this directive is not used or none of its rules match, then Squid
2256 responds with all configured authentication schemes in the order of
2257 auth_param directives in the configuration file.
2259 This directive does not determine when authentication is used or
2260 how each authentication scheme authenticates clients.
2262 The following example sends basic and negotiate authentication
2263 schemes, in that order, when requesting authentication of HTTP
2264 requests matching the isIE ACL (not shown) while sending all
2265 auth_param schemes in their configuration order to other clients:
2267 auth_schemes basic,negotiate isIE
2268 auth_schemes ALL all # explicit default
2270 This directive supports fast ACLs only.
2272 See also: auth_param.
2277 -----------------------------------------------------------------------------
2280 NAME: http_port ascii_port
2285 Usage: port [mode] [options]
2286 hostname:port [mode] [options]
2287 1.2.3.4:port [mode] [options]
2289 The socket addresses where Squid will listen for HTTP client
2290 requests. You may specify multiple socket addresses.
2291 There are three forms: port alone, hostname with port, and
2292 IP address with port. If you specify a hostname or IP
2293 address, Squid binds the socket to that specific
2294 address. Most likely, you do not need to bind to a specific
2295 address, so you can use the port number alone.
2297 If you are running Squid in accelerator mode, you
2298 probably want to listen on port 80 also, or instead.
2300 The -a command line option may be used to specify additional
2301 port(s) where Squid listens for proxy request. Such ports will
2302 be plain proxy ports with no options.
2304 You may specify multiple socket addresses on multiple lines.
2308 intercept Support for IP-Layer NAT interception delivering
2309 traffic to this Squid port.
2310 NP: disables authentication on the port.
2312 tproxy Support Linux TPROXY (or BSD divert-to) with spoofing
2313 of outgoing connections using the client IP address.
2314 NP: disables authentication on the port.
2316 accel Accelerator / reverse proxy mode
2318 ssl-bump For each CONNECT request allowed by ssl_bump ACLs,
2319 establish secure connection with the client and with
2320 the server, decrypt HTTPS messages as they pass through
2321 Squid, and treat them as unencrypted HTTP messages,
2322 becoming the man-in-the-middle.
2324 The ssl_bump option is required to fully enable
2325 bumping of CONNECT requests.
2327 Omitting the mode flag causes default forward proxy mode to be used.
2330 Accelerator Mode Options:
2332 defaultsite=domainname
2333 What to use for the Host: header if it is not present
2334 in a request. Determines what site (not origin server)
2335 accelerators should consider the default.
2337 no-vhost Disable using HTTP/1.1 Host header for virtual domain support.
2339 protocol= Protocol to reconstruct accelerated and intercepted
2340 requests with. Defaults to HTTP/1.1 for http_port and
2341 HTTPS/1.1 for https_port.
2342 When an unsupported value is configured Squid will
2343 produce a FATAL error.
2344 Values: HTTP or HTTP/1.1, HTTPS or HTTPS/1.1
2346 vport Virtual host port support. Using the http_port number
2347 instead of the port passed on Host: headers.
2349 vport=NN Virtual host port support. Using the specified port
2350 number instead of the port passed on Host: headers.
2353 Act as if this Squid is the origin server.
2354 This currently means generate new Date: and Expires:
2355 headers on HIT instead of adding Age:.
2357 ignore-cc Ignore request Cache-Control headers.
2359 WARNING: This option violates HTTP specifications if
2360 used in non-accelerator setups.
2362 allow-direct Allow direct forwarding in accelerator mode. Normally
2363 accelerated requests are denied direct forwarding as if
2364 never_direct was used.
2366 WARNING: this option opens accelerator mode to security
2367 vulnerabilities usually only affecting in interception
2368 mode. Make sure to protect forwarding with suitable
2369 http_access rules when using this.
2372 SSL Bump Mode Options:
2373 In addition to these options ssl-bump requires TLS/SSL options.
2375 generate-host-certificates[=<on|off>]
2376 Dynamically create SSL server certificates for the
2377 destination hosts of bumped CONNECT requests.When
2378 enabled, the cert and key options are used to sign
2379 generated certificates. Otherwise generated
2380 certificate will be selfsigned.
2381 If there is a CA certificate lifetime of the generated
2382 certificate equals lifetime of the CA certificate. If
2383 generated certificate is selfsigned lifetime is three
2385 This option is enabled by default when ssl-bump is used.
2386 See the ssl-bump option above for more information.
2388 dynamic_cert_mem_cache_size=SIZE
2389 Approximate total RAM size spent on cached generated
2390 certificates. If set to zero, caching is disabled. The
2391 default value is 4MB.
2395 tls-cert= Path to file containing an X.509 certificate (PEM format)
2396 to be used in the TLS handshake ServerHello.
2398 If this certificate is constrained by KeyUsage TLS
2399 feature it must allow HTTP server usage, along with
2400 any additional restrictions imposed by your choice
2401 of options= settings.
2403 When OpenSSL is used this file may also contain a
2404 chain of intermediate CA certificates to send in the
2407 When GnuTLS is used this option (and any paired
2408 tls-key= option) may be repeated to load multiple
2409 certificates for different domains.
2411 Also, when generate-host-certificates=on is configured
2412 the first tls-cert= option must be a CA certificate
2413 capable of signing the automatically generated
2416 tls-key= Path to a file containing private key file (PEM format)
2417 for the previous tls-cert= option.
2419 If tls-key= is not specified tls-cert= is assumed to
2420 reference a PEM file containing both the certificate
2423 cipher= Colon separated list of supported ciphers.
2424 NOTE: some ciphers such as EDH ciphers depend on
2425 additional settings. If those settings are
2426 omitted the ciphers may be silently ignored
2427 by the OpenSSL library.
2429 options= Various SSL implementation options. The most important
2432 NO_SSLv3 Disallow the use of SSLv3
2434 NO_TLSv1 Disallow the use of TLSv1.0
2436 NO_TLSv1_1 Disallow the use of TLSv1.1
2438 NO_TLSv1_2 Disallow the use of TLSv1.2
2441 Always create a new key when using
2442 temporary/ephemeral DH key exchanges
2445 Enable ephemeral ECDH key exchange.
2446 The adopted curve should be specified
2447 using the tls-dh option.
2450 Disable use of RFC5077 session tickets.
2451 Some servers may have problems
2452 understanding the TLS extension due
2453 to ambiguous specification in RFC4507.
2455 ALL Enable various bug workarounds
2456 suggested as "harmless" by OpenSSL
2457 Be warned that this reduces SSL/TLS
2458 strength to some attacks.
2460 See the OpenSSL SSL_CTX_set_options documentation for a
2463 clientca= File containing the list of CAs to use when
2464 requesting a client certificate.
2466 tls-cafile= PEM file containing CA certificates to use when verifying
2467 client certificates. If not configured clientca will be
2468 used. May be repeated to load multiple files.
2470 capath= Directory containing additional CA certificates
2471 and CRL lists to use when verifying client certificates.
2472 Requires OpenSSL or LibreSSL.
2474 crlfile= File of additional CRL lists to use when verifying
2475 the client certificate, in addition to CRLs stored in
2476 the capath. Implies VERIFY_CRL flag below.
2479 File containing DH parameters for temporary/ephemeral DH key
2480 exchanges, optionally prefixed by a curve for ephemeral ECDH
2482 See OpenSSL documentation for details on how to create the
2483 DH parameter file. Supported curves for ECDH can be listed
2484 using the "openssl ecparam -list_curves" command.
2485 WARNING: EDH and EECDH ciphers will be silently disabled if
2486 this option is not set.
2488 sslflags= Various flags modifying the use of SSL:
2490 Don't request client certificates
2491 immediately, but wait until acl processing
2492 requires a certificate (not yet implemented).
2494 Request a client certificate during the TLS
2495 handshake, but ignore certificate absence in
2496 the TLS client Hello. If the client does
2497 supply a certificate, it is validated.
2499 Don't allow for session reuse. Each connection
2500 will result in a new SSL session.
2502 Verify CRL lists when accepting client
2505 Verify CRL lists for all certificates in the
2506 client certificate chain.
2508 tls-default-ca[=off]
2509 Whether to use the system Trusted CAs. Default is OFF.
2511 tls-no-npn Do not use the TLS NPN extension to advertise HTTP/1.1.
2513 sslcontext= SSL session ID context identifier.
2517 connection-auth[=on|off]
2518 use connection-auth=off to tell Squid to prevent
2519 forwarding Microsoft connection oriented authentication
2520 (NTLM, Negotiate and Kerberos)
2522 disable-pmtu-discovery=
2523 Control Path-MTU discovery usage:
2524 off lets OS decide on what to do (default).
2525 transparent disable PMTU discovery when transparent
2527 always disable always PMTU discovery.
2529 In many setups of transparently intercepting proxies
2530 Path-MTU discovery can not work on traffic towards the
2531 clients. This is the case when the intercepting device
2532 does not fully track connections and fails to forward
2533 ICMP must fragment messages to the cache server. If you
2534 have such setup and experience that certain clients
2535 sporadically hang or never complete requests set
2536 disable-pmtu-discovery option to 'transparent'.
2538 name= Specifies a internal name for the port. Defaults to
2539 the port specification (port or addr:port)
2541 tcpkeepalive[=idle,interval,timeout]
2542 Enable TCP keepalive probes of idle connections.
2543 In seconds; idle is the initial time before TCP starts
2544 probing the connection, interval how often to probe, and
2545 timeout the time before giving up.
2547 require-proxy-header
2548 Require PROXY protocol version 1 or 2 connections.
2549 The proxy_protocol_access is required to permit
2550 downstream proxies which can be trusted.
2553 Ask TCP stack to maintain a dedicated listening queue
2554 for each worker accepting requests at this port.
2555 Requires TCP stack that supports the SO_REUSEPORT socket
2558 SECURITY WARNING: Enabling worker-specific queues
2559 allows any process running as Squid's effective user to
2560 easily accept requests destined to this port.
2562 If you run Squid on a dual-homed machine with an internal
2563 and an external interface we recommend you to specify the
2564 internal address:port in http_port. This way Squid will only be
2565 visible on the internal address.
2569 # Squid normally listens to port 3128
2570 http_port @DEFAULT_HTTP_PORT@
2575 IFDEF: HAVE_LIBGNUTLS||USE_OPENSSL
2580 Usage: [ip:]port [mode] tls-cert=certificate.pem [options]
2582 The socket address where Squid will listen for client requests made
2583 over TLS or SSL connections. Commonly referred to as HTTPS.
2585 This is most useful for situations where you are running squid in
2586 accelerator mode and you want to do the TLS work at the accelerator
2589 You may specify multiple socket addresses on multiple lines,
2590 each with their own certificate and/or options.
2592 The tls-cert= option is mandatory on HTTPS ports.
2594 See http_port for a list of modes and options.
2595 Not all http_port options are available for https_port.
2596 Among the unavalable options:
2597 - require-proxy-header
2605 Enables Native FTP proxy by specifying the socket address where Squid
2606 listens for FTP client requests. See http_port directive for various
2607 ways to specify the listening address and mode.
2609 Usage: ftp_port address [mode] [options]
2611 WARNING: This is a new, experimental, complex feature that has seen
2612 limited production exposure. Some Squid modules (e.g., caching) do not
2613 currently work with native FTP proxying, and many features have not
2614 even been tested for compatibility. Test well before deploying!
2616 Native FTP proxying differs substantially from proxying HTTP requests
2617 with ftp:// URIs because Squid works as an FTP server and receives
2618 actual FTP commands (rather than HTTP requests with FTP URLs).
2620 Native FTP commands accepted at ftp_port are internally converted or
2621 wrapped into HTTP-like messages. The same happens to Native FTP
2622 responses received from FTP origin servers. Those HTTP-like messages
2623 are shoveled through regular access control and adaptation layers
2624 between the FTP client and the FTP origin server. This allows Squid to
2625 examine, adapt, block, and log FTP exchanges. Squid reuses most HTTP
2626 mechanisms when shoveling wrapped FTP messages. For example,
2627 http_access and adaptation_access directives are used.
2631 intercept Same as http_port intercept. The FTP origin address is
2632 determined based on the intended destination of the
2633 intercepted connection.
2635 tproxy Support Linux TPROXY for spoofing outgoing
2636 connections using the client IP address.
2637 NP: disables authentication and maybe IPv6 on the port.
2639 By default (i.e., without an explicit mode option), Squid extracts the
2640 FTP origin address from the login@origin parameter of the FTP USER
2641 command. Many popular FTP clients support such native FTP proxying.
2645 name=token Specifies an internal name for the port. Defaults to
2646 the port address. Usable with myportname ACL.
2649 Enables tracking of FTP directories by injecting extra
2650 PWD commands and adjusting Request-URI (in wrapping
2651 HTTP requests) to reflect the current FTP server
2652 directory. Tracking is disabled by default.
2654 protocol=FTP Protocol to reconstruct accelerated and intercepted
2655 requests with. Defaults to FTP. No other accepted
2656 values have been tested with. An unsupported value
2657 results in a FATAL error. Accepted values are FTP,
2658 HTTP (or HTTP/1.1), and HTTPS (or HTTPS/1.1).
2660 Other http_port modes and options that are not specific to HTTP and
2661 HTTPS may also work.
2662 Among the options that are not available for ftp_port:
2663 - require-proxy-header
2667 NAME: tcp_outgoing_tos tcp_outgoing_ds tcp_outgoing_dscp
2670 LOC: Ip::Qos::TheConfig.tosToServer
2672 Allows you to select a TOS/Diffserv value for packets outgoing
2673 on the server side, based on an ACL.
2675 tcp_outgoing_tos ds-field [!]aclname ...
2677 Example where normal_service_net uses the TOS value 0x00
2678 and good_service_net uses 0x20
2680 acl normal_service_net src 10.0.0.0/24
2681 acl good_service_net src 10.0.1.0/24
2682 tcp_outgoing_tos 0x00 normal_service_net
2683 tcp_outgoing_tos 0x20 good_service_net
2685 TOS/DSCP values really only have local significance - so you should
2686 know what you're specifying. For more information, see RFC2474,
2687 RFC2475, and RFC3260.
2689 The TOS/DSCP byte must be exactly that - a octet value 0 - 255, or
2690 "default" to use whatever default your host has.
2691 Note that only multiples of 4 are usable as the two rightmost bits have
2692 been redefined for use by ECN (RFC 3168 section 23.1).
2693 The squid parser will enforce this by masking away the ECN bits.
2695 Processing proceeds in the order specified, and stops at first fully
2698 Only fast ACLs are supported.
2701 NAME: clientside_tos
2704 LOC: Ip::Qos::TheConfig.tosToClient
2706 Allows you to select a TOS/DSCP value for packets being transmitted
2707 on the client-side, based on an ACL.
2709 clientside_tos ds-field [!]aclname ...
2711 Example where normal_service_net uses the TOS value 0x00
2712 and good_service_net uses 0x20
2714 acl normal_service_net src 10.0.0.0/24
2715 acl good_service_net src 10.0.1.0/24
2716 clientside_tos 0x00 normal_service_net
2717 clientside_tos 0x20 good_service_net
2719 Note: This feature is incompatible with qos_flows. Any TOS values set here
2720 will be overwritten by TOS values in qos_flows.
2722 The TOS/DSCP byte must be exactly that - a octet value 0 - 255, or
2723 "default" to use whatever default your host has.
2724 Note that only multiples of 4 are usable as the two rightmost bits have
2725 been redefined for use by ECN (RFC 3168 section 23.1).
2726 The squid parser will enforce this by masking away the ECN bits.
2728 This clause only supports fast acl types.
2729 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
2732 NAME: tcp_outgoing_mark
2734 IFDEF: HAVE_LIBCAP&&SO_MARK
2736 LOC: Ip::Qos::TheConfig.nfmarkToServer
2738 Allows you to apply a Netfilter mark value to outgoing packets
2739 on the server side, based on an ACL.
2741 tcp_outgoing_mark mark-value [!]aclname ...
2743 Example where normal_service_net uses the mark value 0x00
2744 and good_service_net uses 0x20
2746 acl normal_service_net src 10.0.0.0/24
2747 acl good_service_net src 10.0.1.0/24
2748 tcp_outgoing_mark 0x00 normal_service_net
2749 tcp_outgoing_mark 0x20 good_service_net
2751 Only fast ACLs are supported.
2754 NAME: mark_client_packet clientside_mark
2756 IFDEF: HAVE_LIBCAP&&SO_MARK
2758 LOC: Ip::Qos::TheConfig.nfmarkToClient
2760 Allows you to apply a Netfilter MARK value to packets being transmitted
2761 on the client-side, based on an ACL.
2763 mark_client_packet mark-value [!]aclname ...
2765 Example where normal_service_net uses the MARK value 0x00
2766 and good_service_net uses 0x20
2768 acl normal_service_net src 10.0.0.0/24
2769 acl good_service_net src 10.0.1.0/24
2770 mark_client_packet 0x00 normal_service_net
2771 mark_client_packet 0x20 good_service_net
2773 Note: This feature is incompatible with qos_flows. Any mark values set here
2774 will be overwritten by mark values in qos_flows.
2776 This clause only supports fast acl types.
2777 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
2780 NAME: mark_client_connection
2782 IFDEF: HAVE_LIBCAP&&SO_MARK
2784 LOC: Ip::Qos::TheConfig.nfConnmarkToClient
2786 Allows you to apply a Netfilter CONNMARK value to a connection
2787 on the client-side, based on an ACL.
2789 mark_client_connection mark-value[/mask] [!]aclname ...
2791 The mark-value and mask are unsigned integers (hex, octal, or decimal).
2792 The mask may be used to preserve marking previously set by other agents
2795 A matching rule replaces the CONNMARK value. If a mask is also
2796 specified, then the masked bits of the original value are zeroed, and
2797 the configured mark-value is ORed with that adjusted value.
2798 For example, applying a mark-value 0xAB/0xF to 0x5F CONNMARK, results
2799 in a 0xFB marking (rather than a 0xAB or 0x5B).
2801 This directive semantics is similar to iptables --set-mark rather than
2802 --set-xmark functionality.
2804 The directive does not interfere with qos_flows (which uses packet MARKs,
2807 Example where squid marks intercepted FTP connections:
2809 acl proto_ftp proto FTP
2810 mark_client_connection 0x200/0xff00 proto_ftp
2812 This clause only supports fast acl types.
2813 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
2820 LOC: Ip::Qos::TheConfig
2822 Allows you to select a TOS/DSCP value to mark outgoing
2823 connections to the client, based on where the reply was sourced.
2824 For platforms using netfilter, allows you to set a netfilter mark
2825 value instead of, or in addition to, a TOS value.
2827 By default this functionality is disabled. To enable it with the default
2828 settings simply use "qos_flows mark" or "qos_flows tos". Default
2829 settings will result in the netfilter mark or TOS value being copied
2830 from the upstream connection to the client. Note that it is the connection
2831 CONNMARK value not the packet MARK value that is copied.
2833 It is not currently possible to copy the mark or TOS value from the
2834 client to the upstream connection request.
2836 TOS values really only have local significance - so you should
2837 know what you're specifying. For more information, see RFC2474,
2838 RFC2475, and RFC3260.
2840 The TOS/DSCP byte must be exactly that - a octet value 0 - 255.
2841 Note that only multiples of 4 are usable as the two rightmost bits have
2842 been redefined for use by ECN (RFC 3168 section 23.1).
2843 The squid parser will enforce this by masking away the ECN bits.
2845 Mark values can be any unsigned 32-bit integer value.
2847 This setting is configured by setting the following values:
2849 tos|mark Whether to set TOS or netfilter mark values
2851 local-hit=0xFF Value to mark local cache hits.
2853 sibling-hit=0xFF Value to mark hits from sibling peers.
2855 parent-hit=0xFF Value to mark hits from parent peers.
2857 miss=0xFF[/mask] Value to mark cache misses. Takes precedence
2858 over the preserve-miss feature (see below), unless
2859 mask is specified, in which case only the bits
2860 specified in the mask are written.
2862 The TOS variant of the following features are only possible on Linux
2863 and require your kernel to be patched with the TOS preserving ZPH
2864 patch, available from http://zph.bratcheda.org
2865 No patch is needed to preserve the netfilter mark, which will work
2866 with all variants of netfilter.
2868 disable-preserve-miss
2869 This option disables the preservation of the TOS or netfilter
2870 mark. By default, the existing TOS or netfilter mark value of
2871 the response coming from the remote server will be retained
2872 and masked with miss-mark.
2873 NOTE: in the case of a netfilter mark, the mark must be set on
2874 the connection (using the CONNMARK target) not on the packet
2878 Allows you to mask certain bits in the TOS or mark value
2879 received from the remote server, before copying the value to
2880 the TOS sent towards clients.
2881 Default for tos: 0xFF (TOS from server is not changed).
2882 Default for mark: 0xFFFFFFFF (mark from server is not changed).
2884 All of these features require the --enable-zph-qos compilation flag
2885 (enabled by default). Netfilter marking also requires the
2886 libnetfilter_conntrack libraries (--with-netfilter-conntrack) and
2887 libcap 2.09+ (--with-libcap).
2891 NAME: tcp_outgoing_address
2894 DEFAULT_DOC: Address selection is performed by the operating system.
2895 LOC: Config.accessList.outgoing_address
2897 Allows you to map requests to different outgoing IP addresses
2898 based on the username or source address of the user making
2901 tcp_outgoing_address ipaddr [[!]aclname] ...
2904 Forwarding clients with dedicated IPs for certain subnets.
2906 acl normal_service_net src 10.0.0.0/24
2907 acl good_service_net src 10.0.2.0/24
2909 tcp_outgoing_address 2001:db8::c001 good_service_net
2910 tcp_outgoing_address 10.1.0.2 good_service_net
2912 tcp_outgoing_address 2001:db8::beef normal_service_net
2913 tcp_outgoing_address 10.1.0.1 normal_service_net
2915 tcp_outgoing_address 2001:db8::1
2916 tcp_outgoing_address 10.1.0.3
2918 Processing proceeds in the order specified, and stops at first fully
2921 Squid will add an implicit IP version test to each line.
2922 Requests going to IPv4 websites will use the outgoing 10.1.0.* addresses.
2923 Requests going to IPv6 websites will use the outgoing 2001:db8:* addresses.
2926 NOTE: The use of this directive using client dependent ACLs is
2927 incompatible with the use of server side persistent connections. To
2928 ensure correct results it is best to set server_persistent_connections
2929 to off when using this directive in such configurations.
2931 NOTE: The use of this directive to set a local IP on outgoing TCP links
2932 is incompatible with using TPROXY to set client IP out outbound TCP links.
2933 When needing to contact peers use the no-tproxy cache_peer option and the
2934 client_dst_passthru directive re-enable normal forwarding such as this.
2936 This clause only supports fast acl types.
2937 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
2940 NAME: host_verify_strict
2943 LOC: Config.onoff.hostStrictVerify
2945 Regardless of this option setting, when dealing with intercepted
2946 traffic, Squid always verifies that the destination IP address matches
2947 the Host header domain or IP (called 'authority form URL').
2949 This enforcement is performed to satisfy a MUST-level requirement in
2950 RFC 2616 section 14.23: "The Host field value MUST represent the naming
2951 authority of the origin server or gateway given by the original URL".
2954 Squid always responds with an HTTP 409 (Conflict) error
2955 page and logs a security warning if there is no match.
2957 Squid verifies that the destination IP address matches
2958 the Host header for forward-proxy and reverse-proxy traffic
2959 as well. For those traffic types, Squid also enables the
2960 following checks, comparing the corresponding Host header
2961 and Request-URI components:
2963 * The host names (domain or IP) must be identical,
2964 but valueless or missing Host header disables all checks.
2965 For the two host names to match, both must be either IP
2968 * Port numbers must be identical, but if a port is missing
2969 the scheme-default port is assumed.
2972 When set to OFF (the default):
2973 Squid allows suspicious requests to continue but logs a
2974 security warning and blocks caching of the response.
2976 * Forward-proxy traffic is not checked at all.
2978 * Reverse-proxy traffic is not checked at all.
2980 * Intercepted traffic which passes verification is handled
2981 according to client_dst_passthru.
2983 * Intercepted requests which fail verification are sent
2984 to the client original destination instead of DIRECT.
2985 This overrides 'client_dst_passthru off'.
2987 For now suspicious intercepted CONNECT requests are always
2988 responded to with an HTTP 409 (Conflict) error page.
2993 As described in CVE-2009-0801 when the Host: header alone is used
2994 to determine the destination of a request it becomes trivial for
2995 malicious scripts on remote websites to bypass browser same-origin
2996 security policy and sandboxing protections.
2998 The cause of this is that such applets are allowed to perform their
2999 own HTTP stack, in which case the same-origin policy of the browser
3000 sandbox only verifies that the applet tries to contact the same IP
3001 as from where it was loaded at the IP level. The Host: header may
3002 be different from the connected IP and approved origin.
3006 NAME: client_dst_passthru
3009 LOC: Config.onoff.client_dst_passthru
3011 With NAT or TPROXY intercepted traffic Squid may pass the request
3012 directly to the original client destination IP or seek a faster
3013 source using the HTTP Host header.
3015 Using Host to locate alternative servers can provide faster
3016 connectivity with a range of failure recovery options.
3017 But can also lead to connectivity trouble when the client and
3018 server are attempting stateful interactions unaware of the proxy.
3020 This option (on by default) prevents alternative DNS entries being
3021 located to send intercepted traffic DIRECT to an origin server.
3022 The clients original destination IP and port will be used instead.
3024 Regardless of this option setting, when dealing with intercepted
3025 traffic Squid will verify the Host: header and any traffic which
3026 fails Host verification will be treated as if this option were ON.
3028 see host_verify_strict for details on the verification process.
3033 -----------------------------------------------------------------------------
3036 NAME: tls_outgoing_options
3037 IFDEF: HAVE_LIBGNUTLS||USE_OPENSSL
3038 TYPE: securePeerOptions
3039 DEFAULT: min-version=1.0
3040 LOC: Security::ProxyOutgoingConfig()
3042 disable Do not support https:// URLs.
3044 cert=/path/to/client/certificate
3045 A client X.509 certificate to use when connecting.
3047 key=/path/to/client/private_key
3048 The private key corresponding to the cert= above.
3050 If key= is not specified cert= is assumed to
3051 reference a PEM file containing both the certificate
3054 cipher=... The list of valid TLS ciphers to use.
3057 The minimum TLS protocol version to permit.
3058 To control SSLv3 use the options= parameter.
3059 Supported Values: 1.0 (default), 1.1, 1.2, 1.3
3061 options=... Specify various TLS/SSL implementation options.
3063 OpenSSL options most important are:
3065 NO_SSLv3 Disallow the use of SSLv3
3068 Always create a new key when using
3069 temporary/ephemeral DH key exchanges
3072 Disable use of RFC5077 session tickets.
3073 Some servers may have problems
3074 understanding the TLS extension due
3075 to ambiguous specification in RFC4507.
3077 ALL Enable various bug workarounds
3078 suggested as "harmless" by OpenSSL
3079 Be warned that this reduces SSL/TLS
3080 strength to some attacks.
3082 See the OpenSSL SSL_CTX_set_options documentation
3083 for a more complete list.
3085 GnuTLS options most important are:
3088 Disable use of RFC5077 session tickets.
3089 Some servers may have problems
3090 understanding the TLS extension due
3091 to ambiguous specification in RFC4507.
3093 See the GnuTLS Priority Strings documentation
3094 for a more complete list.
3095 http://www.gnutls.org/manual/gnutls.html#Priority-Strings
3098 cafile= PEM file containing CA certificates to use when verifying
3099 the peer certificate. May be repeated to load multiple files.
3101 capath= A directory containing additional CA certificates to
3102 use when verifying the peer certificate.
3103 Requires OpenSSL or LibreSSL.
3105 crlfile=... A certificate revocation list file to use when
3106 verifying the peer certificate.
3108 flags=... Specify various flags modifying the TLS implementation:
3111 Accept certificates even if they fail to
3114 Don't verify the peer certificate
3115 matches the server name
3118 Whether to use the system Trusted CAs. Default is ON.
3120 domain= The peer name as advertised in its certificate.
3121 Used for verifying the correctness of the received peer
3122 certificate. If not specified the peer hostname will be
3128 -----------------------------------------------------------------------------
3131 NAME: ssl_unclean_shutdown
3135 LOC: Config.SSL.unclean_shutdown
3137 Some browsers (especially MSIE) bugs out on SSL shutdown
3144 LOC: Config.SSL.ssl_engine
3147 The OpenSSL engine to use. You will need to set this if you
3148 would like to use hardware SSL acceleration for example.
3150 Not supported in builds with OpenSSL 3.0 or newer.
3153 NAME: sslproxy_session_ttl
3156 LOC: Config.SSL.session_ttl
3159 Sets the timeout value for SSL sessions
3162 NAME: sslproxy_session_cache_size
3165 LOC: Config.SSL.sessionCacheSize
3168 Sets the cache size to use for ssl session
3171 NAME: sslproxy_foreign_intermediate_certs
3174 LOC: Config.ssl_client.foreignIntermediateCertsPath
3177 Many origin servers fail to send their full server certificate
3178 chain for verification, assuming the client already has or can
3179 easily locate any missing intermediate certificates.
3181 Squid uses the certificates from the specified file to fill in
3182 these missing chains when trying to validate origin server
3185 The file is expected to contain zero or more PEM-encoded
3186 intermediate certificates. These certificates are not treated
3187 as trusted root certificates, and any self-signed certificate in
3188 this file will be ignored.
3191 NAME: sslproxy_cert_sign_hash
3194 LOC: Config.SSL.certSignHash
3197 Sets the hashing algorithm to use when signing generated certificates.
3198 Valid algorithm names depend on the OpenSSL library used. The following
3199 names are usually available: sha1, sha256, sha512, and md5. Please see
3200 your OpenSSL library manual for the available hashes. By default, Squids
3201 that support this option use sha256 hashes.
3203 Squid does not forcefully purge cached certificates that were generated
3204 with an algorithm other than the currently configured one. They remain
3205 in the cache, subject to the regular cache eviction policy, and become
3206 useful if the algorithm changes again.
3211 TYPE: sslproxy_ssl_bump
3212 LOC: Config.accessList.ssl_bump
3213 DEFAULT_DOC: Become a TCP tunnel without decrypting proxied traffic.
3216 This option is consulted when a CONNECT request is received on
3217 an http_port (or a new connection is intercepted at an
3218 https_port), provided that port was configured with an ssl-bump
3219 flag. The subsequent data on the connection is either treated as
3220 HTTPS and decrypted OR tunneled at TCP level without decryption,
3221 depending on the first matching bumping "action".
3223 ssl_bump <action> [!]acl ...
3225 The following bumping actions are currently supported:
3228 Become a TCP tunnel without decrypting proxied traffic.
3229 This is the default action.
3232 When used on step SslBump1, establishes a secure connection
3233 with the client first, then connect to the server.
3234 When used on step SslBump2 or SslBump3, establishes a secure
3235 connection with the server and, using a mimicked server
3236 certificate, with the client.
3239 Receive client (step SslBump1) or server (step SslBump2)
3240 certificate while preserving the possibility of splicing the
3241 connection. Peeking at the server certificate (during step 2)
3242 usually precludes bumping of the connection at step 3.
3245 Receive client (step SslBump1) or server (step SslBump2)
3246 certificate while preserving the possibility of bumping the
3247 connection. Staring at the server certificate (during step 2)
3248 usually precludes splicing of the connection at step 3.
3251 Close client and server connections.
3253 Backward compatibility actions available at step SslBump1:
3256 Bump the connection. Establish a secure connection with the
3257 client first, then connect to the server. This old mode does
3258 not allow Squid to mimic server SSL certificate and does not
3259 work with intercepted SSL connections.
3262 Bump the connection. Establish a secure connection with the
3263 server first, then establish a secure connection with the
3264 client, using a mimicked server certificate. Works with both
3265 CONNECT requests and intercepted SSL connections, but does
3266 not allow to make decisions based on SSL handshake info.
3269 Decide whether to bump or splice the connection based on
3270 client-to-squid and server-to-squid SSL hello messages.
3274 Same as the "splice" action.
3276 All ssl_bump rules are evaluated at each of the supported bumping
3277 steps. Rules with actions that are impossible at the current step are
3278 ignored. The first matching ssl_bump action wins and is applied at the
3279 end of the current step. If no rules match, the splice action is used.
3280 See the at_step ACL for a list of the supported SslBump steps.
3282 This clause supports both fast and slow acl types.
3283 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
3285 See also: http_port ssl-bump, https_port ssl-bump, and acl at_step.
3288 # Example: Bump all TLS connections except those originating from
3289 # localhost or those going to example.com.
3291 acl broken_sites ssl::server_name .example.com
3292 ssl_bump splice localhost
3293 ssl_bump splice broken_sites
3297 NAME: sslproxy_cert_error
3300 DEFAULT_DOC: Server certificate errors terminate the transaction.
3301 LOC: Config.ssl_client.cert_error
3304 Use this ACL to bypass server certificate validation errors.
3306 For example, the following lines will bypass all validation errors
3307 when talking to servers for example.com. All other
3308 validation errors will result in ERR_SECURE_CONNECT_FAIL error.
3310 acl BrokenButTrustedServers dstdomain example.com
3311 sslproxy_cert_error allow BrokenButTrustedServers
3312 sslproxy_cert_error deny all
3314 This clause only supports fast acl types.
3315 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
3316 Using slow acl types may result in server crashes
3318 Without this option, all server certificate validation errors
3319 terminate the transaction to protect Squid and the client.
3321 SQUID_X509_V_ERR_INFINITE_VALIDATION error cannot be bypassed
3322 but should not happen unless your OpenSSL library is buggy.
3325 Bypassing validation errors is dangerous because an
3326 error usually implies that the server cannot be trusted
3327 and the connection may be insecure.
3329 See also: sslproxy_flags and DONT_VERIFY_PEER.
3332 NAME: sslproxy_cert_sign
3335 POSTSCRIPTUM: signUntrusted ssl::certUntrusted
3336 POSTSCRIPTUM: signSelf ssl::certSelfSigned
3337 POSTSCRIPTUM: signTrusted all
3338 TYPE: sslproxy_cert_sign
3339 LOC: Config.ssl_client.cert_sign
3342 sslproxy_cert_sign <signing algorithm> acl ...
3344 The following certificate signing algorithms are supported:
3347 Sign using the configured CA certificate which is usually
3348 placed in and trusted by end-user browsers. This is the
3349 default for trusted origin server certificates.
3352 Sign to guarantee an X509_V_ERR_CERT_UNTRUSTED browser error.
3353 This is the default for untrusted origin server certificates
3354 that are not self-signed (see ssl::certUntrusted).
3357 Sign using a self-signed certificate with the right CN to
3358 generate a X509_V_ERR_DEPTH_ZERO_SELF_SIGNED_CERT error in the
3359 browser. This is the default for self-signed origin server
3360 certificates (see ssl::certSelfSigned).
3362 This clause only supports fast acl types.
3364 When sslproxy_cert_sign acl(s) match, Squid uses the corresponding
3365 signing algorithm to generate the certificate and ignores all
3366 subsequent sslproxy_cert_sign options (the first match wins). If no
3367 acl(s) match, the default signing algorithm is determined by errors
3368 detected when obtaining and validating the origin server certificate.
3370 WARNING: SQUID_X509_V_ERR_DOMAIN_MISMATCH and ssl:certDomainMismatch can
3371 be used with sslproxy_cert_adapt, but if and only if Squid is bumping a
3372 CONNECT request that carries a domain name. In all other cases (CONNECT
3373 to an IP address or an intercepted SSL connection), Squid cannot detect
3374 the domain mismatch at certificate generation time when
3375 bump-server-first is used.
3378 NAME: sslproxy_cert_adapt
3381 TYPE: sslproxy_cert_adapt
3382 LOC: Config.ssl_client.cert_adapt
3385 sslproxy_cert_adapt <adaptation algorithm> acl ...
3387 The following certificate adaptation algorithms are supported:
3390 Sets the "Not After" property to the "Not After" property of
3391 the CA certificate used to sign generated certificates.
3394 Sets the "Not Before" property to the "Not Before" property of
3395 the CA certificate used to sign generated certificates.
3397 setCommonName or setCommonName{CN}
3398 Sets Subject.CN property to the host name specified as a
3399 CN parameter or, if no explicit CN parameter was specified,
3400 extracted from the CONNECT request. It is a misconfiguration
3401 to use setCommonName without an explicit parameter for
3402 intercepted or tproxied SSL connections.
3404 This clause only supports fast acl types.
3406 Squid first groups sslproxy_cert_adapt options by adaptation algorithm.
3407 Within a group, when sslproxy_cert_adapt acl(s) match, Squid uses the
3408 corresponding adaptation algorithm to generate the certificate and
3409 ignores all subsequent sslproxy_cert_adapt options in that algorithm's
3410 group (i.e., the first match wins within each algorithm group). If no
3411 acl(s) match, the default mimicking action takes place.
3413 WARNING: SQUID_X509_V_ERR_DOMAIN_MISMATCH and ssl:certDomainMismatch can
3414 be used with sslproxy_cert_adapt, but if and only if Squid is bumping a
3415 CONNECT request that carries a domain name. In all other cases (CONNECT
3416 to an IP address or an intercepted SSL connection), Squid cannot detect
3417 the domain mismatch at certificate generation time when
3418 bump-server-first is used.
3421 NAME: sslpassword_program
3424 LOC: Config.Program.ssl_password
3427 Specify a program used for entering SSL key passphrases
3428 when using encrypted SSL certificate keys. If not specified
3429 keys must either be unencrypted, or Squid started with the -N
3430 option to allow it to query interactively for the passphrase.
3432 The key file name is given as argument to the program allowing
3433 selection of the right password if you have multiple encrypted
3438 OPTIONS RELATING TO EXTERNAL SSL_CRTD
3439 -----------------------------------------------------------------------------
3442 NAME: sslcrtd_program
3445 DEFAULT: @DEFAULT_SSL_CRTD@ -s @DEFAULT_SSL_DB_DIR@ -M 4MB
3446 LOC: Ssl::TheConfig.ssl_crtd
3448 Specify the location and options of the executable for certificate
3451 @DEFAULT_SSL_CRTD@ program can use a disk cache to improve response
3452 times on repeated requests. To enable caching, specify -s and -M
3453 parameters. If those parameters are not given, the program generates
3454 a new certificate on every request.
3456 For more information use:
3457 @DEFAULT_SSL_CRTD@ -h
3460 NAME: sslcrtd_children
3461 TYPE: HelperChildConfig
3463 DEFAULT: 32 startup=5 idle=1
3464 LOC: Ssl::TheConfig.ssl_crtdChildren
3466 Specifies the maximum number of certificate generation processes that
3467 Squid may spawn (numberofchildren) and several related options. Using
3468 too few of these helper processes (a.k.a. "helpers") creates request
3469 queues. Using too many helpers wastes your system resources. Squid
3470 does not support spawning more than 32 helpers.
3472 Usage: numberofchildren [option]...
3474 The startup= and idle= options allow some measure of skew in your
3479 Sets the minimum number of processes to spawn when Squid
3480 starts or reconfigures. When set to zero the first request will
3481 cause spawning of the first child process to handle it.
3483 Starting too few children temporary slows Squid under load while it
3484 tries to spawn enough additional processes to cope with traffic.
3488 Sets a minimum of how many processes Squid is to try and keep available
3489 at all times. When traffic begins to rise above what the existing
3490 processes can handle this many more will be spawned up to the maximum
3491 configured. A minimum setting of 1 is required.
3495 Sets the maximum number of queued requests. A request is queued when
3496 no existing child is idle and no new child can be started due to
3497 numberofchildren limit. If the queued requests exceed queue size for
3498 more than 3 minutes squid aborts its operation. The default value is
3499 set to 2*numberofchildren.
3501 You must have at least one ssl_crtd process.
3504 NAME: sslcrtvalidator_program
3508 LOC: Ssl::TheConfig.ssl_crt_validator
3510 Specify the location and options of the executable for ssl_crt_validator
3513 Usage: sslcrtvalidator_program [ttl=...] [cache=n] path ...
3518 Limits how much memory Squid can use for caching validator
3519 responses. The default is 67108864 (i.e. 64 MB).
3520 Reconfiguration purges any excess entries. To disable caching,
3521 use cache=0. Currently, cache entry sizes are seriously
3522 underestimated. Even with that bug, a typical estimate for a
3523 single cache entry size would be at least a few kilobytes (the
3524 size of the PEM certificates sent to the validator).
3526 ttl=<seconds|"infinity">
3527 Approximately how long Squid may reuse the validator results
3528 for. The default is 3600 (i.e. 1 hour). Using ttl=infinity
3529 disables TTL checks. Reconfiguration does not affect TTLs of
3530 the already cached entries. To disable caching, use zero cache
3531 size, not zero TTL -- zero TTL allows reuse for the remainder
3532 of the second when the result was cached.
3535 NAME: sslcrtvalidator_children
3536 TYPE: HelperChildConfig
3538 DEFAULT: 32 startup=5 idle=1 concurrency=1
3539 LOC: Ssl::TheConfig.ssl_crt_validator_Children
3541 Specifies the maximum number of certificate validation processes that
3542 Squid may spawn (numberofchildren) and several related options. Using
3543 too few of these helper processes (a.k.a. "helpers") creates request
3544 queues. Using too many helpers wastes your system resources. Squid
3545 does not support spawning more than 32 helpers.
3547 Usage: numberofchildren [option]...
3549 The startup= and idle= options allow some measure of skew in your
3554 Sets the minimum number of processes to spawn when Squid
3555 starts or reconfigures. When set to zero the first request will
3556 cause spawning of the first child process to handle it.
3558 Starting too few children temporary slows Squid under load while it
3559 tries to spawn enough additional processes to cope with traffic.
3563 Sets a minimum of how many processes Squid is to try and keep available
3564 at all times. When traffic begins to rise above what the existing
3565 processes can handle this many more will be spawned up to the maximum
3566 configured. A minimum setting of 1 is required.
3570 The number of requests each certificate validator helper can handle in
3571 parallel. A value of 0 indicates the certificate validator does not
3572 support concurrency. Defaults to 1.
3574 When this directive is set to a value >= 1 then the protocol
3575 used to communicate with the helper is modified to include
3576 a request ID in front of the request/response. The request
3577 ID from the request must be echoed back with the response
3582 Sets the maximum number of queued requests. A request is queued when
3583 no existing child can accept it due to concurrency limit and no new
3584 child can be started due to numberofchildren limit. If the queued
3585 requests exceed queue size for more than 3 minutes squid aborts its
3586 operation. The default value is set to 2*numberofchildren.
3588 You must have at least one ssl_crt_validator process.
3592 OPTIONS WHICH AFFECT THE NEIGHBOR SELECTION ALGORITHM
3593 -----------------------------------------------------------------------------
3601 To specify other caches in a hierarchy, use the format:
3603 cache_peer hostname type http-port icp-port [options]
3608 # hostname type port port options
3609 # -------------------- -------- ----- ----- -----------
3610 cache_peer parent.foo.net parent 3128 3130 default
3611 cache_peer sib1.foo.net sibling 3128 3130 proxy-only
3612 cache_peer sib2.foo.net sibling 3128 3130 proxy-only
3613 cache_peer example.com parent 80 0 default
3614 cache_peer cdn.example.com sibling 3128 0
3616 type: either 'parent', 'sibling', or 'multicast'.
3618 proxy-port: The port number where the peer accept HTTP requests.
3619 For other Squid proxies this is usually 3128
3620 For web servers this is usually 80
3622 icp-port: Used for querying neighbor caches about objects.
3623 Set to 0 if the peer does not support ICP or HTCP.
3624 See ICP and HTCP options below for additional details.
3627 ==== ICP OPTIONS ====
3629 You MUST also set icp_port and icp_access explicitly when using these options.
3630 The defaults will prevent peer traffic using ICP.
3633 no-query Disable ICP queries to this neighbor.
3636 Indicates the named peer is a member of a multicast group.
3637 ICP queries will not be sent directly to the peer, but ICP
3638 replies will be accepted from it.
3640 closest-only Indicates that, for ICP_OP_MISS replies, we'll only forward
3641 CLOSEST_PARENT_MISSes and never FIRST_PARENT_MISSes.
3644 To only send ICP queries to this neighbor infrequently.
3645 This is used to keep the neighbor round trip time updated
3646 and is usually used in conjunction with weighted-round-robin.
3649 ==== HTCP OPTIONS ====
3651 You MUST also set htcp_port and htcp_access explicitly when using these options.
3652 The defaults will prevent peer traffic using HTCP.
3655 htcp Send HTCP, instead of ICP, queries to the neighbor.
3656 You probably also want to set the "icp-port" to 4827
3657 instead of 3130. This directive accepts a comma separated
3658 list of options described below.
3660 htcp=oldsquid Send HTCP to old Squid versions (2.5 or earlier).
3662 htcp=no-clr Send HTCP to the neighbor but without
3663 sending any CLR requests. This cannot be used with
3666 htcp=only-clr Send HTCP to the neighbor but ONLY CLR requests.
3667 This cannot be used with no-clr.
3670 Send HTCP to the neighbor including CLRs but only when
3671 they do not result from PURGE requests.
3674 Forward any HTCP CLR requests this proxy receives to the peer.
3677 ==== PEER SELECTION METHODS ====
3679 The default peer selection method is ICP, with the first responding peer
3680 being used as source. These options can be used for better load balancing.
3683 default This is a parent cache which can be used as a "last-resort"
3684 if a peer cannot be located by any of the peer-selection methods.
3685 If specified more than once, only the first is used.
3687 round-robin Load-Balance parents which should be used in a round-robin
3688 fashion in the absence of any ICP queries.
3689 weight=N can be used to add bias.
3691 weighted-round-robin
3692 Load-Balance parents which should be used in a round-robin
3693 fashion with the frequency of each parent being based on the
3694 round trip time. Closer parents are used more often.
3695 Usually used for background-ping parents.
3696 weight=N can be used to add bias.
3698 carp Load-Balance parents which should be used as a CARP array.
3699 The requests will be distributed among the parents based on the
3700 CARP load balancing hash function based on their weight.
3702 userhash Load-balance parents based on the client proxy_auth username.
3704 sourcehash Load-balance parents based on the client source IP.
3707 To be used only for cache peers of type "multicast".
3708 ALL members of this multicast group have "sibling"
3709 relationship with it, not "parent". This is to a multicast
3710 group when the requested object would be fetched only from
3711 a "parent" cache, anyway. It's useful, e.g., when
3712 configuring a pool of redundant Squid proxies, being
3713 members of the same multicast group.
3716 ==== PEER SELECTION OPTIONS ====
3718 weight=N use to affect the selection of a peer during any weighted
3719 peer-selection mechanisms.
3720 The weight must be an integer; default is 1,
3721 larger weights are favored more.
3722 This option does not affect parent selection if a peering
3723 protocol is not in use.
3725 basetime=N Specify a base amount to be subtracted from round trip
3727 It is subtracted before division by weight in calculating
3728 which parent to fectch from. If the rtt is less than the
3729 base time the rtt is set to a minimal value.
3731 ttl=N Specify a TTL to use when sending multicast ICP queries
3733 Only useful when sending to a multicast group.
3734 Because we don't accept ICP replies from random
3735 hosts, you must configure other group members as
3736 peers with the 'multicast-responder' option.
3738 no-delay To prevent access to this neighbor from influencing the
3741 digest-url=URL Tell Squid to fetch the cache digest (if digests are
3742 enabled) for this host from the specified URL rather
3743 than the Squid default location.
3746 ==== CARP OPTIONS ====
3748 carp-key=key-specification
3749 use a different key than the full URL to hash against the peer.
3750 the key-specification is a comma-separated list of the keywords
3751 scheme, host, port, path, params
3752 Order is not important.
3754 ==== ACCELERATOR / REVERSE-PROXY OPTIONS ====
3756 originserver Causes this parent to be contacted as an origin server.
3757 Meant to be used in accelerator setups when the peer
3761 Set the Host header of requests forwarded to this peer.
3762 Useful in accelerator setups where the server (peer)
3763 expects a certain domain name but clients may request
3764 others. ie example.com or www.example.com
3766 no-digest Disable request of cache digests.
3769 Disables requesting ICMP RTT database (NetDB).
3772 ==== AUTHENTICATION OPTIONS ====
3775 If this is a personal/workgroup proxy and your parent
3776 requires proxy authentication.
3778 Note: The string can include URL escapes (i.e. %20 for
3779 spaces). This also means % must be written as %%.
3782 Send login details received from client to this peer.
3783 Both Proxy- and WWW-Authorization headers are passed
3784 without alteration to the peer.
3785 Authentication is not required by Squid for this to work.
3787 Note: This will pass any form of authentication but
3788 only Basic auth will work through a proxy unless the
3789 connection-auth options are also used.
3791 login=PASS Send login details received from client to this peer.
3792 Authentication is not required by this option.
3794 If there are no client-provided authentication headers
3795 to pass on, but username and password are available
3796 from an external ACL user= and password= result tags
3797 they may be sent instead.
3799 Note: To combine this with proxy_auth both proxies must
3800 share the same user database as HTTP only allows for
3801 a single login (one for proxy, one for origin server).
3802 Also be warned this will expose your users proxy
3803 password to the peer. USE WITH CAUTION
3806 Send the username to the upstream cache, but with a
3807 fixed password. This is meant to be used when the peer
3808 is in another administrative domain, but it is still
3809 needed to identify each user.
3810 The star can optionally be followed by some extra
3811 information which is added to the username. This can
3812 be used to identify this proxy to the peer, similar to
3813 the login=username:password option above.
3816 If this is a personal/workgroup proxy and your parent
3817 requires a secure proxy authentication.
3818 The first principal from the default keytab or defined by
3819 the environment variable KRB5_KTNAME will be used.
3821 WARNING: The connection may transmit requests from multiple
3822 clients. Negotiate often assumes end-to-end authentication
3823 and a single-client. Which is not strictly true here.
3825 login=NEGOTIATE:principal_name
3826 If this is a personal/workgroup proxy and your parent
3827 requires a secure proxy authentication.
3828 The principal principal_name from the default keytab or
3829 defined by the environment variable KRB5_KTNAME will be
3832 WARNING: The connection may transmit requests from multiple
3833 clients. Negotiate often assumes end-to-end authentication
3834 and a single-client. Which is not strictly true here.
3836 connection-auth=on|off
3837 Tell Squid that this peer does or not support Microsoft
3838 connection oriented authentication, and any such
3839 challenges received from there should be ignored.
3840 Default is auto to automatically determine the status
3844 Do not use a keytab to authenticate to a peer when
3845 login=NEGOTIATE is specified. Let the GSSAPI
3846 implementation determine which already existing
3847 credentials cache to use instead.
3850 ==== SSL / HTTPS / TLS OPTIONS ====
3852 tls Encrypt connections to this peer with TLS.
3854 sslcert=/path/to/ssl/certificate
3855 A client X.509 certificate to use when connecting to
3858 sslkey=/path/to/ssl/key
3859 The private key corresponding to sslcert above.
3861 If sslkey= is not specified sslcert= is assumed to
3862 reference a PEM file containing both the certificate
3865 sslcipher=... The list of valid SSL ciphers to use when connecting
3869 The minimum TLS protocol version to permit. To control
3870 SSLv3 use the tls-options= parameter.
3871 Supported Values: 1.0 (default), 1.1, 1.2
3873 tls-options=... Specify various TLS implementation options.
3875 OpenSSL options most important are:
3877 NO_SSLv3 Disallow the use of SSLv3
3880 Always create a new key when using
3881 temporary/ephemeral DH key exchanges
3884 Disable use of RFC5077 session tickets.
3885 Some servers may have problems
3886 understanding the TLS extension due
3887 to ambiguous specification in RFC4507.
3889 ALL Enable various bug workarounds
3890 suggested as "harmless" by OpenSSL
3891 Be warned that this reduces SSL/TLS
3892 strength to some attacks.
3894 See the OpenSSL SSL_CTX_set_options documentation for a
3897 GnuTLS options most important are:
3900 Disable use of RFC5077 session tickets.
3901 Some servers may have problems
3902 understanding the TLS extension due
3903 to ambiguous specification in RFC4507.
3905 See the GnuTLS Priority Strings documentation
3906 for a more complete list.
3907 http://www.gnutls.org/manual/gnutls.html#Priority-Strings
3909 tls-cafile= PEM file containing CA certificates to use when verifying
3910 the peer certificate. May be repeated to load multiple files.
3912 sslcapath=... A directory containing additional CA certificates to
3913 use when verifying the peer certificate.
3914 Requires OpenSSL or LibreSSL.
3916 sslcrlfile=... A certificate revocation list file to use when
3917 verifying the peer certificate.
3919 sslflags=... Specify various flags modifying the SSL implementation:
3922 Accept certificates even if they fail to
3926 Don't verify the peer certificate
3927 matches the server name
3929 ssldomain= The peer name as advertised in it's certificate.
3930 Used for verifying the correctness of the received peer
3931 certificate. If not specified the peer hostname will be
3934 front-end-https[=off|on|auto]
3935 Enable the "Front-End-Https: On" header needed when
3936 using Squid as a SSL frontend in front of Microsoft OWA.
3937 See MS KB document Q307347 for details on this header.
3938 If set to auto the header will only be added if the
3939 request is forwarded as a https:// URL.
3941 tls-default-ca[=off]
3942 Whether to use the system Trusted CAs. Default is ON.
3944 tls-no-npn Do not use the TLS NPN extension to advertise HTTP/1.1.
3946 ==== GENERAL OPTIONS ====
3949 A peer-specific connect timeout.
3950 Also see the peer_connect_timeout directive.
3952 connect-fail-limit=N
3953 How many times connecting to a peer must fail before
3954 it is marked as down. Standby connection failures
3955 count towards this limit. Default is 10.
3957 allow-miss Disable Squid's use of only-if-cached when forwarding
3958 requests to siblings. This is primarily useful when
3959 icp_hit_stale is used by the sibling. Excessive use
3960 of this option may result in forwarding loops. One way
3961 to prevent peering loops when using this option, is to
3962 deny cache peer usage on requests from a peer:
3964 cache_peer_access peerName deny fromPeer
3966 max-conn=N Limit the number of concurrent connections the Squid
3967 may open to this peer, including already opened idle
3968 and standby connections. There is no peer-specific
3969 connection limit by default.
3971 A peer exceeding the limit is not used for new
3972 requests unless a standby connection is available.
3974 max-conn currently works poorly with idle persistent
3975 connections: When a peer reaches its max-conn limit,
3976 and there are idle persistent connections to the peer,
3977 the peer may not be selected because the limiting code
3978 does not know whether Squid can reuse those idle
3981 standby=N Maintain a pool of N "hot standby" connections to an
3982 UP peer, available for requests when no idle
3983 persistent connection is available (or safe) to use.
3984 By default and with zero N, no such pool is maintained.
3985 N must not exceed the max-conn limit (if any).
3987 At start or after reconfiguration, Squid opens new TCP
3988 standby connections until there are N connections
3989 available and then replenishes the standby pool as
3990 opened connections are used up for requests. A used
3991 connection never goes back to the standby pool, but
3992 may go to the regular idle persistent connection pool
3993 shared by all peers and origin servers.
3995 Squid never opens multiple new standby connections
3996 concurrently. This one-at-a-time approach minimizes
3997 flooding-like effect on peers. Furthermore, just a few
3998 standby connections should be sufficient in most cases
3999 to supply most new requests with a ready-to-use
4002 Standby connections obey server_idle_pconn_timeout.
4003 For the feature to work as intended, the peer must be
4004 configured to accept and keep them open longer than
4005 the idle timeout at the connecting Squid, to minimize
4006 race conditions typical to idle used persistent
4007 connections. Default request_timeout and
4008 server_idle_pconn_timeout values ensure such a
4011 name=xxx Unique name for the peer.
4012 Required if you have multiple cache_peers with the same hostname.
4013 Defaults to cache_peer hostname when not explicitly specified.
4015 Other directives (e.g., cache_peer_access), cache manager reports,
4016 and cache.log messages use this name to refer to this cache_peer.
4018 The cache_peer name value affects hashing-based peer selection
4019 methods (e.g., carp and sourcehash).
4021 Can be used by outgoing access controls through the
4024 The name value preserves configured spelling, but name uniqueness
4025 checks and name-based search are case-insensitive.
4027 no-tproxy Do not use the client-spoof TPROXY support when forwarding
4028 requests to this peer. Use normal address selection instead.
4029 This overrides the spoof_client_ip ACL.
4031 proxy-only objects fetched from the peer will not be stored locally.
4035 NAME: cache_peer_access
4038 DEFAULT_DOC: No peer usage restrictions.
4041 Restricts usage of cache_peer proxies.
4044 cache_peer_access peer-name allow|deny [!]aclname ...
4046 For the required peer-name parameter, use either the value of the
4047 cache_peer name=value parameter or, if name=value is missing, the
4048 cache_peer hostname parameter.
4050 This directive narrows down the selection of peering candidates, but
4051 does not determine the order in which the selected candidates are
4052 contacted. That order is determined by the peer selection algorithms
4053 (see PEER SELECTION sections in the cache_peer documentation).
4055 If a deny rule matches, the corresponding peer will not be contacted
4056 for the current transaction -- Squid will not send ICP queries and
4057 will not forward HTTP requests to that peer. An allow match leaves
4058 the corresponding peer in the selection. The first match for a given
4059 peer wins for that peer.
4061 The relative order of cache_peer_access directives for the same peer
4062 matters. The relative order of any two cache_peer_access directives
4063 for different peers does not matter. To ease interpretation, it is a
4064 good idea to group cache_peer_access directives for the same peer
4067 A single cache_peer_access directive may be evaluated multiple times
4068 for a given transaction because individual peer selection algorithms
4069 may check it independently from each other. These redundant checks
4070 may be optimized away in future Squid versions.
4072 This clause only supports fast acl types.
4073 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
4077 NAME: neighbor_type_domain
4078 TYPE: hostdomaintype
4080 DEFAULT_DOC: The peer type from cache_peer directive is used for all requests to that peer.
4083 Modify the cache_peer neighbor type when passing requests
4084 about specific domains to the peer.
4087 neighbor_type_domain peer-name parent|sibling domain...
4089 For the required peer-name parameter, use either the value of the
4090 cache_peer name=value parameter or, if name=value is missing, the
4091 cache_peer hostname parameter.
4094 cache_peer foo.example.com parent 3128 3130
4095 neighbor_type_domain foo.example.com sibling .au .de
4097 The above configuration treats all requests to foo.example.com as a
4098 parent proxy unless the request is for a .au or .de ccTLD domain name.
4101 NAME: dead_peer_timeout
4105 LOC: Config.Timeout.deadPeer
4107 This controls how long Squid waits to declare a peer cache
4108 as "dead." If there are no ICP replies received in this
4109 amount of time, Squid will declare the peer dead and not
4110 expect to receive any further ICP replies. However, it
4111 continues to send ICP queries, and will mark the peer as
4112 alive upon receipt of the first subsequent ICP reply.
4114 This timeout also affects when Squid expects to receive ICP
4115 replies from peers. If more than 'dead_peer' seconds have
4116 passed since the last ICP reply was received, Squid will not
4117 expect to receive an ICP reply on the next query. Thus, if
4118 your time between requests is greater than this timeout, you
4119 will see a lot of requests sent DIRECT to origin servers
4120 instead of to your parents.
4123 NAME: forward_max_tries
4126 LOC: Config.forward_max_tries
4128 Limits the number of attempts to forward the request.
4130 For the purpose of this limit, Squid counts all high-level request
4131 forwarding attempts, including any same-destination retries after
4132 certain persistent connection failures and any attempts to use a
4133 different peer. However, these low-level attempts are not counted:
4134 * connection reopening attempts (enabled using connect_retries)
4135 * unfinished Happy Eyeballs connection attempts (prevented by setting
4136 happy_eyeballs_connect_limit to 0)
4138 See also: forward_timeout, connect_retries, and %request_attempts.
4142 MEMORY CACHE OPTIONS
4143 -----------------------------------------------------------------------------
4150 LOC: Config.memMaxSize
4152 NOTE: THIS PARAMETER DOES NOT SPECIFY THE MAXIMUM PROCESS SIZE.
4153 IT ONLY PLACES A LIMIT ON HOW MUCH ADDITIONAL MEMORY SQUID WILL
4154 USE AS A MEMORY CACHE OF OBJECTS. SQUID USES MEMORY FOR OTHER
4155 THINGS AS WELL. SEE THE SQUID FAQ SECTION 8 FOR DETAILS.
4157 'cache_mem' specifies the ideal amount of memory to be used
4159 * In-Transit objects
4161 * Negative-Cached objects
4163 Data for these objects are stored in 4 KB blocks. This
4164 parameter specifies the ideal upper limit on the total size of
4165 4 KB blocks allocated. In-Transit objects take the highest
4168 In-transit objects have priority over the others. When
4169 additional space is needed for incoming data, negative-cached
4170 and hot objects will be released. In other words, the
4171 negative-cached and hot objects will fill up any unused space
4172 not needed for in-transit objects.
4174 If circumstances require, this limit will be exceeded.
4175 Specifically, if your incoming request rate requires more than
4176 'cache_mem' of memory to hold in-transit objects, Squid will
4177 exceed this limit to satisfy the new requests. When the load
4178 decreases, blocks will be freed until the high-water mark is
4179 reached. Thereafter, blocks will be used to store hot
4182 If shared memory caching is enabled, Squid does not use the shared
4183 cache space for in-transit objects, but they still consume as much
4184 local memory as they need. For more details about the shared memory
4185 cache, see memory_cache_shared.
4188 NAME: maximum_object_size_in_memory
4192 LOC: Config.Store.maxInMemObjSize
4194 Objects greater than this size will not be attempted to kept in
4195 the memory cache. This should be set high enough to keep objects
4196 accessed frequently in memory to improve performance whilst low
4197 enough to keep larger objects from hoarding cache_mem.
4200 NAME: memory_cache_shared
4203 LOC: Config.memShared
4205 DEFAULT_DOC: "on" where supported if doing memory caching with multiple SMP workers.
4207 Controls whether the memory cache is shared among SMP workers.
4209 The shared memory cache is meant to occupy cache_mem bytes and replace
4210 the non-shared memory cache, although some entities may still be
4211 cached locally by workers for now (e.g., internal and in-transit
4212 objects may be served from a local memory cache even if shared memory
4213 caching is enabled).
4215 By default, the memory cache is shared if and only if all of the
4216 following conditions are satisfied: Squid runs in SMP mode with
4217 multiple workers, cache_mem is positive, and Squid environment
4218 supports required IPC primitives (e.g., POSIX shared memory segments
4219 and GCC-style atomic operations).
4221 To avoid blocking locks, shared memory uses opportunistic algorithms
4222 that do not guarantee that every cachable entity that could have been
4223 shared among SMP workers will actually be shared.
4226 NAME: memory_cache_mode
4230 DEFAULT_DOC: Keep the most recently fetched objects in memory
4232 Controls which objects to keep in the memory cache (cache_mem)
4234 always Keep most recently fetched objects in memory (default)
4236 disk Only disk cache hits are kept in memory, which means
4237 an object must first be cached on disk and then hit
4238 a second time before cached in memory.
4240 network Only objects fetched from network is kept in memory
4243 NAME: memory_replacement_policy
4245 LOC: Config.memPolicy
4248 The memory replacement policy parameter determines which
4249 objects are purged from memory when memory space is needed.
4251 See cache_replacement_policy for details on algorithms.
4256 -----------------------------------------------------------------------------
4259 NAME: cache_replacement_policy
4261 LOC: Config.replPolicy
4264 The cache replacement policy parameter determines which
4265 objects are evicted (replaced) when disk space is needed.
4267 lru : Squid's original list based LRU policy
4268 heap GDSF : Greedy-Dual Size Frequency
4269 heap LFUDA: Least Frequently Used with Dynamic Aging
4270 heap LRU : LRU policy implemented using a heap
4272 Applies to any cache_dir lines listed below this directive.
4274 The LRU policies keeps recently referenced objects.
4276 The heap GDSF policy optimizes object hit rate by keeping smaller
4277 popular objects in cache so it has a better chance of getting a
4278 hit. It achieves a lower byte hit rate than LFUDA though since
4279 it evicts larger (possibly popular) objects.
4281 The heap LFUDA policy keeps popular objects in cache regardless of
4282 their size and thus optimizes byte hit rate at the expense of
4283 hit rate since one large, popular object will prevent many
4284 smaller, slightly less popular objects from being cached.
4286 Both policies utilize a dynamic aging mechanism that prevents
4287 cache pollution that can otherwise occur with frequency-based
4288 replacement policies.
4290 NOTE: if using the LFUDA replacement policy you should increase
4291 the value of maximum_object_size above its default of 4 MB to
4292 to maximize the potential byte hit rate improvement of LFUDA.
4294 For more information about the GDSF and LFUDA cache replacement
4295 policies see http://www.hpl.hp.com/techreports/1999/HPL-1999-69.html
4296 and http://fog.hpl.external.hp.com/techreports/98/HPL-98-173.html.
4299 NAME: minimum_object_size
4303 DEFAULT_DOC: no limit
4304 LOC: Config.Store.minObjectSize
4306 Objects smaller than this size will NOT be saved on disk. The
4307 value is specified in bytes, and the default is 0 KB, which
4308 means all responses can be stored.
4311 NAME: maximum_object_size
4315 LOC: Config.Store.maxObjectSize
4317 Set the default value for max-size parameter on any cache_dir.
4318 The value is specified in bytes, and the default is 4 MB.
4320 If you wish to get a high BYTES hit ratio, you should probably
4321 increase this (one 32 MB object hit counts for 3200 10KB
4324 If you wish to increase hit ratio more than you want to
4325 save bandwidth you should leave this low.
4327 NOTE: if using the LFUDA replacement policy you should increase
4328 this value to maximize the byte hit rate improvement of LFUDA!
4329 See cache_replacement_policy for a discussion of this policy.
4335 DEFAULT_DOC: No disk cache. Store cache objects only in memory.
4336 LOC: Config.cacheSwap
4339 cache_dir Type Directory-Name Fs-specific-data [options]
4341 You can specify multiple cache_dir lines to spread the
4342 cache among different disk partitions.
4344 Type specifies the kind of storage system to use. Only "ufs"
4345 is built by default. To enable any of the other storage systems
4346 see the --enable-storeio configure option.
4348 'Directory' is a top-level directory where cache swap
4349 files will be stored. If you want to use an entire disk
4350 for caching, this can be the mount-point directory.
4351 The directory must exist and be writable by the Squid
4352 process. Squid will NOT create this directory for you.
4354 Rock is currently the only SMP-aware cache_dir type. Using other
4355 store types in configurations with multiple workers is not
4356 supported and may lead to HTTP violations or undefined behavior,
4357 even when each such cache_dir is given a dedicated worker using
4358 configuration conditionals.
4361 ==== The ufs store type ====
4363 "ufs" is the old well-known Squid storage format that has always
4367 cache_dir ufs Directory-Name Mbytes L1 L2 [options]
4369 'Mbytes' is the amount of disk space (MB) to use under this
4370 directory. The default is 100 MB. Change this to suit your
4371 configuration. Do NOT put the size of your disk drive here.
4372 Instead, if you want Squid to use the entire disk drive,
4373 subtract 20% and use that value.
4375 'L1' is the number of first-level subdirectories which
4376 will be created under the 'Directory'. The default is 16.
4378 'L2' is the number of second-level subdirectories which
4379 will be created under each first-level directory. The default
4383 ==== The aufs store type ====
4385 "aufs" uses the same storage format as "ufs", utilizing
4386 POSIX-threads to avoid blocking the main Squid process on
4387 disk-I/O. This was formerly known in Squid as async-io.
4390 cache_dir aufs Directory-Name Mbytes L1 L2 [options]
4392 see argument descriptions under ufs above
4395 ==== The diskd store type ====
4397 "diskd" uses the same storage format as "ufs", utilizing a
4398 separate process to avoid blocking the main Squid process on
4402 cache_dir diskd Directory-Name Mbytes L1 L2 [options] [Q1=n] [Q2=n]
4404 see argument descriptions under ufs above
4406 Q1 specifies the number of unacknowledged I/O requests when Squid
4407 stops opening new files. If this many messages are in the queues,
4408 Squid won't open new files. Default is 64
4410 Q2 specifies the number of unacknowledged messages when Squid
4411 starts blocking. If this many messages are in the queues,
4412 Squid blocks until it receives some replies. Default is 72
4414 When Q1 < Q2 (the default), the cache directory is optimized
4415 for lower response time at the expense of a decrease in hit
4416 ratio. If Q1 > Q2, the cache directory is optimized for
4417 higher hit ratio at the expense of an increase in response
4421 ==== The rock store type ====
4424 cache_dir rock Directory-Name Mbytes [options]
4426 The Rock Store type is a database-style storage. All cached
4427 entries are stored in a "database" file, using fixed-size slots.
4428 A single entry occupies one or more slots.
4430 If possible, Squid using Rock Store creates a dedicated kid
4431 process called "disker" to avoid blocking Squid worker(s) on disk
4432 I/O. One disker kid is created for each rock cache_dir. Diskers
4433 are created only when Squid, running in daemon mode, has support
4434 for the IpcIo disk I/O module.
4436 swap-timeout=msec: Squid will not start writing a miss to or
4437 reading a hit from disk if it estimates that the swap operation
4438 will take more than the specified number of milliseconds. By
4439 default and when set to zero, disables the disk I/O time limit
4440 enforcement. Ignored when using blocking I/O module because
4441 blocking synchronous I/O does not allow Squid to estimate the
4442 expected swap wait time.
4444 max-swap-rate=swaps/sec: Artificially limits disk access using
4445 the specified I/O rate limit. Swap out requests that
4446 would cause the average I/O rate to exceed the limit are
4447 delayed. Individual swap in requests (i.e., hits or reads) are
4448 not delayed, but they do contribute to measured swap rate and
4449 since they are placed in the same FIFO queue as swap out
4450 requests, they may wait longer if max-swap-rate is smaller.
4451 This is necessary on file systems that buffer "too
4452 many" writes and then start blocking Squid and other processes
4453 while committing those writes to disk. Usually used together
4454 with swap-timeout to avoid excessive delays and queue overflows
4455 when disk demand exceeds available disk "bandwidth". By default
4456 and when set to zero, disables the disk I/O rate limit
4457 enforcement. Currently supported by IpcIo module only.
4459 slot-size=bytes: The size of a database "record" used for
4460 storing cached responses. A cached response occupies at least
4461 one slot and all database I/O is done using individual slots so
4462 increasing this parameter leads to more disk space waste while
4463 decreasing it leads to more disk I/O overheads. Should be a
4464 multiple of your operating system I/O page size. Defaults to
4465 16KBytes. A housekeeping header is stored with each slot and
4466 smaller slot-sizes will be rejected. The header is smaller than
4470 ==== COMMON OPTIONS ====
4472 no-store no new objects should be stored to this cache_dir.
4474 min-size=n the minimum object size in bytes this cache_dir
4475 will accept. It's used to restrict a cache_dir
4476 to only store large objects (e.g. AUFS) while
4477 other stores are optimized for smaller objects
4481 max-size=n the maximum object size in bytes this cache_dir
4483 The value in maximum_object_size directive sets
4484 the default unless more specific details are
4485 available (ie a small store capacity).
4487 Note: To make optimal use of the max-size limits you should order
4488 the cache_dir lines with the smallest max-size value first.
4492 # Uncomment and adjust the following to add a disk cache directory.
4493 #cache_dir ufs @DEFAULT_SWAP_DIR@ 100 16 256
4497 NAME: store_dir_select_algorithm
4499 LOC: Config.store_dir_select_algorithm
4502 How Squid selects which cache_dir to use when the response
4503 object will fit into more than one.
4505 Regardless of which algorithm is used the cache_dir min-size
4506 and max-size parameters are obeyed. As such they can affect
4507 the selection algorithm by limiting the set of considered
4514 This algorithm is suited to caches with similar cache_dir
4515 sizes and disk speeds.
4517 The disk with the least I/O pending is selected.
4518 When there are multiple disks with the same I/O load ranking
4519 the cache_dir with most available capacity is selected.
4521 When a mix of cache_dir sizes are configured the faster disks
4522 have a naturally lower I/O loading and larger disks have more
4523 capacity. So space used to store objects and data throughput
4524 may be very unbalanced towards larger disks.
4529 This algorithm is suited to caches with unequal cache_dir
4532 Each cache_dir is selected in a rotation. The next suitable
4535 Available cache_dir capacity is only considered in relation
4536 to whether the object will fit and meets the min-size and
4537 max-size parameters.
4539 Disk I/O loading is only considered to prevent overload on slow
4540 disks. This algorithm does not spread objects by size, so any
4541 I/O loading per-disk may appear very unbalanced and volatile.
4543 If several cache_dirs use similar min-size, max-size, or other
4544 limits to to reject certain responses, then do not group such
4545 cache_dir lines together, to avoid round-robin selection bias
4546 towards the first cache_dir after the group. Instead, interleave
4547 cache_dir lines from different groups. For example:
4549 store_dir_select_algorithm round-robin
4550 cache_dir rock /hdd1 ... min-size=100000
4551 cache_dir rock /ssd1 ... max-size=99999
4552 cache_dir rock /hdd2 ... min-size=100000
4553 cache_dir rock /ssd2 ... max-size=99999
4554 cache_dir rock /hdd3 ... min-size=100000
4555 cache_dir rock /ssd3 ... max-size=99999
4558 NAME: paranoid_hit_validation
4559 COMMENT: time-units-small
4560 TYPE: time_nanoseconds
4562 DEFAULT_DOC: validation disabled
4563 LOC: Config.paranoid_hit_validation
4565 Controls whether Squid should perform paranoid validation of cache entry
4566 metadata integrity every time a cache entry is hit. This low-level
4567 validation should always succeed. Each failed validation results in a
4568 cache miss, a BUG line reported to cache.log, and the invalid entry
4569 marked as unusable (and eventually purged from the cache).
4571 Squid can only validate shared cache memory and rock cache_dir entries.
4573 * Zero (default) value means that the validation is disabled.
4575 * Positive values enable validation:
4576 - values less than 1 day approximate the maximum time that Squid is allowed
4577 to spend validating a single cache hit.
4578 - values greater or equal to 1 day are considered as no limitation:
4579 in this case all checks will be performed, regardless of how much time
4582 Hits are usually stored using 16KB slots (for rock, the size is
4583 configurable via cache_dir slot-size). Larger hits require scanning more
4584 slots and, hence, take more time. When validation is enabled, at least one
4585 slot is always validated, regardless of the configured time limit.
4587 A worker process validating an entry cannot do anything else (i.e. the
4588 validation is blocking). The validation overhead is environment dependent,
4589 but developers have observed Squid spending 3-10 microseconds to check each
4590 slot of a Rock or shared memory hit entry. If Squid cuts validation short
4591 because it runs out of configured time, it treats the entry as valid.
4593 When hit validation is enabled, its statistics is included in Cache
4594 Manager mgr:counters, mgr:5min, and mgr:60min reports.
4597 NAME: max_open_disk_fds
4599 LOC: Config.max_open_disk_fds
4601 DEFAULT_DOC: no limit
4603 To avoid having disk as the I/O bottleneck Squid can optionally
4604 bypass the on-disk cache if more than this amount of disk file
4605 descriptors are open.
4607 A value of 0 indicates no limit.
4610 NAME: cache_swap_low
4611 COMMENT: (percent, 0-100)
4614 LOC: Config.Swap.lowWaterMark
4616 The low-water mark for AUFS/UFS/diskd cache object eviction by
4617 the cache_replacement_policy algorithm.
4619 Removal begins when the swap (disk) usage of a cache_dir is
4620 above this low-water mark and attempts to maintain utilization
4621 near the low-water mark.
4623 As swap utilization increases towards the high-water mark set
4624 by cache_swap_high object eviction becomes more aggressive.
4626 The value difference in percentages between low- and high-water
4627 marks represent an eviction rate of 300 objects per second and
4628 the rate continues to scale in aggressiveness by multiples of
4629 this above the high-water mark.
4631 Defaults are 90% and 95%. If you have a large cache, 5% could be
4632 hundreds of MB. If this is the case you may wish to set these
4633 numbers closer together.
4635 See also cache_swap_high and cache_replacement_policy
4638 NAME: cache_swap_high
4639 COMMENT: (percent, 0-100)
4642 LOC: Config.Swap.highWaterMark
4644 The high-water mark for AUFS/UFS/diskd cache object eviction by
4645 the cache_replacement_policy algorithm.
4647 Removal begins when the swap (disk) usage of a cache_dir is
4648 above the low-water mark set by cache_swap_low and attempts to
4649 maintain utilization near the low-water mark.
4651 As swap utilization increases towards this high-water mark object
4652 eviction becomes more aggressive.
4654 The value difference in percentages between low- and high-water
4655 marks represent an eviction rate of 300 objects per second and
4656 the rate continues to scale in aggressiveness by multiples of
4657 this above the high-water mark.
4659 Defaults are 90% and 95%. If you have a large cache, 5% could be
4660 hundreds of MB. If this is the case you may wish to set these
4661 numbers closer together.
4663 See also cache_swap_low and cache_replacement_policy
4668 -----------------------------------------------------------------------------
4675 DEFAULT_DOC: The format definitions squid, common, combined, referrer, useragent are built in.
4679 logformat <name> <format specification>
4681 Defines an access log format.
4683 The <format specification> is a string with embedded % format codes
4685 % format codes all follow the same basic structure where all
4686 components but the formatcode are optional and usually unnecessary,
4687 especially when dealing with common codes.
4689 % [encoding] [-] [[0]width] [{arg}] formatcode [{arg}]
4691 encoding escapes or otherwise protects "special" characters:
4693 " Quoted string encoding where quote(") and
4694 backslash(\) characters are \-escaped while
4695 CR, LF, and TAB characters are encoded as \r,
4696 \n, and \t two-character sequences.
4698 [ Custom Squid encoding where percent(%), square
4699 brackets([]), backslash(\) and characters with
4700 codes outside of [32,126] range are %-encoded.
4701 SP is not encoded. Used by log_mime_hdrs.
4703 # URL encoding (a.k.a. percent-encoding) where
4704 all URL unsafe and control characters (per RFC
4705 1738) are %-encoded.
4707 / Shell-like encoding where quote(") and
4708 backslash(\) characters are \-escaped while CR
4709 and LF characters are encoded as \r and \n
4710 two-character sequences. Values containing SP
4711 character(s) are surrounded by quotes(").
4713 ' Raw/as-is encoding with no escaping/quoting.
4715 Default encoding: When no explicit encoding is
4716 specified, each %code determines its own encoding.
4717 Most %codes use raw/as-is encoding, but some codes use
4718 a so called "pass-through URL encoding" where all URL
4719 unsafe and control characters (per RFC 1738) are
4720 %-encoded, but the percent character(%) is left as is.
4724 width minimum and/or maximum field width:
4725 [width_min][.width_max]
4726 When minimum starts with 0, the field is zero-padded.
4727 String values exceeding maximum width are truncated.
4729 {arg} argument such as header name etc. This field may be
4730 placed before or after the token, but not both at once.
4734 % a literal % character
4736 byte{value} Adds a single byte with the given value (e.g., %byte{10}
4737 adds an ASCII LF character a.k.a. "new line" or "\n"). The value
4738 parameter is required and must be a positive decimal integer not
4739 exceeding 255. Zero-valued bytes (i.e. ASCII NUL characters) are
4742 sn Unique sequence number per log line entry
4743 err_code The ID of an error response served by Squid or
4744 a similar internal error identifier.
4746 err_detail Additional err_code-dependent error information. Multiple
4747 details are separated by the plus sign ('+'). Admins should not
4748 rely on a particular detail listing order, the uniqueness of the
4749 entries, or individual detail text stability. All those properties
4750 depend on many unstable factors, including external libraries.
4752 note The annotation specified by the argument. Also
4753 logs the adaptation meta headers set by the
4754 adaptation_meta configuration parameter.
4755 If no argument given all annotations logged.
4756 The argument may include a separator to use with
4759 By default, multiple note values are separated with ","
4760 and multiple notes are separated with "\r\n".
4761 When logging named notes with %{name}note, the
4762 explicitly configured separator is used between note
4763 values. When logging all notes with %note, the
4764 explicitly configured separator is used between
4765 individual notes. There is currently no way to
4766 specify both value and notes separators when logging
4767 all notes with %note.
4768 master_xaction The master transaction identifier is an unsigned
4769 integer. These IDs are guaranteed to monotonically
4770 increase within a single worker process lifetime, with
4771 higher values corresponding to transactions that were
4772 accepted or initiated later. Due to current implementation
4773 deficiencies, some IDs are skipped (i.e. never logged).
4774 Concurrent workers and restarted workers use similar,
4775 overlapping sequences of master transaction IDs.
4777 Connection related format codes:
4779 >a Client source IP address
4781 >p Client source port
4782 >eui Client source EUI (MAC address, EUI-48 or EUI-64 identifier)
4783 >la Local IP address the client connected to
4784 >lp Local port number the client connected to
4785 >qos Client connection TOS/DSCP value set by Squid
4786 >nfmark Client connection netfilter packet MARK set by Squid
4788 transport::>connection_id Identifies a transport connection
4789 accepted by Squid (e.g., a connection carrying the
4790 logged HTTP request). Currently, Squid only supports
4791 TCP transport connections.
4793 The logged identifier is an unsigned integer. These
4794 IDs are guaranteed to monotonically increase within a
4795 single worker process lifetime, with higher values
4796 corresponding to connections that were accepted later.
4797 Many IDs are skipped (i.e. never logged). Concurrent
4798 workers and restarted workers use similar, partially
4799 overlapping sequences of IDs.
4801 la Local listening IP address the client connection was connected to.
4802 lp Local listening port number the client connection was connected to.
4804 <a Server IP address of the last server or peer connection
4805 <A Server FQDN or peer name
4806 <p Server port number of the last server or peer connection
4807 <la Local IP address of the last server or peer connection
4808 <lp Local port number of the last server or peer connection
4809 <qos Server connection TOS/DSCP value set by Squid
4810 <nfmark Server connection netfilter packet MARK set by Squid
4812 >handshake Raw client handshake
4813 Initial client bytes received by Squid on a newly
4814 accepted TCP connection or inside a just established
4815 CONNECT tunnel. Squid stops accumulating handshake
4816 bytes as soon as the handshake parser succeeds or
4817 fails (determining whether the client is using the
4820 For HTTP clients, the handshake is the request line.
4821 For TLS clients, the handshake consists of all TLS
4822 records up to and including the TLS record that
4823 contains the last byte of the first ClientHello
4824 message. For clients using an unsupported protocol,
4825 this field contains the bytes received by Squid at the
4826 time of the handshake parsing failure.
4828 See the on_unsupported_protocol directive for more
4829 information on Squid handshake traffic expectations.
4831 Current support is limited to these contexts:
4832 - http_port connections, but only when the
4833 on_unsupported_protocol directive is in use.
4834 - https_port connections (and CONNECT tunnels) that
4835 are subject to the ssl_bump peek or stare action.
4837 To protect binary handshake data, this field is always
4838 base64-encoded (RFC 4648 Section 4). If logformat
4839 field encoding is configured, that encoding is applied
4840 on top of base64. Otherwise, the computed base64 value
4843 Time related format codes:
4845 ts Seconds since epoch
4846 tu subsecond time (milliseconds)
4847 tl Local time. Optional strftime format argument
4848 default %d/%b/%Y:%H:%M:%S %z
4849 tg GMT time. Optional strftime format argument
4850 default %d/%b/%Y:%H:%M:%S %z
4851 tr Response time (milliseconds)
4852 dt Total time spent making DNS lookups (milliseconds)
4853 tS Approximate master transaction start time in
4854 <full seconds since epoch>.<fractional seconds> format.
4855 Currently, Squid considers the master transaction
4856 started when a complete HTTP request header initiating
4857 the transaction is received from the client. This is
4858 the same value that Squid uses to calculate transaction
4859 response time when logging %tr to access.log. Currently,
4860 Squid uses millisecond resolution for %tS values,
4861 similar to the default access.log "current time" field
4864 busy_time Time spent in transaction-related code (nanoseconds)
4865 This cumulative measurement excludes periods of time when the
4866 transaction was waiting (e.g., for a server or helper response)
4867 while Squid worked on other transactions or was engaged in
4868 transaction-unrelated activities (e.g., generating a cache index).
4869 In other words, this measurement represents the total amount of
4870 physical time when Squid was busy working on this transaction.
4872 WARNING: This measurement relies on Squid transaction context
4873 tracking features that currently have known context leak bugs and
4874 coverage gaps. Until those features are fully implemented, logged
4875 values may significantly understate or exaggerate actual times.
4876 Do not use this measurement unless you know it works in your case.
4878 Access Control related format codes:
4880 et Tag returned by external acl
4881 ea Log string returned by external acl
4882 un User name (any available)
4883 ul User name from authentication
4884 ue User name from external acl helper
4885 un A user name. Expands to the first available name
4886 from the following list of information sources:
4887 - authenticated user name, like %ul
4888 - user name supplied by an external ACL, like %ue
4889 - SSL client name, like %us
4890 credentials Client credentials. The exact meaning depends on
4891 the authentication scheme: For Basic authentication,
4892 it is the password; for Digest, the realm sent by the
4893 client; for NTLM and Negotiate, the client challenge
4894 or client credentials prefixed with "YR " or "KK ".
4896 HTTP related format codes:
4900 [http::]rm Request method (GET/POST etc)
4901 [http::]>rm Request method from client
4902 [http::]<rm Request method sent to server or peer
4904 [http::]ru Request URL received (or computed) and sanitized
4906 Logs request URI received from the client, a
4907 request adaptation service, or a request
4908 redirector (whichever was applied last).
4910 Computed URLs are URIs of internally generated
4911 requests and various "error:..." URIs.
4913 Honors strip_query_terms and uri_whitespace.
4915 This field is not encoded by default. Encoding
4916 this field using variants of %-encoding will
4917 clash with uri_whitespace modifications that
4918 also use %-encoding.
4920 [http::]>ru Request URL received from the client (or computed)
4922 Computed URLs are URIs of internally generated
4923 requests and various "error:..." URIs.
4925 Unlike %ru, this request URI is not affected
4926 by request adaptation, URL rewriting services,
4927 and strip_query_terms.
4929 Honors uri_whitespace.
4931 This field is using pass-through URL encoding
4932 by default. Encoding this field using other
4933 variants of %-encoding will clash with
4934 uri_whitespace modifications that also use
4937 [http::]<ru Request URL sent to server or peer
4938 [http::]>rs Request URL scheme from client
4939 [http::]<rs Request URL scheme sent to server or peer
4940 [http::]>rd Request URL domain from client
4941 [http::]<rd Request URL domain sent to server or peer
4942 [http::]>rP Request URL port from client
4943 [http::]<rP Request URL port sent to server or peer
4944 [http::]rp Request URL path excluding hostname
4945 [http::]>rp Request URL path excluding hostname from client
4946 [http::]<rp Request URL path excluding hostname sent to server or peer
4947 [http::]rv Request protocol version
4948 [http::]>rv Request protocol version from client
4949 [http::]<rv Request protocol version sent to server or peer
4951 [http::]>h Original received request header.
4952 Usually differs from the request header sent by
4953 Squid, although most fields are often preserved.
4954 Accepts optional header field name/value filter
4955 argument using name[:[separator]element] format.
4956 [http::]>ha Received request header after adaptation and
4957 redirection (pre-cache REQMOD vectoring point).
4958 Usually differs from the request header sent by
4959 Squid, although most fields are often preserved.
4960 Optional header name argument as for >h
4964 [http::]<Hs HTTP status code received from the next hop
4965 [http::]>Hs HTTP status code sent to the client
4967 [http::]<h Reply header. Optional header name argument
4970 [http::]mt MIME content type
4975 [http::]st Total size of request + reply traffic with client
4976 [http::]>st Total size of request received from client.
4977 Excluding chunked encoding bytes.
4978 [http::]<st Total size of reply sent to client (after adaptation)
4980 [http::]>sh Size of request headers received from client
4981 [http::]<sh Size of reply headers sent to client (after adaptation)
4983 [http::]<sH Reply high offset sent
4984 [http::]<sS Upstream object size
4986 [http::]<bs Number of HTTP-equivalent message body bytes
4987 received from the next hop, excluding chunked
4988 transfer encoding and control messages.
4989 Generated FTP listings are treated as
4994 [http::]<pt Peer response time in milliseconds. The timer starts
4995 when the last request byte is sent to the next hop
4996 and stops when the last response byte is received.
4997 [http::]<tt Total time spent forwarding to origin servers or
4998 cache_peers (milliseconds).
5000 The timer starts when Squid decides to forward the request (to
5001 an origin server or cache_peer) and peer selection begins. The
5002 timer stops when relevant forwarding activities (including any
5005 Between those two timer events, Squid may perform DNS lookups,
5006 query external ACL helpers, adapt responses using pre-cache
5007 RESPMOD services, and participate in other concurrent
5008 secondary activities. Most secondary activities increase
5009 peering time. In some cases, a secondary activity may start
5010 before the timer starts or end after the timer stops, leading
5011 to misleading results of simple computations like %<tt - %dt.
5013 If this logformat %code is used before its timer starts, the
5014 corresponding measurement has no value (and the %code expands
5015 to a single dash ("-") character).
5017 If this code is used while its timer is running, the time
5018 spent so far is used as the measurement value.
5020 When Squid re-forwards the request (e.g., after certain cache
5021 revalidation failures), the timer may restart. In this case,
5022 the new measurement is added to the value accumulated from
5023 previous forwarding attempts. The time interval between
5024 forwarding attempts is not added to the final result.
5026 Squid handling related format codes:
5028 Ss Squid request status (TCP_MISS etc)
5029 Sh Squid hierarchy status (DEFAULT_PARENT etc)
5031 [http::]request_attempts Number of request forwarding attempts
5033 See forward_max_tries documentation that details what Squid counts
5034 as a forwarding attempt. Pure cache hits log zero, but cache hits
5035 that triggered HTTP cache revalidation log the number of attempts
5036 made when sending an internal revalidation request. DNS, ICMP,
5037 ICP, HTCP, ICAP, eCAP, helper, and other secondary requests
5038 sent by Squid as a part of a master transaction do not increment
5039 the counter logged for the received request.
5041 SSL-related format codes:
5043 ssl::bump_mode SslBump decision for the transaction:
5045 For CONNECT requests that initiated bumping of
5046 a connection and for any request received on
5047 an already bumped connection, Squid logs the
5048 corresponding SslBump mode ("splice", "bump",
5049 "peek", "stare", "terminate", "server-first"
5050 or "client-first"). See the ssl_bump option
5051 for more information about these modes.
5053 A "none" token is logged for requests that
5054 triggered "ssl_bump" ACL evaluation matching
5057 In all other cases, a single dash ("-") is
5060 ssl::>sni SSL client SNI sent to Squid.
5063 The Subject field of the received client
5064 SSL certificate or a dash ('-') if Squid has
5065 received an invalid/malformed certificate or
5066 no certificate at all. Consider encoding the
5067 logged value because Subject often has spaces.
5070 The Issuer field of the received client
5071 SSL certificate or a dash ('-') if Squid has
5072 received an invalid/malformed certificate or
5073 no certificate at all. Consider encoding the
5074 logged value because Issuer often has spaces.
5077 The Subject field of the received server
5078 TLS certificate or a dash ('-') if this is
5079 not available. Consider encoding the logged
5080 value because Subject often has spaces.
5083 The Issuer field of the received server
5084 TLS certificate or a dash ('-') if this is
5085 not available. Consider encoding the logged
5086 value because Issuer often has spaces.
5089 The received server x509 certificate in PEM
5090 format, including BEGIN and END lines (or a
5091 dash ('-') if the certificate is unavailable).
5093 WARNING: Large certificates will exceed the
5094 current 8KB access.log record limit, resulting
5095 in truncated records. Such truncation usually
5096 happens in the middle of a record field. The
5097 limit applies to all access logging modules.
5099 The logged certificate may have failed
5100 validation and may not be trusted by Squid.
5101 This field does not include any intermediate
5102 certificates that may have been received from
5103 the server or fetched during certificate
5106 Currently, Squid only collects server
5107 certificates during step3 of SslBump
5108 processing; connections that were not subject
5109 to ssl_bump rules or that did not match a peek
5110 or stare rule at step2 will not have the
5111 server certificate information.
5113 This field is using pass-through URL encoding
5117 The list of certificate validation errors
5118 detected by Squid (including OpenSSL and
5119 certificate validation helper components). The
5120 errors are listed in the discovery order. By
5121 default, the error codes are separated by ':'.
5122 Accepts an optional separator argument.
5124 %ssl::>negotiated_version The negotiated TLS version of the
5127 %ssl::<negotiated_version The negotiated TLS version of the
5128 last server or peer connection.
5130 %ssl::>received_hello_version The TLS version of the Hello
5131 message received from TLS client.
5133 %ssl::<received_hello_version The TLS version of the Hello
5134 message received from TLS server.
5136 %ssl::>received_supported_version The maximum TLS version
5137 supported by the TLS client.
5139 %ssl::<received_supported_version The maximum TLS version
5140 supported by the TLS server.
5142 %ssl::>negotiated_cipher The negotiated cipher of the
5145 %ssl::<negotiated_cipher The negotiated cipher of the
5146 last server or peer connection.
5148 If ICAP is enabled, the following code becomes available (as
5149 well as ICAP log codes documented with the icap_log option):
5151 icap::tt Total ICAP "blocking" time for the HTTP transaction. The
5152 timer ticks while Squid checks adaptation_access and while
5153 ICAP transaction(s) expect ICAP response headers, including
5154 the embedded adapted HTTP message headers (where applicable).
5155 This measurement is meant to estimate ICAP impact on HTTP
5156 transaction response times, but it does not currently account
5157 for slow ICAP response body delivery blocking HTTP progress.
5159 Once Squid receives the final ICAP response headers (e.g.,
5160 ICAP 200 or 204) and the associated adapted HTTP message
5161 headers (if any) from the ICAP service, the corresponding ICAP
5162 transaction stops affecting this measurement, even though the
5163 transaction itself may continue for a long time (e.g., to
5164 finish sending the ICAP request and/or to finish receiving the
5165 ICAP response body).
5167 When "blocking" sections of multiple concurrent ICAP
5168 transactions overlap in time, the overlapping segment is
5171 To see complete ICAP transaction response times (rather than
5172 the cumulative effect of their blocking sections) use the
5173 %adapt::all_trs logformat code or the icap_log directive.
5175 If adaptation is enabled the following codes become available:
5177 adapt::<last_h The header of the last ICAP response or
5178 meta-information from the last eCAP
5179 transaction related to the HTTP transaction.
5180 Like <h, accepts an optional header name
5183 adapt::sum_trs Summed adaptation transaction response
5184 times recorded as a comma-separated list in
5185 the order of transaction start time. Each time
5186 value is recorded as an integer number,
5187 representing response time of one or more
5188 adaptation (ICAP or eCAP) transaction in
5189 milliseconds. When a failed transaction is
5190 being retried or repeated, its time is not
5191 logged individually but added to the
5192 replacement (next) transaction. Lifetimes of individually
5193 listed adaptation transactions may overlap.
5194 See also: %icap::tt and %adapt::all_trs.
5196 adapt::all_trs All adaptation transaction response times.
5197 Same as %adapt::sum_trs but response times of
5198 individual transactions are never added
5199 together. Instead, all transaction response
5200 times are recorded individually.
5202 You can prefix adapt::*_trs format codes with adaptation
5203 service name in curly braces to record response time(s) specific
5204 to that service. For example: %{my_service}adapt::sum_trs
5206 Format codes related to the PROXY protocol:
5208 proxy_protocol::>h PROXY protocol header, including optional TLVs.
5210 Supports the same field and element reporting/extraction logic
5211 as %http::>h. For configuration and reporting purposes, Squid
5212 maps each PROXY TLV to an HTTP header field: the TLV type
5213 (configured as a decimal integer) is the field name, and the
5214 TLV value is the field value. All TLVs of "LOCAL" connections
5215 (in PROXY protocol terminology) are currently skipped/ignored.
5217 Squid also maps the following standard PROXY protocol header
5218 blocks to pseudo HTTP headers (their names use PROXY
5219 terminology and start with a colon, following HTTP tradition
5220 for pseudo headers): :command, :version, :src_addr, :dst_addr,
5221 :src_port, and :dst_port.
5223 Without optional parameters, this logformat code logs
5224 pseudo headers and TLVs.
5226 This format code uses pass-through URL encoding by default.
5229 # relay custom PROXY TLV #224 to adaptation services
5230 adaptation_meta Client-Foo "%proxy_protocol::>h{224}"
5234 The default formats available (which do not need re-defining) are:
5236 logformat squid %ts.%03tu %6tr %>a %Ss/%03>Hs %<st %rm %ru %[un %Sh/%<a %mt
5237 logformat common %>a - %[un [%tl] "%rm %ru HTTP/%rv" %>Hs %<st %Ss:%Sh
5238 logformat combined %>a - %[un [%tl] "%rm %ru HTTP/%rv" %>Hs %<st "%{Referer}>h" "%{User-Agent}>h" %Ss:%Sh
5239 logformat referrer %ts.%03tu %>a %{Referer}>h %ru
5240 logformat useragent %>a [%tl] "%{User-Agent}>h"
5242 NOTE: When the log_mime_hdrs directive is set to ON.
5243 The squid, common and combined formats have a safely encoded copy
5244 of the mime headers appended to each line within a pair of brackets.
5246 NOTE: The common and combined formats are not quite true to the Apache definition.
5247 The logs from Squid contain an extra status and hierarchy code appended.
5251 NAME: access_log cache_access_log
5253 LOC: Config.Log.accesslogs
5254 DEFAULT_IF_NONE: daemon:@DEFAULT_ACCESS_LOG@ squid
5256 Configures whether and how Squid logs HTTP and ICP transactions.
5257 If access logging is enabled, a single line is logged for every
5258 matching HTTP or ICP request. The recommended directive formats are:
5260 access_log <module>:<place> [option ...] [acl acl ...]
5261 access_log none [acl acl ...]
5263 The following directive format is accepted but may be deprecated:
5264 access_log <module>:<place> [<logformat name> [acl acl ...]]
5266 In most cases, the first ACL name must not contain the '=' character
5267 and should not be equal to an existing logformat name. You can always
5268 start with an 'all' ACL to work around those restrictions.
5270 Will log to the specified module:place using the specified format (which
5271 must be defined in a logformat directive) those entries which match
5272 ALL the acl's specified (which must be defined in acl clauses).
5273 If no acl is specified, all requests will be logged to this destination.
5275 ===== Available options for the recommended directive format =====
5277 logformat=name Names log line format (either built-in or
5278 defined by a logformat directive). Defaults
5281 buffer-size=64KB Defines approximate buffering limit for log
5282 records (see buffered_logs). Squid should not
5283 keep more than the specified size and, hence,
5284 should flush records before the buffer becomes
5285 full to avoid overflows under normal
5286 conditions (the exact flushing algorithm is
5287 module-dependent though). The on-error option
5288 controls overflow handling.
5290 on-error=die|drop Defines action on unrecoverable errors. The
5291 'drop' action ignores (i.e., does not log)
5292 affected log records. The default 'die' action
5293 kills the affected worker. The drop action
5294 support has not been tested for modules other
5297 rotate=N Specifies the number of log file rotations to
5298 make when you run 'squid -k rotate'. The default
5299 is to obey the logfile_rotate directive. Setting
5300 rotate=0 will disable the file name rotation,
5301 but the log files are still closed and re-opened.
5302 This will enable you to rename the logfiles
5303 yourself just before sending the rotate signal.
5304 Only supported by the stdio module.
5306 ===== Modules Currently available =====
5308 none Do not log any requests matching these ACL.
5309 Do not specify Place or logformat name.
5311 stdio Write each log line to disk immediately at the completion of
5313 Place: the filename and path to be written.
5315 daemon Very similar to stdio. But instead of writing to disk the log
5316 line is passed to a daemon helper for asynchronous handling instead.
5317 Place: varies depending on the daemon.
5319 log_file_daemon Place: the file name and path to be written.
5321 syslog To log each request via syslog facility.
5322 Place: The syslog facility and priority level for these entries.
5323 Place Format: facility.priority
5325 where facility could be any of:
5326 authpriv, daemon, local0 ... local7 or user.
5328 And priority could be any of:
5329 err, warning, notice, info, debug.
5331 udp To send each log line as text data to a UDP receiver.
5332 Place: The destination host name or IP and port.
5333 Place Format: //host:port
5335 tcp To send each log line as text data to a TCP receiver.
5336 Lines may be accumulated before sending (see buffered_logs).
5337 Place: The destination host name or IP and port.
5338 Place Format: //host:port
5341 access_log daemon:@DEFAULT_ACCESS_LOG@ squid
5347 LOC: Config.Log.icaplogs
5350 ICAP log files record ICAP transaction summaries, one line per
5353 The icap_log option format is:
5354 icap_log <filepath> [<logformat name> [acl acl ...]]
5355 icap_log none [acl acl ...]]
5357 Please see access_log option documentation for details. The two
5358 kinds of logs share the overall configuration approach and many
5361 ICAP processing of a single HTTP message or transaction may
5362 require multiple ICAP transactions. In such cases, multiple
5363 ICAP transaction log lines will correspond to a single access
5366 ICAP log supports many access.log logformat %codes. In ICAP context,
5367 HTTP message-related %codes are applied to the HTTP message embedded
5368 in an ICAP message. Logformat "%http::>..." codes are used for HTTP
5369 messages embedded in ICAP requests while "%http::<..." codes are used
5370 for HTTP messages embedded in ICAP responses. For example:
5372 http::>h To-be-adapted HTTP message headers sent by Squid to
5373 the ICAP service. For REQMOD transactions, these are
5374 HTTP request headers. For RESPMOD, these are HTTP
5375 response headers, but Squid currently cannot log them
5376 (i.e., %http::>h will expand to "-" for RESPMOD).
5378 http::<h Adapted HTTP message headers sent by the ICAP
5379 service to Squid (i.e., HTTP request headers in regular
5380 REQMOD; HTTP response headers in RESPMOD and during
5381 request satisfaction in REQMOD).
5383 ICAP OPTIONS transactions do not embed HTTP messages.
5385 Several logformat codes below deal with ICAP message bodies. An ICAP
5386 message body, if any, typically includes a complete HTTP message
5387 (required HTTP headers plus optional HTTP message body). When
5388 computing HTTP message body size for these logformat codes, Squid
5389 either includes or excludes chunked encoding overheads; see
5390 code-specific documentation for details.
5392 For Secure ICAP services, all size-related information is currently
5393 computed before/after TLS encryption/decryption, as if TLS was not
5396 The following format codes are also available for ICAP logs:
5398 icap::<A ICAP server IP address. Similar to <A.
5400 icap::<service_name ICAP service name from the icap_service
5401 option in Squid configuration file.
5403 icap::ru ICAP Request-URI. Similar to ru.
5405 icap::rm ICAP request method (REQMOD, RESPMOD, or
5406 OPTIONS). Similar to existing rm.
5408 icap::>st The total size of the ICAP request sent to the ICAP
5409 server (ICAP headers + ICAP body), including chunking
5412 icap::<st The total size of the ICAP response received from the
5413 ICAP server (ICAP headers + ICAP body), including
5414 chunking metadata (if any).
5416 icap::<bs The size of the ICAP response body received from the
5417 ICAP server, excluding chunking metadata (if any).
5419 icap::tr Transaction response time (in
5420 milliseconds). The timer starts when
5421 the ICAP transaction is created and
5422 stops when the transaction is completed.
5425 icap::tio Transaction I/O time (in milliseconds). The
5426 timer starts when the first ICAP request
5427 byte is scheduled for sending. The timers
5428 stops when the last byte of the ICAP response
5431 icap::to Transaction outcome: ICAP_ERR* for all
5432 transaction errors, ICAP_OPT for OPTION
5433 transactions, ICAP_ECHO for 204
5434 responses, ICAP_MOD for message
5435 modification, and ICAP_SAT for request
5436 satisfaction. Similar to Ss.
5438 icap::Hs ICAP response status code. Similar to Hs.
5440 icap::>h ICAP request header(s). Similar to >h.
5442 icap::<h ICAP response header(s). Similar to <h.
5444 The default ICAP log format, which can be used without an explicit
5445 definition, is called icap_squid:
5447 logformat icap_squid %ts.%03tu %6icap::tr %>A %icap::to/%03icap::Hs %icap::<st %icap::rm %icap::ru %un -/%icap::<A -
5449 See also: logformat and %adapt::<last_h
5452 NAME: logfile_daemon
5454 DEFAULT: @DEFAULT_LOGFILED@
5455 LOC: Log::TheConfig.logfile_daemon
5457 Specify the path to the logfile-writing daemon. This daemon is
5458 used to write the access and store logs, if configured.
5460 Squid sends a number of commands to the log daemon:
5461 L<data>\n - logfile data
5466 r<n>\n - set rotate count to <n>
5467 b<n>\n - 1 = buffer output, 0 = don't buffer output
5469 No responses is expected.
5472 NAME: stats_collection
5474 LOC: Config.accessList.stats_collection
5476 DEFAULT_DOC: Allow logging for all transactions.
5477 COMMENT: allow|deny acl acl...
5479 This options allows you to control which requests gets accounted
5480 in performance counters.
5482 This clause only supports fast acl types.
5483 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
5486 NAME: cache_store_log
5489 LOC: Config.Log.store
5491 Logs the activities of the storage manager. Shows which
5492 objects are ejected from the cache, and which objects are
5493 saved and for how long.
5494 There are not really utilities to analyze this data, so you can safely
5495 disable it (the default).
5497 Store log uses modular logging outputs. See access_log for the list
5498 of modules supported.
5501 cache_store_log stdio:@DEFAULT_STORE_LOG@
5502 cache_store_log daemon:@DEFAULT_STORE_LOG@
5505 NAME: cache_swap_state cache_swap_log
5507 LOC: Config.Log.swap
5509 DEFAULT_DOC: Store the journal inside its cache_dir
5511 Location for the cache "swap.state" file. This index file holds
5512 the metadata of objects saved on disk. It is used to rebuild
5513 the cache during startup. Normally this file resides in each
5514 'cache_dir' directory, but you may specify an alternate
5515 pathname here. Note you must give a full filename, not just
5516 a directory. Since this is the index for the whole object
5517 list you CANNOT periodically rotate it!
5519 If %s can be used in the file name it will be replaced with a
5520 a representation of the cache_dir name where each / is replaced
5521 with '.'. This is needed to allow adding/removing cache_dir
5522 lines when cache_swap_log is being used.
5524 If have more than one 'cache_dir', and %s is not used in the name
5525 these swap logs will have names such as:
5531 The numbered extension (which is added automatically)
5532 corresponds to the order of the 'cache_dir' lines in this
5533 configuration file. If you change the order of the 'cache_dir'
5534 lines in this file, these index files will NOT correspond to
5535 the correct 'cache_dir' entry (unless you manually rename
5536 them). We recommend you do NOT use this option. It is
5537 better to keep these index files in each 'cache_dir' directory.
5540 NAME: logfile_rotate
5543 LOC: Config.Log.rotateNumber
5545 Specifies the default number of logfile rotations to make when you
5546 type 'squid -k rotate'. The default is 10, which will rotate
5547 with extensions 0 through 9. Setting logfile_rotate to 0 will
5548 disable the file name rotation, but the logfiles are still closed
5549 and re-opened. This will enable you to rename the logfiles
5550 yourself just before sending the rotate signal.
5552 Note, from Squid-3.1 this option is only a default for cache.log,
5553 that log can be rotated separately by using debug_options.
5555 Note, from Squid-4 this option is only a default for access.log
5556 recorded by stdio: module. Those logs can be rotated separately by
5557 using the rotate=N option on their access_log directive.
5559 Note, the 'squid -k rotate' command normally sends a USR1
5560 signal to the running squid process. In certain situations
5561 (e.g. on Linux with Async I/O), USR1 is used for other
5562 purposes, so -k rotate uses another signal. It is best to get
5563 in the habit of using 'squid -k rotate' instead of 'kill -USR1
5570 DEFAULT: @DEFAULT_MIME_TABLE@
5571 LOC: Config.mimeTablePathname
5573 Path to Squid's icon configuration file.
5575 You shouldn't need to change this, but the default file contains
5576 examples and formatting information if you do.
5582 LOC: Config.onoff.log_mime_hdrs
5585 The Cache can record both the request and the response MIME
5586 headers for each HTTP transaction. The headers are encoded
5587 safely and will appear as two bracketed fields at the end of
5588 the access log (for either the native or httpd-emulated log
5589 formats). To enable this logging set log_mime_hdrs to 'on'.
5594 DEFAULT: @DEFAULT_PID_FILE@
5595 LOC: Config.pidFilename
5597 A filename to write the process-id to. To disable, enter "none".
5600 NAME: client_netmask
5602 LOC: Config.Addrs.client_netmask
5604 DEFAULT_DOC: Log full client IP address
5606 A netmask for client addresses in logfiles and cachemgr output.
5607 Change this to protect the privacy of your cache clients.
5608 A netmask of 255.255.255.0 will log all IP's in that range with
5609 the last digit set to '0'.
5612 NAME: strip_query_terms
5614 LOC: Config.onoff.strip_query_terms
5617 By default, Squid strips query terms from requested URLs before
5618 logging. This protects your user's privacy and reduces log size.
5620 When investigating HIT/MISS or other caching behaviour you
5621 will need to disable this to see the full URL used by Squid.
5628 LOC: Config.onoff.buffered_logs
5630 Whether to write/send access_log records ASAP or accumulate them and
5631 then write/send them in larger chunks. Buffering may improve
5632 performance because it decreases the number of I/Os. However,
5633 buffering increases the delay before log records become available to
5634 the final recipient (e.g., a disk file or logging daemon) and,
5635 hence, increases the risk of log records loss.
5637 Note that even when buffered_logs are off, Squid may have to buffer
5638 records if it cannot write/send them immediately due to pending I/Os
5639 (e.g., the I/O writing the previous log record) or connectivity loss.
5641 Currently honored by 'daemon', 'tcp' and 'udp' access_log modules only.
5644 NAME: netdb_filename
5646 DEFAULT: stdio:@DEFAULT_NETDB_FILE@
5647 LOC: Config.netdbFilename
5650 Where Squid stores it's netdb journal.
5651 When enabled this journal preserves netdb state between restarts.
5653 To disable, enter "none".
5657 TYPE: Security::KeyLog*
5659 LOC: Config.Log.tlsKeys
5662 Configures whether and where Squid records pre-master secret and
5663 related encryption details for TLS connections accepted or established
5664 by Squid. These connections include connections accepted at
5665 https_port, TLS connections opened to origin servers/cache_peers/ICAP
5666 services, and TLS tunnels bumped by Squid using the SslBump feature.
5667 This log (a.k.a. SSLKEYLOGFILE) is meant for triage with traffic
5668 inspection tools like Wireshark.
5670 tls_key_log <destination> [options] [if [!]<acl>...]
5672 WARNING: This log allows anybody to decrypt the corresponding
5673 encrypted TLS connections, both in-flight and postmortem.
5675 At most one log file is supported at this time. Repeated tls_key_log
5676 directives are treated as fatal configuration errors. By default, no
5677 log is created or updated.
5679 If the log file does not exist, Squid creates it. Otherwise, Squid
5680 appends an existing log file.
5682 The directive is consulted whenever a TLS connection is accepted or
5683 established by Squid. TLS connections that fail the handshake may be
5684 logged if Squid got enough information to form a log record. A record
5685 is logged only if all of the configured ACLs match.
5687 While transport-related ACLs like src and dst should work, Squid may
5688 not have access to higher-level information. For example, when logging
5689 accepted https_port connections, Squid does not yet have access to the
5690 expected HTTPS request. Similarly, an HTTPS response is not available
5691 when logging most TLS connections established by Squid.
5693 The log record format is meant to be compatible with TLS deciphering
5694 features of Wireshark which relies on fields like CLIENT_RANDOM and
5695 RSA Master-Key. A single log record usually spans multiple lines.
5696 Technical documentation for that format is maintained inside the
5697 Wireshark code (e.g., see tls_keylog_process_lines() comments as of
5698 Wireshark commit e3d44136f0f0026c5e893fa249f458073f3b7328). TLS key
5699 log does not support custom record formats.
5701 This clause only supports fast acl types.
5702 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
5704 See access_log's <module>:<place> parameter for a list of supported
5705 logging destinations.
5707 TLS key log supports all access_log key=value options with the
5708 exception of logformat=name.
5710 Requires Squid built with OpenSSL support.
5715 OPTIONS FOR TROUBLESHOOTING
5716 -----------------------------------------------------------------------------
5721 DEFAULT_IF_NONE: @DEFAULT_CACHE_LOG@
5722 LOC: Debug::cache_log
5724 Squid administrative logging file.
5726 This is where general information about Squid behavior goes. You can
5727 increase the amount of data logged to this file and how often it is
5728 rotated with "debug_options"
5731 NAME: cache_log_message
5732 TYPE: cache_log_message
5734 DEFAULT_DOC: Use debug_options.
5735 LOC: DebugMessagesConfig
5737 Configures logging of individual cache.log messages.
5739 cache_log_message id=<number> option...
5740 cache_log_message ids=<number>-<number> option...
5742 Most messages have _not_ been instrumented to support this directive
5743 yet. For the list of instrumented messages and their IDs, please see
5744 the doc/debug-messages.dox file.
5746 Message ID corresponds to the message semantics rather than message
5747 text or source code location. The ID is stable across Squid
5748 instances and versions. Substantial changes in message semantics
5749 result in a new ID assignment. To reduce the danger of suppressing
5750 an important log message, the old IDs of removed (or substantially
5751 changed) messages are never reused.
5753 If more than one cache_log_message directive refers to the same
5754 message ID, the last directive wins.
5756 Use ids=min-max syntax to apply the same message configuration to an
5757 inclusive range of message IDs. An ID range with N values has
5758 exactly the same effect as typing N cache_log_message lines.
5760 At least one option is required. Supported options are:
5762 level=<number>: The logging level to use for the message. Squid
5763 command line options (-s and -d) as well as the debug_options
5764 directive control which levels go to syslog, stderr, and/or
5765 cache.log. In most environments, using level=2 or higher stops
5766 Squid from logging the message anywhere. By default, the
5767 hard-coded message-specific level is used.
5769 limit=<number>: After logging the specified number of messages at
5770 the configured (or default) debugging level DL, start using
5771 level 3 (for DL 0 and 1) or 8 (for higher DL values). Usually,
5772 level-3+ messages are not logged anywhere so this option can
5773 often be used to effectively suppress the message. Each SMP
5774 Squid process gets the same limit.
5780 DEFAULT_DOC: Log all critical and important messages.
5781 LOC: Debug::debugOptions
5783 Logging options are set as section,level where each source file
5784 is assigned a unique section. Lower levels result in less
5785 output, Full debugging (level 9) can result in a very large
5786 log file, so be careful.
5788 The magic word "ALL" sets debugging levels for all sections.
5789 The default is to run with "ALL,1" to record important warnings.
5791 The rotate=N option can be used to keep more or less of these logs
5792 than would otherwise be kept by logfile_rotate.
5793 For most uses a single log should be enough to monitor current
5794 events affecting Squid.
5799 LOC: Config.coredump_dir
5800 DEFAULT_IF_NONE: none
5801 DEFAULT_DOC: Use the directory from where Squid was started.
5803 By default Squid leaves core files in the directory from where
5804 it was started. If you set 'coredump_dir' to a directory
5805 that exists, Squid will chdir() to that directory at startup
5806 and coredump files will be left there.
5808 In addition to changing the directory, the process permissions are updated
5809 to enable process tracing and/or coredump file generation. The details are
5810 OS-specific, but look for prctl(2) PR_SET_DUMPABLE and procctl(2)
5811 PROC_TRACE_CTL documentation as guiding examples.
5815 # Leave coredumps in the first cache dir
5816 coredump_dir @DEFAULT_SWAP_DIR@
5822 OPTIONS FOR FTP GATEWAYING
5823 -----------------------------------------------------------------------------
5829 LOC: Config.Ftp.anon_user
5831 If you want the anonymous login password to be more informative
5832 (and enable the use of picky FTP servers), set this to something
5833 reasonable for your domain, like wwwuser@somewhere.net
5835 The reason why this is domainless by default is the
5836 request can be made on the behalf of a user in any domain,
5837 depending on how the cache is used.
5838 Some FTP server also validate the email address is valid
5839 (for example perl.com).
5845 LOC: Config.Ftp.passive
5847 If your firewall does not allow Squid to use passive
5848 connections, turn off this option.
5850 Use of ftp_epsv_all option requires this to be ON.
5856 LOC: Config.Ftp.epsv_all
5858 FTP Protocol extensions permit the use of a special "EPSV ALL" command.
5860 NATs may be able to put the connection on a "fast path" through the
5861 translator, as the EPRT command will never be used and therefore,
5862 translation of the data portion of the segments will never be needed.
5864 When a client only expects to do two-way FTP transfers this may be
5866 If squid finds that it must do a three-way FTP transfer after issuing
5867 an EPSV ALL command, the FTP session will fail.
5869 If you have any doubts about this option do not use it.
5870 Squid will nicely attempt all other connection methods.
5872 Requires ftp_passive to be ON (default) for any effect.
5878 LOC: Config.accessList.ftp_epsv
5880 FTP Protocol extensions permit the use of a special "EPSV" command.
5882 NATs may be able to put the connection on a "fast path" through the
5883 translator using EPSV, as the EPRT command will never be used
5884 and therefore, translation of the data portion of the segments
5885 will never be needed.
5887 EPSV is often required to interoperate with FTP servers on IPv6
5888 networks. On the other hand, it may break some IPv4 servers.
5890 By default, EPSV may try EPSV with any FTP server. To fine tune
5891 that decision, you may restrict EPSV to certain clients or servers
5894 ftp_epsv allow|deny al1 acl2 ...
5896 WARNING: Disabling EPSV may cause problems with external NAT and IPv6.
5898 Only fast ACLs are supported.
5899 Requires ftp_passive to be ON (default) for any effect.
5905 LOC: Config.Ftp.eprt
5907 FTP Protocol extensions permit the use of a special "EPRT" command.
5909 This extension provides a protocol neutral alternative to the
5910 IPv4-only PORT command. When supported it enables active FTP data
5911 channels over IPv6 and efficient NAT handling.
5913 Turning this OFF will prevent EPRT being attempted and will skip
5914 straight to using PORT for IPv4 servers.
5916 Some devices are known to not handle this extension correctly and
5917 may result in crashes. Devices which support EPRT enough to fail
5918 cleanly will result in Squid attempting PORT anyway. This directive
5919 should only be disabled when EPRT results in device failures.
5921 WARNING: Doing so will convert Squid back to the old behavior with all
5922 the related problems with external NAT devices/layers and IPv4-only FTP.
5925 NAME: ftp_sanitycheck
5928 LOC: Config.Ftp.sanitycheck
5930 For security and data integrity reasons Squid by default performs
5931 sanity checks of the addresses of FTP data connections ensure the
5932 data connection is to the requested server. If you need to allow
5933 FTP connections to servers using another IP address for the data
5934 connection turn this off.
5937 NAME: ftp_telnet_protocol
5940 LOC: Config.Ftp.telnet
5942 The FTP protocol is officially defined to use the telnet protocol
5943 as transport channel for the control connection. However, many
5944 implementations are broken and does not respect this aspect of
5947 If you have trouble accessing files with ASCII code 255 in the
5948 path or similar problems involving this ASCII code you can
5949 try setting this directive to off. If that helps, report to the
5950 operator of the FTP server in question that their FTP server
5951 is broken and does not follow the FTP standard.
5955 OPTIONS FOR EXTERNAL SUPPORT PROGRAMS
5956 -----------------------------------------------------------------------------
5961 DEFAULT: @DEFAULT_DISKD@
5962 LOC: Config.Program.diskd
5964 Specify the location of the diskd executable.
5965 Note this is only useful if you have compiled in
5966 diskd as one of the store io modules.
5969 NAME: unlinkd_program
5972 DEFAULT: @DEFAULT_UNLINKD@
5973 LOC: Config.Program.unlinkd
5975 Specify the location of the executable for file deletion process.
5978 NAME: pinger_program
5981 DEFAULT: @DEFAULT_PINGER@
5984 Specify the location of the executable for the pinger process.
5993 Control whether the pinger is active at run-time.
5994 Enables turning ICMP pinger on and off with a simple
5995 squid -k reconfigure.
6000 OPTIONS FOR URL REWRITING
6001 -----------------------------------------------------------------------------
6004 NAME: url_rewrite_program redirect_program
6006 LOC: Config.Program.redirect
6009 The name and command line parameters of an admin-provided executable
6010 for redirecting clients or adjusting/replacing client request URLs.
6012 This helper is consulted after the received request is cleared by
6013 http_access and adapted using eICAP/ICAP services (if any). If the
6014 helper does not redirect the client, Squid checks adapted_http_access
6015 and may consult the cache or forward the request to the next hop.
6018 For each request, the helper gets one line in the following format:
6020 [channel-ID <SP>] request-URL [<SP> extras] <NL>
6022 Use url_rewrite_extras to configure what Squid sends as 'extras'.
6025 The helper must reply to each query using a single line:
6027 [channel-ID <SP>] result [<SP> kv-pairs] <NL>
6029 The result section must match exactly one of the following outcomes:
6031 OK [status=30N] url="..."
6033 Redirect the client to a URL supplied in the 'url' parameter.
6034 Optional 'status' specifies the status code to send to the
6035 client in Squid's HTTP redirect response. It must be one of
6036 the standard HTTP redirect status codes: 301, 302, 303, 307,
6037 or 308. When no specific status is requested, Squid uses 302.
6039 OK rewrite-url="..."
6041 Replace the current request URL with the one supplied in the
6042 'rewrite-url' parameter. Squid fetches the resource specified
6043 by the new URL and forwards the received response (or its
6044 cached copy) to the client.
6046 WARNING: Avoid rewriting URLs! When possible, redirect the
6047 client using an "OK url=..." helper response instead.
6048 Rewriting URLs may create inconsistent requests and/or break
6049 synchronization between internal client and origin server
6050 states, especially when URLs or other message parts contain
6051 snippets of that state. For example, Squid does not adjust
6052 Location headers and embedded URLs after the helper rewrites
6055 If Squid cannot parse the URL value returned by the helper, it logs a
6056 critical cache.log ERROR message and uses the original request URL.
6057 Supported URL format depends on the request method. For example,
6058 CONNECT request URLs must use `host:port` form, while GET URLs start
6059 with a `scheme:` prefix (e.g., `https://example.com/`).
6062 Keep the client request intact.
6065 Keep the client request intact.
6068 A helper problem that should be reported to the Squid admin
6069 via a level-1 cache.log message. The 'message' parameter is
6070 reserved for specifying the log message.
6072 In addition to the kv-pairs mentioned above, Squid also understands
6073 the following optional kv-pairs in URL rewriter responses:
6076 Associates a TAG with the client TCP connection.
6078 The clt_conn_tag=TAG pair is treated as a regular transaction
6079 annotation for the current request and also annotates future
6080 requests on the same client connection. A helper may update
6081 the TAG during subsequent requests by returning a new kv-pair.
6084 Helper messages contain the channel-ID part if and only if the
6085 url_rewrite_children directive specifies positive concurrency. As a
6086 channel-ID value, Squid sends a number between 0 and concurrency-1.
6087 The helper must echo back the received channel-ID in its response.
6089 By default, Squid does not use a URL rewriter.
6092 NAME: url_rewrite_children redirect_children
6093 TYPE: HelperChildConfig
6094 DEFAULT: 20 startup=0 idle=1 concurrency=0
6095 LOC: Config.redirectChildren
6097 Specifies the maximum number of redirector processes that Squid may
6098 spawn (numberofchildren) and several related options. Using too few of
6099 these helper processes (a.k.a. "helpers") creates request queues.
6100 Using too many helpers wastes your system resources.
6102 Usage: numberofchildren [option]...
6104 The startup= and idle= options allow some measure of skew in your
6109 Sets a minimum of how many processes are to be spawned when Squid
6110 starts or reconfigures. When set to zero the first request will
6111 cause spawning of the first child process to handle it.
6113 Starting too few will cause an initial slowdown in traffic as Squid
6114 attempts to simultaneously spawn enough processes to cope.
6118 Sets a minimum of how many processes Squid is to try and keep available
6119 at all times. When traffic begins to rise above what the existing
6120 processes can handle this many more will be spawned up to the maximum
6121 configured. A minimum setting of 1 is required.
6125 The number of requests each redirector helper can handle in
6126 parallel. Defaults to 0 which indicates the redirector
6127 is a old-style single threaded redirector.
6129 When this directive is set to a value >= 1 then the protocol
6130 used to communicate with the helper is modified to include
6131 an ID in front of the request/response. The ID from the request
6132 must be echoed back with the response to that request.
6136 Sets the maximum number of queued requests. A request is queued when
6137 no existing child can accept it due to concurrency limit and no new
6138 child can be started due to numberofchildren limit. The default
6139 maximum is zero if url_rewrite_bypass is enabled and
6140 2*numberofchildren otherwise. If the queued requests exceed queue size
6141 and redirector_bypass configuration option is set, then redirector is
6142 bypassed. Otherwise, Squid is allowed to temporarily exceed the
6143 configured maximum, marking the affected helper as "overloaded". If
6144 the helper overload lasts more than 3 minutes, the action prescribed
6145 by the on-persistent-overload option applies.
6147 on-persistent-overload=action
6149 Specifies Squid reaction to a new helper request arriving when the helper
6150 has been overloaded for more that 3 minutes already. The number of queued
6151 requests determines whether the helper is overloaded (see the queue-size
6154 Two actions are supported:
6156 die Squid worker quits. This is the default behavior.
6158 ERR Squid treats the helper request as if it was
6159 immediately submitted, and the helper immediately
6160 replied with an ERR response. This action has no effect
6161 on the already queued and in-progress helper requests.
6164 NAME: url_rewrite_host_header redirect_rewrites_host_header
6167 LOC: Config.onoff.redir_rewrites_host
6169 To preserve same-origin security policies in browsers and
6170 prevent Host: header forgery by redirectors Squid rewrites
6171 any Host: header in redirected requests.
6173 If you are running an accelerator this may not be a wanted
6174 effect of a redirector. This directive enables you disable
6175 Host: alteration in reverse-proxy traffic.
6177 WARNING: Entries are cached on the result of the URL rewriting
6178 process, so be careful if you have domain-virtual hosts.
6180 WARNING: Squid and other software verifies the URL and Host
6181 are matching, so be careful not to relay through other proxies
6182 or inspecting firewalls with this disabled.
6185 NAME: url_rewrite_access redirector_access
6188 DEFAULT_DOC: Allow, unless rules exist in squid.conf.
6189 LOC: Config.accessList.redirector
6191 If defined, this access list specifies which requests are
6192 sent to the redirector processes.
6194 This clause supports both fast and slow acl types.
6195 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
6198 NAME: url_rewrite_bypass redirector_bypass
6200 LOC: Config.onoff.redirector_bypass
6203 When this is 'on', a request will not go through the
6204 redirector if all the helpers are busy. If this is 'off' and the
6205 redirector queue grows too large, the action is prescribed by the
6206 on-persistent-overload option. You should only enable this if the
6207 redirectors are not critical to your caching system. If you use
6208 redirectors for access control, and you enable this option,
6209 users may have access to pages they should not
6210 be allowed to request.
6212 Enabling this option sets the default url_rewrite_children queue-size
6216 NAME: url_rewrite_extras
6217 TYPE: TokenOrQuotedString
6218 LOC: Config.redirector_extras
6219 DEFAULT: "%>a/%>A %un %>rm myip=%la myport=%lp"
6221 Specifies a string to be append to request line format for the
6222 rewriter helper. "Quoted" format values may contain spaces and
6223 logformat %macros. In theory, any logformat %macro can be used.
6224 In practice, a %macro expands as a dash (-) if the helper request is
6225 sent before the required macro information is available to Squid.
6228 NAME: url_rewrite_timeout
6229 TYPE: UrlHelperTimeout
6230 LOC: Config.onUrlRewriteTimeout
6232 DEFAULT_DOC: Squid waits for the helper response forever
6234 Squid times active requests to redirector. The timeout value and Squid
6235 reaction to a timed out request are configurable using the following
6238 url_rewrite_timeout timeout time-units on_timeout=<action> [response=<quoted-response>]
6240 supported timeout actions:
6241 fail Squid return a ERR_GATEWAY_FAILURE error page
6243 bypass Do not re-write the URL
6245 retry Send the lookup to the helper again
6247 use_configured_response
6248 Use the <quoted-response> as helper response
6252 OPTIONS FOR STORE ID
6253 -----------------------------------------------------------------------------
6256 NAME: store_id_program storeurl_rewrite_program
6258 LOC: Config.Program.store_id
6261 Specify the location of the executable StoreID helper to use.
6262 Since they can perform almost any function there isn't one included.
6264 For each requested URL, the helper will receive one line with the format
6266 [channel-ID <SP>] URL [<SP> extras]<NL>
6269 After processing the request the helper must reply using the following format:
6271 [channel-ID <SP>] result [<SP> kv-pairs]
6273 The result code can be:
6276 Use the StoreID supplied in 'store-id='.
6279 The default is to use HTTP request URL as the store ID.
6282 An internal error occurred in the helper, preventing
6283 a result being identified.
6285 In addition to the above kv-pairs Squid also understands the following
6286 optional kv-pairs received from URL rewriters:
6288 Associates a TAG with the client TCP connection.
6289 Please see url_rewrite_program related documentation for this
6292 Helper programs should be prepared to receive and possibly ignore
6293 additional whitespace-separated tokens on each input line.
6295 When using the concurrency= option the protocol is changed by
6296 introducing a query channel tag in front of the request/response.
6297 The query channel tag is a number between 0 and concurrency-1.
6298 This value must be echoed back unchanged to Squid as the first part
6299 of the response relating to its request.
6301 NOTE: when using StoreID refresh_pattern will apply to the StoreID
6302 returned from the helper and not the URL.
6304 WARNING: Wrong StoreID value returned by a careless helper may result
6305 in the wrong cached response returned to the user.
6307 By default, a StoreID helper is not used.
6310 NAME: store_id_extras
6311 TYPE: TokenOrQuotedString
6312 LOC: Config.storeId_extras
6313 DEFAULT: "%>a/%>A %un %>rm myip=%la myport=%lp"
6315 Specifies a string to be append to request line format for the
6316 StoreId helper. "Quoted" format values may contain spaces and
6317 logformat %macros. In theory, any logformat %macro can be used.
6318 In practice, a %macro expands as a dash (-) if the helper request is
6319 sent before the required macro information is available to Squid.
6322 NAME: store_id_children storeurl_rewrite_children
6323 TYPE: HelperChildConfig
6324 DEFAULT: 20 startup=0 idle=1 concurrency=0
6325 LOC: Config.storeIdChildren
6327 Specifies the maximum number of StoreID helper processes that Squid
6328 may spawn (numberofchildren) and several related options. Using
6329 too few of these helper processes (a.k.a. "helpers") creates request
6330 queues. Using too many helpers wastes your system resources.
6332 Usage: numberofchildren [option]...
6334 The startup= and idle= options allow some measure of skew in your
6339 Sets a minimum of how many processes are to be spawned when Squid
6340 starts or reconfigures. When set to zero the first request will
6341 cause spawning of the first child process to handle it.
6343 Starting too few will cause an initial slowdown in traffic as Squid
6344 attempts to simultaneously spawn enough processes to cope.
6348 Sets a minimum of how many processes Squid is to try and keep available
6349 at all times. When traffic begins to rise above what the existing
6350 processes can handle this many more will be spawned up to the maximum
6351 configured. A minimum setting of 1 is required.
6355 The number of requests each storeID helper can handle in
6356 parallel. Defaults to 0 which indicates the helper
6357 is a old-style single threaded program.
6359 When this directive is set to a value >= 1 then the protocol
6360 used to communicate with the helper is modified to include
6361 an ID in front of the request/response. The ID from the request
6362 must be echoed back with the response to that request.
6366 Sets the maximum number of queued requests to N. A request is queued
6367 when no existing child can accept it due to concurrency limit and no
6368 new child can be started due to numberofchildren limit. The default
6369 maximum is 2*numberofchildren. If the queued requests exceed queue
6370 size and redirector_bypass configuration option is set, then
6371 redirector is bypassed. Otherwise, Squid is allowed to temporarily
6372 exceed the configured maximum, marking the affected helper as
6373 "overloaded". If the helper overload lasts more than 3 minutes, the
6374 action prescribed by the on-persistent-overload option applies.
6376 on-persistent-overload=action
6378 Specifies Squid reaction to a new helper request arriving when the helper
6379 has been overloaded for more that 3 minutes already. The number of queued
6380 requests determines whether the helper is overloaded (see the queue-size
6383 Two actions are supported:
6385 die Squid worker quits. This is the default behavior.
6387 ERR Squid treats the helper request as if it was
6388 immediately submitted, and the helper immediately
6389 replied with an ERR response. This action has no effect
6390 on the already queued and in-progress helper requests.
6393 NAME: store_id_access storeurl_rewrite_access
6396 DEFAULT_DOC: Allow, unless rules exist in squid.conf.
6397 LOC: Config.accessList.store_id
6399 If defined, this access list specifies which requests are
6400 sent to the StoreID processes. By default all requests
6403 This clause supports both fast and slow acl types.
6404 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
6407 NAME: store_id_bypass storeurl_rewrite_bypass
6409 LOC: Config.onoff.store_id_bypass
6412 When this is 'on', a request will not go through the
6413 helper if all helpers are busy. If this is 'off' and the helper
6414 queue grows too large, the action is prescribed by the
6415 on-persistent-overload option. You should only enable this if the
6416 helpers are not critical to your caching system. If you use
6417 helpers for critical caching components, and you enable this
6418 option, users may not get objects from cache.
6419 This options sets default queue-size option of the store_id_children
6424 OPTIONS FOR TUNING THE CACHE
6425 -----------------------------------------------------------------------------
6428 NAME: cache no_cache
6431 DEFAULT_DOC: By default, this directive is unused and has no effect.
6432 LOC: Config.accessList.noCache
6434 Requests denied by this directive will not be served from the cache
6435 and their responses will not be stored in the cache. This directive
6436 has no effect on other transactions and on already cached responses.
6438 This clause supports both fast and slow acl types.
6439 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
6441 This and the two other similar caching directives listed below are
6442 checked at different transaction processing stages, have different
6443 access to response information, affect different cache operations,
6444 and differ in slow ACLs support:
6446 * cache: Checked before Squid makes a hit/miss determination.
6447 No access to reply information!
6448 Denies both serving a hit and storing a miss.
6449 Supports both fast and slow ACLs.
6450 * send_hit: Checked after a hit was detected.
6451 Has access to reply (hit) information.
6452 Denies serving a hit only.
6453 Supports fast ACLs only.
6454 * store_miss: Checked before storing a cachable miss.
6455 Has access to reply (miss) information.
6456 Denies storing a miss only.
6457 Supports fast ACLs only.
6459 If you are not sure which of the three directives to use, apply the
6460 following decision logic:
6462 * If your ACL(s) are of slow type _and_ need response info, redesign.
6463 Squid does not support that particular combination at this time.
6465 * If your directive ACL(s) are of slow type, use "cache"; and/or
6466 * if your directive ACL(s) need no response info, use "cache".
6468 * If you do not want the response cached, use store_miss; and/or
6469 * if you do not want a hit on a cached response, use send_hit.
6475 DEFAULT_DOC: By default, this directive is unused and has no effect.
6476 LOC: Config.accessList.sendHit
6478 Responses denied by this directive will not be served from the cache
6479 (but may still be cached, see store_miss). This directive has no
6480 effect on the responses it allows and on the cached objects. This
6481 directive is applied to both regular from-cache responses and responses
6482 reused by collapsed requests (see collapsed_forwarding).
6484 Please see the "cache" directive for a summary of differences among
6485 store_miss, send_hit, and cache directives.
6487 Unlike the "cache" directive, send_hit only supports fast acl
6488 types. See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
6492 # apply custom Store ID mapping to some URLs
6493 acl MapMe dstdomain .c.example.com
6494 store_id_program ...
6495 store_id_access allow MapMe
6497 # but prevent caching of special responses
6498 # such as 302 redirects that cause StoreID loops
6499 acl Ordinary http_status 200-299
6500 store_miss deny MapMe !Ordinary
6502 # and do not serve any previously stored special responses
6503 # from the cache (in case they were already cached before
6504 # the above store_miss rule was in effect).
6505 send_hit deny MapMe !Ordinary
6511 DEFAULT_DOC: By default, this directive is unused and has no effect.
6512 LOC: Config.accessList.storeMiss
6514 Responses denied by this directive will not be cached (but may still
6515 be served from the cache, see send_hit). This directive has no
6516 effect on the responses it allows and on the already cached responses.
6518 Please see the "cache" directive for a summary of differences among
6519 store_miss, send_hit, and cache directives. See the
6520 send_hit directive for a usage example.
6522 Unlike the "cache" directive, store_miss only supports fast acl
6523 types. See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
6529 LOC: Config.maxStale
6532 This option puts an upper limit on how stale content Squid
6533 will serve from the cache if cache validation fails.
6534 Can be overridden by the refresh_pattern max-stale option.
6537 NAME: refresh_pattern
6538 TYPE: refreshpattern
6542 usage: refresh_pattern [-i] regex min percent max [options]
6544 By default, regular expressions are CASE-SENSITIVE. To make
6545 them case-insensitive, use the -i option.
6547 'Min' is the time (in minutes) an object without an explicit
6548 expiry time should be considered fresh. The recommended
6549 value is 0, any higher values may cause dynamic applications
6550 to be erroneously cached unless the application designer
6551 has taken the appropriate actions.
6553 'Percent' is used to compute the max-age value for responses
6554 with a Last-Modified header and no Cache-Control:max-age nor Expires.
6555 Cache-Control:max-age = ( Date - Last-Modified ) * percent
6557 'Max' is an upper limit on how long objects without an explicit
6558 expiry time will be considered fresh. The value is also used
6559 to form Cache-Control: max-age header for a request sent from
6560 Squid to origin/parent.
6562 options: override-expire
6572 override-expire enforces min age even if the server
6573 sent an explicit expiry time (e.g., with the
6574 Expires: header or Cache-Control: max-age). Doing this
6575 VIOLATES the HTTP standard. Enabling this feature
6576 could make you liable for problems which it causes.
6578 Note: override-expire does not enforce staleness - it only extends
6579 freshness / min. If the server returns a Expires time which
6580 is longer than your max time, Squid will still consider
6581 the object fresh for that period of time.
6583 override-lastmod enforces min age even on objects
6584 that were modified recently.
6586 reload-into-ims changes a client no-cache or ``reload''
6587 request for a cached entry into a conditional request using
6588 If-Modified-Since and/or If-None-Match headers, provided the
6589 cached entry has a Last-Modified and/or a strong ETag header.
6590 Doing this VIOLATES the HTTP standard. Enabling this feature
6591 could make you liable for problems which it causes.
6593 ignore-reload ignores a client no-cache or ``reload''
6594 header. Doing this VIOLATES the HTTP standard. Enabling
6595 this feature could make you liable for problems which
6598 ignore-no-store ignores any ``Cache-control: no-store''
6599 headers received from a server. Doing this VIOLATES
6600 the HTTP standard. Enabling this feature could make you
6601 liable for problems which it causes.
6603 ignore-private ignores any ``Cache-control: private''
6604 headers received from a server. Doing this VIOLATES
6605 the HTTP standard. Enabling this feature could make you
6606 liable for problems which it causes.
6608 refresh-ims causes squid to contact the origin server
6609 when a client issues an If-Modified-Since request. This
6610 ensures that the client will receive an updated version
6611 if one is available.
6613 store-stale stores responses even if they don't have explicit
6614 freshness or a validator (i.e., Last-Modified or an ETag)
6615 present, or if they're already stale. By default, Squid will
6616 not cache such responses because they usually can't be
6617 reused. Note that such responses will be stale by default.
6619 max-stale=NN provide a maximum staleness factor. Squid won't
6620 serve objects more stale than this even if it failed to
6621 validate the object. Default: use the max_stale global limit.
6623 Basically a cached object is:
6625 FRESH if expire > now, else STALE
6627 FRESH if lm-factor < percent, else STALE
6631 The refresh_pattern lines are checked in the order listed here.
6632 The first entry which matches is used. If none of the entries
6633 match the default will be used.
6635 Note, you must uncomment all the default lines if you want
6636 to change one. The default setting is only active if none is
6642 # Add any of your own refresh_pattern entries above these.
6644 refresh_pattern ^ftp: 1440 20% 10080
6645 refresh_pattern -i (/cgi-bin/|\?) 0 0% 0
6646 refresh_pattern . 0 20% 4320
6650 NAME: quick_abort_min
6654 LOC: Config.quickAbort.min
6657 NAME: quick_abort_max
6661 LOC: Config.quickAbort.max
6664 NAME: quick_abort_pct
6668 LOC: Config.quickAbort.pct
6670 Continuing to download a cachable response after its request is aborted is
6671 going to waste resources if the received response is not requested again.
6672 On the other hand, aborting an in-progress download may effectively waste
6673 (already spent) resources if the received cachable response is requested
6674 again. Such waste is especially noticeable when, for example, impatient
6675 users repeatedly request and then abort slow downloads. To balance these
6676 trade-offs when a request is aborted during response download, Squid may
6677 check quick_abort_* directives to decide whether to finish the retrieval:
6679 If the transfer has less than 'quick_abort_min' KB remaining,
6680 it will finish the retrieval.
6682 If the transfer has more than 'quick_abort_max' KB remaining,
6683 it will abort the retrieval.
6685 If more than 'quick_abort_pct' of the transfer has completed,
6686 it will finish the retrieval.
6688 If you do not want any retrieval to continue after the client
6689 has aborted, set both 'quick_abort_min' and 'quick_abort_max'
6692 If you want retrievals to always continue if they are being
6693 cached set 'quick_abort_min' to '-1 KB'.
6695 Many other conditions affect Squid decision to abort or continue download.
6696 For example, Squid continues to download responses that feed other
6697 requests but aborts responses with unknown body length.
6700 NAME: read_ahead_gap
6701 COMMENT: buffer-size
6703 LOC: Config.readAheadGap
6706 The amount of data the cache will buffer ahead of what has been
6707 sent to the client when retrieving an object from another server.
6711 IFDEF: USE_HTTP_VIOLATIONS
6714 LOC: Config.negativeTtl
6717 Set the Default Time-to-Live (TTL) for failed requests.
6718 Certain types of failures (such as "connection refused" and
6719 "404 Not Found") are able to be negatively-cached for a short time.
6720 Modern web servers should provide Expires: header, however if they
6721 do not this can provide a minimum TTL.
6722 The default is not to cache errors with unknown expiry details.
6724 Note that this is different from negative caching of DNS lookups.
6726 WARNING: Doing this VIOLATES the HTTP standard. Enabling
6727 this feature could make you liable for problems which it
6731 NAME: positive_dns_ttl
6734 LOC: Config.positiveDnsTtl
6737 Upper limit on how long Squid will cache positive DNS responses.
6738 Default is 6 hours (360 minutes). This directive must be set
6739 larger than negative_dns_ttl.
6742 NAME: negative_dns_ttl
6745 LOC: Config.negativeDnsTtl
6748 Time-to-Live (TTL) for negative caching of failed DNS lookups.
6749 This also sets the lower cache limit on positive lookups.
6750 Minimum value is 1 second, and it is not recommendable to go
6751 much below 10 seconds.
6754 NAME: range_offset_limit
6755 COMMENT: size [acl acl...]
6757 LOC: Config.rangeOffsetLimit
6760 usage: (size) [units] [[!]aclname]
6762 Sets an upper limit on how far (number of bytes) into the file
6763 a Range request may be to cause Squid to prefetch the whole file.
6764 If beyond this limit, Squid forwards the Range request as it is and
6765 the result is NOT cached.
6767 This is to stop a far ahead range request (lets say start at 17MB)
6768 from making Squid fetch the whole object up to that point before
6769 sending anything to the client.
6771 Multiple range_offset_limit lines may be specified, and they will
6772 be searched from top to bottom on each request until a match is found.
6773 The first match found will be used. If no line matches a request, the
6774 default limit of 0 bytes will be used.
6776 'size' is the limit specified as a number of units.
6778 'units' specifies whether to use bytes, KB, MB, etc.
6779 If no units are specified bytes are assumed.
6781 A size of 0 causes Squid to never fetch more than the
6782 client requested. (default)
6784 A size of 'none' causes Squid to always fetch the object from the
6785 beginning so it may cache the result. (2.0 style)
6787 'aclname' is the name of a defined ACL.
6789 NP: Using 'none' as the byte value here will override any quick_abort settings
6790 that may otherwise apply to the range request. The range request will
6791 be fully fetched from start to finish regardless of the client
6792 actions. This affects bandwidth usage.
6795 NAME: minimum_expiry_time
6798 LOC: Config.minimum_expiry_time
6801 The minimum caching time according to (Expires - Date)
6802 headers Squid honors if the object can't be revalidated.
6803 The default is 60 seconds.
6805 In reverse proxy environments it might be desirable to honor
6806 shorter object lifetimes. It is most likely better to make
6807 your server return a meaningful Last-Modified header however.
6810 NAME: store_avg_object_size
6814 LOC: Config.Store.avgObjectSize
6816 Average object size, used to estimate number of objects your
6817 cache can hold. The default is 13 KB.
6819 This is used to pre-seed the cache index memory allocation to
6820 reduce expensive reallocate operations while handling clients
6821 traffic. Too-large values may result in memory allocation during
6822 peak traffic, too-small values will result in wasted memory.
6824 Check the cache manager 'info' report metrics for the real
6825 object sizes seen by your Squid before tuning this.
6828 NAME: store_objects_per_bucket
6831 LOC: Config.Store.objectsPerBucket
6833 Target number of objects per bucket in the store hash table.
6834 Lowering this value increases the total number of buckets and
6835 also the storage maintenance rate. The default is 20.
6840 -----------------------------------------------------------------------------
6843 NAME: request_header_max_size
6847 LOC: Config.maxRequestHeaderSize
6849 This directives limits the header size of a received HTTP request
6850 (including request-line). Increasing this limit beyond its 64 KB default
6851 exposes certain old Squid code to various denial-of-service attacks. This
6852 limit also applies to received FTP commands.
6854 This limit has no direct affect on Squid memory consumption.
6856 Squid does not check this limit when sending requests.
6859 NAME: reply_header_max_size
6863 LOC: Config.maxReplyHeaderSize
6865 This directives limits the header size of a received HTTP response
6866 (including status-line). Increasing this limit beyond its 64 KB default
6867 exposes certain old Squid code to various denial-of-service attacks. This
6868 limit also applies to FTP command responses.
6870 Squid also checks this limit when loading hit responses from disk cache.
6872 Squid does not check this limit when sending responses.
6875 NAME: request_body_max_size
6879 DEFAULT_DOC: No limit.
6880 LOC: Config.maxRequestBodySize
6882 This specifies the maximum size for an HTTP request body.
6883 In other words, the maximum size of a PUT/POST request.
6884 A user who attempts to send a request with a body larger
6885 than this limit receives an "Invalid Request" error message.
6886 If you set this parameter to a zero (the default), there will
6887 be no limit imposed.
6889 See also client_request_buffer_max_size for an alternative
6890 limitation on client uploads which can be configured.
6893 NAME: client_request_buffer_max_size
6897 LOC: Config.maxRequestBufferSize
6899 This specifies the maximum buffer size of a client request.
6900 It prevents squid eating too much memory when somebody uploads
6905 IFDEF: USE_HTTP_VIOLATIONS
6908 DEFAULT_DOC: Obey RFC 2616.
6909 LOC: Config.accessList.brokenPosts
6911 A list of ACL elements which, if matched, causes Squid to send
6912 an extra CRLF pair after the body of a PUT/POST request.
6914 Some HTTP servers has broken implementations of PUT/POST,
6915 and rely on an extra CRLF pair sent by some WWW clients.
6917 Quote from RFC2616 section 4.1 on this matter:
6919 Note: certain buggy HTTP/1.0 client implementations generate an
6920 extra CRLF's after a POST request. To restate what is explicitly
6921 forbidden by the BNF, an HTTP/1.1 client must not preface or follow
6922 a request with an extra CRLF.
6924 This clause only supports fast acl types.
6925 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
6928 acl buggy_server url_regex ^http://....
6929 broken_posts allow buggy_server
6932 NAME: adaptation_uses_indirect_client icap_uses_indirect_client
6935 IFDEF: FOLLOW_X_FORWARDED_FOR&&USE_ADAPTATION
6937 LOC: Adaptation::Config::use_indirect_client
6939 Controls whether the indirect client IP address (instead of the direct
6940 client IP address) is passed to adaptation services.
6942 See also: follow_x_forwarded_for adaptation_send_client_ip
6946 IFDEF: USE_HTTP_VIOLATIONS
6950 LOC: Config.onoff.via
6952 If set (default), Squid will include a Via header in requests and
6953 replies as required by RFC2616.
6956 NAME: vary_ignore_expire
6959 LOC: Config.onoff.vary_ignore_expire
6962 Many HTTP servers supporting Vary gives such objects
6963 immediate expiry time with no cache-control header
6964 when requested by a HTTP/1.0 client. This option
6965 enables Squid to ignore such expiry times until
6966 HTTP/1.1 is fully implemented.
6968 WARNING: If turned on this may eventually cause some
6969 varying objects not intended for caching to get cached.
6972 NAME: request_header_access
6973 IFDEF: USE_HTTP_VIOLATIONS
6974 TYPE: http_header_access
6975 LOC: Config.request_header_access
6977 DEFAULT_DOC: No limits.
6979 Usage: request_header_access header_name allow|deny [!]aclname ...
6981 WARNING: Doing this VIOLATES the HTTP standard. Enabling
6982 this feature could make you liable for problems which it
6985 This option replaces the old 'anonymize_headers' and the
6986 older 'http_anonymizer' option with something that is much
6987 more configurable. A list of ACLs for each header name allows
6988 removal of specific header fields under specific conditions.
6990 This option only applies to outgoing HTTP request headers (i.e.,
6991 headers sent by Squid to the next HTTP hop such as a cache peer
6992 or an origin server). The option has no effect during cache hit
6993 detection. The equivalent adaptation vectoring point in ICAP
6994 terminology is post-cache REQMOD.
6996 The option is applied to individual outgoing request header
6997 fields. For each request header field F, Squid uses the first
6998 qualifying sets of request_header_access rules:
7000 1. Rules with header_name equal to F's name.
7001 2. Rules with header_name 'Other', provided F's name is not
7002 on the hard-coded list of commonly used HTTP header names.
7003 3. Rules with header_name 'All'.
7005 Within that qualifying rule set, rule ACLs are checked as usual.
7006 If ACLs of an "allow" rule match, the header field is allowed to
7007 go through as is. If ACLs of a "deny" rule match, the header is
7008 removed and request_header_replace is then checked to identify
7009 if the removed header has a replacement. If no rules within the
7010 set have matching ACLs, the header field is left as is.
7012 For example, to achieve the same behavior as the old
7013 'http_anonymizer standard' option, you should use:
7015 request_header_access From deny all
7016 request_header_access Referer deny all
7017 request_header_access User-Agent deny all
7019 Or, to reproduce the old 'http_anonymizer paranoid' feature
7022 request_header_access Authorization allow all
7023 request_header_access Proxy-Authorization allow all
7024 request_header_access Cache-Control allow all
7025 request_header_access Content-Length allow all
7026 request_header_access Content-Type allow all
7027 request_header_access Date allow all
7028 request_header_access Host allow all
7029 request_header_access If-Modified-Since allow all
7030 request_header_access Pragma allow all
7031 request_header_access Accept allow all
7032 request_header_access Accept-Charset allow all
7033 request_header_access Accept-Encoding allow all
7034 request_header_access Accept-Language allow all
7035 request_header_access Connection allow all
7036 request_header_access All deny all
7038 HTTP reply headers are controlled with the reply_header_access directive.
7040 By default, all headers are allowed (no anonymizing is performed).
7043 NAME: reply_header_access
7044 IFDEF: USE_HTTP_VIOLATIONS
7045 TYPE: http_header_access
7046 LOC: Config.reply_header_access
7048 DEFAULT_DOC: No limits.
7050 Usage: reply_header_access header_name allow|deny [!]aclname ...
7052 WARNING: Doing this VIOLATES the HTTP standard. Enabling
7053 this feature could make you liable for problems which it
7056 This option only applies to reply headers, i.e., from the
7057 server to the client.
7059 This is the same as request_header_access, but in the other
7060 direction. Please see request_header_access for detailed
7063 For example, to achieve the same behavior as the old
7064 'http_anonymizer standard' option, you should use:
7066 reply_header_access Server deny all
7067 reply_header_access WWW-Authenticate deny all
7068 reply_header_access Link deny all
7070 Or, to reproduce the old 'http_anonymizer paranoid' feature
7073 reply_header_access Allow allow all
7074 reply_header_access WWW-Authenticate allow all
7075 reply_header_access Proxy-Authenticate allow all
7076 reply_header_access Cache-Control allow all
7077 reply_header_access Content-Encoding allow all
7078 reply_header_access Content-Length allow all
7079 reply_header_access Content-Type allow all
7080 reply_header_access Date allow all
7081 reply_header_access Expires allow all
7082 reply_header_access Last-Modified allow all
7083 reply_header_access Location allow all
7084 reply_header_access Pragma allow all
7085 reply_header_access Content-Language allow all
7086 reply_header_access Retry-After allow all
7087 reply_header_access Title allow all
7088 reply_header_access Content-Disposition allow all
7089 reply_header_access Connection allow all
7090 reply_header_access All deny all
7092 HTTP request headers are controlled with the request_header_access directive.
7094 By default, all headers are allowed (no anonymizing is
7098 NAME: request_header_replace header_replace
7099 IFDEF: USE_HTTP_VIOLATIONS
7100 TYPE: http_header_replace
7101 LOC: Config.request_header_access
7104 Usage: request_header_replace header_name message
7105 Example: request_header_replace User-Agent Nutscrape/1.0 (CP/M; 8-bit)
7107 This option allows you to change the contents of headers
7108 denied with request_header_access above, by replacing them
7109 with some fixed string.
7111 This only applies to request headers, not reply headers.
7113 By default, headers are removed if denied.
7116 NAME: reply_header_replace
7117 IFDEF: USE_HTTP_VIOLATIONS
7118 TYPE: http_header_replace
7119 LOC: Config.reply_header_access
7122 Usage: reply_header_replace header_name message
7123 Example: reply_header_replace Server Foo/1.0
7125 This option allows you to change the contents of headers
7126 denied with reply_header_access above, by replacing them
7127 with some fixed string.
7129 This only applies to reply headers, not request headers.
7131 By default, headers are removed if denied.
7134 NAME: request_header_add
7135 TYPE: HeaderWithAclList
7136 LOC: Config.request_header_add
7139 Usage: request_header_add field-name field-value [ acl ... ]
7140 Example: request_header_add X-Client-CA "CA=%ssl::>cert_issuer" all
7142 This option adds header fields to outgoing HTTP requests (i.e.,
7143 request headers sent by Squid to the next HTTP hop such as a
7144 cache peer or an origin server). The option has no effect during
7145 cache hit detection. The equivalent adaptation vectoring point
7146 in ICAP terminology is post-cache REQMOD.
7148 Field-name is a token specifying an HTTP header name. If a
7149 standard HTTP header name is used, Squid does not check whether
7150 the new header conflicts with any existing headers or violates
7151 HTTP rules. If the request to be modified already contains a
7152 field with the same name, the old field is preserved but the
7153 header field values are not merged.
7155 Field-value is either a token or a quoted string. If quoted
7156 string format is used, then the surrounding quotes are removed
7157 while escape sequences and %macros are processed.
7159 One or more Squid ACLs may be specified to restrict header
7160 injection to matching requests. As always in squid.conf, all
7161 ACLs in the ACL list must be satisfied for the insertion to
7162 happen. The request_header_add supports fast ACLs only.
7164 See also: reply_header_add.
7167 NAME: reply_header_add
7168 TYPE: HeaderWithAclList
7169 LOC: Config.reply_header_add
7172 Usage: reply_header_add field-name field-value [ acl ... ]
7173 Example: reply_header_add X-Client-CA "CA=%ssl::>cert_issuer" all
7175 This option adds header fields to outgoing HTTP responses (i.e., response
7176 headers delivered by Squid to the client). This option has no effect on
7177 cache hit detection. The equivalent adaptation vectoring point in
7178 ICAP terminology is post-cache RESPMOD. This option does not apply to
7179 successful CONNECT replies.
7181 Field-name is a token specifying an HTTP header name. If a
7182 standard HTTP header name is used, Squid does not check whether
7183 the new header conflicts with any existing headers or violates
7184 HTTP rules. If the response to be modified already contains a
7185 field with the same name, the old field is preserved but the
7186 header field values are not merged.
7188 Field-value is either a token or a quoted string. If quoted
7189 string format is used, then the surrounding quotes are removed
7190 while escape sequences and %macros are processed.
7192 One or more Squid ACLs may be specified to restrict header
7193 injection to matching responses. As always in squid.conf, all
7194 ACLs in the ACL list must be satisfied for the insertion to
7195 happen. The reply_header_add option supports fast ACLs only.
7197 See also: request_header_add.
7205 This option used to log custom information about the master
7206 transaction. For example, an admin may configure Squid to log
7207 which "user group" the transaction belongs to, where "user group"
7208 will be determined based on a set of ACLs and not [just]
7209 authentication information.
7210 Values of key/value pairs can be logged using %{key}note macros:
7212 note key value acl ...
7213 logformat myFormat ... %{key}note ...
7215 This clause only supports fast acl types.
7216 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
7219 NAME: relaxed_header_parser
7220 COMMENT: on|off|warn
7222 LOC: Config.onoff.relaxed_header_parser
7225 In the default "on" setting Squid accepts certain forms
7226 of non-compliant HTTP messages where it is unambiguous
7227 what the sending application intended even if the message
7228 is not correctly formatted. The messages is then normalized
7229 to the correct form when forwarded by Squid.
7231 If set to "warn" then a warning will be emitted in cache.log
7232 each time such HTTP error is encountered.
7234 If set to "off" then such HTTP errors will cause the request
7235 or response to be rejected.
7238 NAME: collapsed_forwarding
7241 LOC: Config.onoff.collapsed_forwarding
7244 This option controls whether Squid is allowed to merge multiple
7245 potentially cachable requests for the same URI before Squid knows
7246 whether the response is going to be cachable.
7248 When enabled, instead of forwarding each concurrent request for
7249 the same URL, Squid just sends the first of them. The other, so
7250 called "collapsed" requests, wait for the response to the first
7251 request and, if it happens to be cachable, use that response.
7252 Here, "concurrent requests" means "received after the first
7253 request headers were parsed and before the corresponding response
7254 headers were parsed".
7256 This feature is disabled by default: enabling collapsed
7257 forwarding needlessly delays forwarding requests that look
7258 cachable (when they are collapsed) but then need to be forwarded
7259 individually anyway because they end up being for uncachable
7260 content. However, in some cases, such as acceleration of highly
7261 cachable content with periodic or grouped expiration times, the
7262 gains from collapsing [large volumes of simultaneous refresh
7263 requests] outweigh losses from such delays.
7265 Squid collapses two kinds of requests: regular client requests
7266 received on one of the listening ports and internal "cache
7267 revalidation" requests which are triggered by those regular
7268 requests hitting a stale cached object. Revalidation collapsing
7269 is currently disabled for Squid instances containing SMP-aware
7270 disk or memory caches and for Vary-controlled cached objects.
7272 A response reused by the collapsed request is deemed fresh in that
7273 request processing context -- Squid does not apply refresh_pattern and
7274 internal freshness validation checks to collapsed transactions. Squid
7275 does apply send_hit rules.
7278 NAME: collapsed_forwarding_access
7281 DEFAULT_DOC: Requests may be collapsed if collapsed_forwarding is on.
7282 LOC: Config.accessList.collapsedForwardingAccess
7284 Use this directive to restrict collapsed forwarding to a subset of
7285 eligible requests. The directive is checked for regular HTTP
7286 requests, internal revalidation requests, and HTCP/ICP requests.
7288 collapsed_forwarding_access allow|deny [!]aclname ...
7290 This directive cannot force collapsing. It has no effect on
7291 collapsing unless collapsed_forwarding is 'on', and all other
7292 collapsing preconditions are satisfied.
7294 * A denied request will not collapse, and future transactions will
7295 not collapse on it (even if they are allowed to collapse).
7297 * An allowed request may collapse, or future transactions may
7298 collapse on it (provided they are allowed to collapse).
7300 This directive is evaluated before receiving HTTP response headers
7301 and without access to Squid-to-peer connection (if any).
7303 Only fast ACLs are supported.
7305 See also: collapsed_forwarding.
7308 NAME: shared_transient_entries_limit collapsed_forwarding_shared_entries_limit
7309 COMMENT: (number of entries)
7311 LOC: Config.shared_transient_entries_limit
7314 This directive limits the size of a table used for sharing current
7315 transaction information among SMP workers. A table entry stores meta
7316 information about a single cache entry being delivered to Squid
7317 client(s) by one or more SMP workers. A single table entry consumes
7318 less than 128 shared memory bytes.
7320 The limit should be significantly larger than the number of
7321 concurrent non-collapsed cachable responses leaving Squid. For a
7322 cache that handles less than 5000 concurrent requests, the default
7323 setting of 16384 should be plenty.
7325 Using excessively large values wastes shared memory. Limiting the
7326 table size too much results in hash collisions, leading to lower hit
7327 ratio and missed SMP request collapsing opportunities: Transactions
7328 left without a table entry cannot cache their responses and are
7329 invisible to other concurrent requests for the same resource.
7331 A zero limit is allowed but unsupported. A positive small limit
7332 lowers hit ratio, but zero limit disables a lot of essential
7333 synchronization among SMP workers, leading to HTTP violations (e.g.,
7334 stale hit responses). It also disables shared collapsed forwarding:
7335 A worker becomes unable to collapse its requests on transactions in
7336 other workers, resulting in more trips to the origin server and more
7342 -----------------------------------------------------------------------------
7345 NAME: forward_timeout
7348 LOC: Config.Timeout.forward
7351 This parameter specifies how long Squid should at most attempt in
7352 finding a forwarding path for the request before giving up.
7355 NAME: connect_timeout
7358 LOC: Config.Timeout.connect
7361 This parameter specifies how long to wait for the TCP connect to
7362 the requested server or peer to complete before Squid should
7363 attempt to find another path where to forward the request.
7366 NAME: peer_connect_timeout
7369 LOC: Config.Timeout.peer_connect
7372 This parameter specifies how long to wait for a pending TCP
7373 connection to a peer cache. The default is 30 seconds. You
7374 may also set different timeout values for individual neighbors
7375 with the 'connect-timeout' option on a 'cache_peer' line.
7381 LOC: Config.Timeout.read
7384 Applied on peer server connections.
7386 After each successful read(), the timeout will be extended by this
7387 amount. If no data is read again after this amount of time,
7388 the request is aborted and logged with ERR_READ_TIMEOUT.
7390 The default is 15 minutes.
7396 LOC: Config.Timeout.write
7399 This timeout is tracked for all connections that have data
7400 available for writing and are waiting for the socket to become
7401 ready. After each successful write, the timeout is extended by
7402 the configured amount. If Squid has data to write but the
7403 connection is not ready for the configured duration, the
7404 transaction associated with the connection is terminated. The
7405 default is 15 minutes.
7408 NAME: request_timeout
7410 LOC: Config.Timeout.request
7413 How long to wait for complete HTTP request headers after initial
7414 connection establishment.
7417 NAME: request_start_timeout
7419 LOC: Config.Timeout.request_start_timeout
7422 How long to wait for the first request byte after initial
7423 connection establishment.
7426 NAME: client_idle_pconn_timeout persistent_request_timeout
7428 LOC: Config.Timeout.clientIdlePconn
7431 How long to wait for the next HTTP request on a persistent
7432 client connection after the previous request completes.
7435 NAME: ftp_client_idle_timeout
7437 LOC: Config.Timeout.ftpClientIdle
7440 How long to wait for an FTP request on a connection to Squid ftp_port.
7441 Many FTP clients do not deal with idle connection closures well,
7442 necessitating a longer default timeout than client_idle_pconn_timeout
7443 used for incoming HTTP requests.
7446 NAME: client_lifetime
7449 LOC: Config.Timeout.lifetime
7452 The maximum amount of time a client (browser) is allowed to
7453 remain connected to the cache process. This protects the Cache
7454 from having a lot of sockets (and hence file descriptors) tied up
7455 in a CLOSE_WAIT state from remote clients that go away without
7456 properly shutting down (either because of a network failure or
7457 because of a poor client implementation). The default is one
7460 NOTE: The default value is intended to be much larger than any
7461 client would ever need to be connected to your cache. You
7462 should probably change client_lifetime only as a last resort.
7463 If you seem to have many client connections tying up
7464 filedescriptors, we recommend first tuning the read_timeout,
7465 request_timeout, persistent_request_timeout and quick_abort values.
7468 NAME: pconn_lifetime
7471 LOC: Config.Timeout.pconnLifetime
7474 Desired maximum lifetime of a persistent connection.
7475 When set, Squid will close a now-idle persistent connection that
7476 exceeded configured lifetime instead of moving the connection into
7477 the idle connection pool (or equivalent). No effect on ongoing/active
7478 transactions. Connection lifetime is the time period from the
7479 connection acceptance or opening time until "now".
7481 This limit is useful in environments with long-lived connections
7482 where Squid configuration or environmental factors change during a
7483 single connection lifetime. If unrestricted, some connections may
7484 last for hours and even days, ignoring those changes that should
7485 have affected their behavior or their existence.
7487 Currently, a new lifetime value supplied via Squid reconfiguration
7488 has no effect on already idle connections unless they become busy.
7490 When set to '0' this limit is not used.
7493 NAME: half_closed_clients
7495 LOC: Config.onoff.half_closed_clients
7498 Some clients may shutdown the sending side of their TCP
7499 connections, while leaving their receiving sides open. Sometimes,
7500 Squid can not tell the difference between a half-closed and a
7501 fully-closed TCP connection.
7503 By default, Squid will immediately close client connections when
7504 read(2) returns "no more data to read."
7506 Change this option to 'on' and Squid will keep open connections
7507 until a read(2) or write(2) on the socket returns an error.
7508 This may show some benefits for reverse proxies. But if not
7509 it is recommended to leave OFF.
7512 NAME: server_idle_pconn_timeout pconn_timeout
7514 LOC: Config.Timeout.serverIdlePconn
7517 Timeout for idle persistent connections to servers and other
7521 NAME: shutdown_lifetime
7524 LOC: Config.shutdownLifetime
7527 When SIGTERM or SIGHUP is received, the cache is put into
7528 "shutdown pending" mode until all active sockets are closed.
7529 This value is the lifetime to set for all open descriptors
7530 during shutdown mode. Any active clients after this many
7531 seconds will receive a 'timeout' message.
7535 ADMINISTRATIVE PARAMETERS
7536 -----------------------------------------------------------------------------
7542 LOC: Config.adminEmail
7544 Email-address of local cache manager who will receive
7545 mail if the cache dies. The default is "webmaster".
7551 LOC: Config.EmailFrom
7553 From: email-address for mail sent when the cache dies.
7554 The default is to use 'squid@unique_hostname'.
7556 See also: unique_hostname directive.
7562 LOC: Config.EmailProgram
7564 Email program used to send mail if the cache dies.
7565 The default is "mail". The specified program must comply
7566 with the standard Unix mail syntax:
7567 mail-program recipient < mailfile
7569 Optional command line options can be specified.
7572 NAME: cache_effective_user
7574 DEFAULT: @DEFAULT_CACHE_EFFECTIVE_USER@
7575 LOC: Config.effectiveUser
7577 If you start Squid as root, it will change its effective/real
7578 UID/GID to the user specified below. The default is to change
7579 to UID of @DEFAULT_CACHE_EFFECTIVE_USER@.
7580 see also; cache_effective_group
7583 NAME: cache_effective_group
7586 DEFAULT_DOC: Use system group memberships of the cache_effective_user account
7587 LOC: Config.effectiveGroup
7589 Squid sets the GID to the effective user's default group ID
7590 (taken from the password file) and supplementary group list
7591 from the groups membership.
7593 If you want Squid to run with a specific GID regardless of
7594 the group memberships of the effective user then set this
7595 to the group (or GID) you want Squid to run as. When set
7596 all other group privileges of the effective user are ignored
7597 and only this GID is effective. If Squid is not started as
7598 root the user starting Squid MUST be member of the specified
7601 This option is not recommended by the Squid Team.
7602 Our preference is for administrators to configure a secure
7603 user account for squid with UID/GID matching system policies.
7606 NAME: httpd_suppress_version_string
7610 LOC: Config.onoff.httpd_suppress_version_string
7612 Do not send Squid version string in HTTP metadata and generated content
7613 such as HTML error pages. Squid version string is still present in certain
7614 SNMP responses, HTTP(S) Server response header field,
7615 various console output, and cache.log.
7618 NAME: visible_hostname
7620 LOC: Config.visibleHostname
7622 DEFAULT_DOC: Automatically detect the system host name
7624 If you want to present a special hostname in error messages, etc,
7625 define this. Otherwise, the return value of gethostname()
7626 will be used. If you have multiple caches in a cluster and
7627 get errors about IP-forwarding you must set them to have individual
7628 names with this setting.
7631 NAME: unique_hostname
7633 LOC: Config.uniqueHostname
7635 DEFAULT_DOC: Copy the value from visible_hostname
7637 If you want to have multiple machines with the same
7638 'visible_hostname' you must give each machine a different
7639 'unique_hostname' so forwarding loops can be detected.
7642 NAME: hostname_aliases
7644 LOC: Config.hostnameAliases
7647 A list of other DNS names your cache has.
7655 Minimum umask which should be enforced while the proxy
7656 is running, in addition to the umask set at startup.
7658 For a traditional octal representation of umasks, start
7663 HTTPD-ACCELERATOR OPTIONS
7664 -----------------------------------------------------------------------------
7667 NAME: httpd_accel_surrogate_id
7670 DEFAULT_DOC: visible_hostname is used if no specific ID is set.
7671 LOC: Config.Accel.surrogate_id
7673 Surrogates (http://www.esi.org/architecture_spec_1.0.html)
7674 need an identification token to allow control targeting. Because
7675 a farm of surrogates may all perform the same tasks, they may share
7676 an identification token.
7678 When the surrogate is a reverse-proxy, this ID is also
7679 used as cdn-id for CDN-Loop detection (RFC 8586).
7682 NAME: http_accel_surrogate_remote
7686 LOC: Config.onoff.surrogate_is_remote
7688 Remote surrogates (such as those in a CDN) honour the header
7689 "Surrogate-Control: no-store-remote".
7691 Set this to on to have squid behave as a remote surrogate.
7695 DELAY POOL PARAMETERS
7696 -----------------------------------------------------------------------------
7700 TYPE: delay_pool_count
7702 IFDEF: USE_DELAY_POOLS
7705 This represents the number of delay pools to be used. For example,
7706 if you have one class 2 delay pool and one class 3 delays pool, you
7707 have a total of 2 delay pools.
7709 See also delay_parameters, delay_class, delay_access for pool
7710 configuration details.
7714 TYPE: delay_pool_class
7716 IFDEF: USE_DELAY_POOLS
7719 This defines the class of each delay pool. There must be exactly one
7720 delay_class line for each delay pool. For example, to define two
7721 delay pools, one of class 2 and one of class 3, the settings above
7725 delay_pools 4 # 4 delay pools
7726 delay_class 1 2 # pool 1 is a class 2 pool
7727 delay_class 2 3 # pool 2 is a class 3 pool
7728 delay_class 3 4 # pool 3 is a class 4 pool
7729 delay_class 4 5 # pool 4 is a class 5 pool
7731 The delay pool classes are:
7733 class 1 Everything is limited by a single aggregate
7736 class 2 Everything is limited by a single aggregate
7737 bucket as well as an "individual" bucket chosen
7738 from bits 25 through 32 of the IPv4 address.
7740 class 3 Everything is limited by a single aggregate
7741 bucket as well as a "network" bucket chosen
7742 from bits 17 through 24 of the IP address and a
7743 "individual" bucket chosen from bits 17 through
7744 32 of the IPv4 address.
7746 class 4 Everything in a class 3 delay pool, with an
7747 additional limit on a per user basis. This
7748 only takes effect if the username is established
7749 in advance - by forcing authentication in your
7752 class 5 Requests are grouped according their tag (see
7753 external_acl's tag= reply).
7756 Each pool also requires a delay_parameters directive to configure the pool size
7757 and speed limits used whenever the pool is applied to a request. Along with
7758 a set of delay_access directives to determine when it is used.
7760 NOTE: If an IP address is a.b.c.d
7761 -> bits 25 through 32 are "d"
7762 -> bits 17 through 24 are "c"
7763 -> bits 17 through 32 are "c * 256 + d"
7765 NOTE-2: Due to the use of bitmasks in class 2,3,4 pools they only apply to
7766 IPv4 traffic. Class 1 and 5 pools may be used with IPv6 traffic.
7768 This clause only supports fast acl types.
7769 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
7771 See also delay_parameters and delay_access.
7775 TYPE: delay_pool_access
7777 DEFAULT_DOC: Deny using the pool, unless allow rules exist in squid.conf for the pool.
7778 IFDEF: USE_DELAY_POOLS
7781 This is used to determine which delay pool a request falls into.
7783 delay_access is sorted per pool and the matching starts with pool 1,
7784 then pool 2, ..., and finally pool N. The first delay pool where the
7785 request is allowed is selected for the request. If it does not allow
7786 the request to any pool then the request is not delayed (default).
7788 For example, if you want some_big_clients in delay
7789 pool 1 and lotsa_little_clients in delay pool 2:
7791 delay_access 1 allow some_big_clients
7792 delay_access 1 deny all
7793 delay_access 2 allow lotsa_little_clients
7794 delay_access 2 deny all
7795 delay_access 3 allow authenticated_clients
7797 See also delay_parameters and delay_class.
7801 NAME: delay_parameters
7802 TYPE: delay_pool_rates
7804 IFDEF: USE_DELAY_POOLS
7807 This defines the parameters for a delay pool. Each delay pool has
7808 a number of "buckets" associated with it, as explained in the
7809 description of delay_class.
7811 For a class 1 delay pool, the syntax is:
7813 delay_parameters pool aggregate
7815 For a class 2 delay pool:
7817 delay_parameters pool aggregate individual
7819 For a class 3 delay pool:
7821 delay_parameters pool aggregate network individual
7823 For a class 4 delay pool:
7825 delay_parameters pool aggregate network individual user
7827 For a class 5 delay pool:
7829 delay_parameters pool tagrate
7831 The option variables are:
7833 pool a pool number - ie, a number between 1 and the
7834 number specified in delay_pools as used in
7837 aggregate the speed limit parameters for the aggregate bucket
7840 individual the speed limit parameters for the individual
7841 buckets (class 2, 3).
7843 network the speed limit parameters for the network buckets
7846 user the speed limit parameters for the user buckets
7849 tagrate the speed limit parameters for the tag buckets
7852 A pair of delay parameters is written restore/maximum, where restore is
7853 the number of bytes (not bits - modem and network speeds are usually
7854 quoted in bits) per second placed into the bucket, and maximum is the
7855 maximum number of bytes which can be in the bucket at any time.
7857 There must be one delay_parameters line for each delay pool.
7860 For example, if delay pool number 1 is a class 2 delay pool as in the
7861 above example, and is being used to strictly limit each host to 64Kbit/sec
7862 (plus overheads), with no overall limit, the line is:
7864 delay_parameters 1 none 8000/8000
7866 Note that 8 x 8K Byte/sec -> 64K bit/sec.
7868 Note that the word 'none' is used to represent no limit.
7871 And, if delay pool number 2 is a class 3 delay pool as in the above
7872 example, and you want to limit it to a total of 256Kbit/sec (strict limit)
7873 with each 8-bit network permitted 64Kbit/sec (strict limit) and each
7874 individual host permitted 4800bit/sec with a bucket maximum size of 64Kbits
7875 to permit a decent web page to be downloaded at a decent speed
7876 (if the network is not being limited due to overuse) but slow down
7877 large downloads more significantly:
7879 delay_parameters 2 32000/32000 8000/8000 600/8000
7881 Note that 8 x 32K Byte/sec -> 256K bit/sec.
7882 8 x 8K Byte/sec -> 64K bit/sec.
7883 8 x 600 Byte/sec -> 4800 bit/sec.
7886 Finally, for a class 4 delay pool as in the example - each user will
7887 be limited to 128Kbits/sec no matter how many workstations they are logged into.:
7889 delay_parameters 4 32000/32000 8000/8000 600/64000 16000/16000
7892 See also delay_class and delay_access.
7896 NAME: delay_initial_bucket_level
7897 COMMENT: (percent, 0-100)
7900 IFDEF: USE_DELAY_POOLS
7901 LOC: Config.Delay.initial
7903 The initial bucket percentage is used to determine how much is put
7904 in each bucket when squid starts, is reconfigured, or first notices
7905 a host accessing it (in class 2 and class 3, individual hosts and
7906 networks only have buckets associated with them once they have been
7911 CLIENT DELAY POOL PARAMETERS
7912 -----------------------------------------------------------------------------
7915 NAME: client_delay_pools
7916 TYPE: client_delay_pool_count
7918 IFDEF: USE_DELAY_POOLS
7919 LOC: Config.ClientDelay
7921 This option specifies the number of client delay pools used. It must
7922 preceed other client_delay_* options.
7925 client_delay_pools 2
7927 See also client_delay_parameters and client_delay_access.
7930 NAME: client_delay_initial_bucket_level
7931 COMMENT: (percent, 0-no_limit)
7934 IFDEF: USE_DELAY_POOLS
7935 LOC: Config.ClientDelay.initial
7937 This option determines the initial bucket size as a percentage of
7938 max_bucket_size from client_delay_parameters. Buckets are created
7939 at the time of the "first" connection from the matching IP. Idle
7940 buckets are periodically deleted up.
7942 You can specify more than 100 percent but note that such "oversized"
7943 buckets are not refilled until their size goes down to max_bucket_size
7944 from client_delay_parameters.
7947 client_delay_initial_bucket_level 50
7950 NAME: client_delay_parameters
7951 TYPE: client_delay_pool_rates
7953 IFDEF: USE_DELAY_POOLS
7954 LOC: Config.ClientDelay
7957 This option configures client-side bandwidth limits using the
7960 client_delay_parameters pool speed_limit max_bucket_size
7962 pool is an integer ID used for client_delay_access matching.
7964 speed_limit is bytes added to the bucket per second.
7966 max_bucket_size is the maximum size of a bucket, enforced after any
7967 speed_limit additions.
7969 Please see the delay_parameters option for more information and
7973 client_delay_parameters 1 1024 2048
7974 client_delay_parameters 2 51200 16384
7976 See also client_delay_access.
7980 NAME: client_delay_access
7981 TYPE: client_delay_pool_access
7983 DEFAULT_DOC: Deny use of the pool, unless allow rules exist in squid.conf for the pool.
7984 IFDEF: USE_DELAY_POOLS
7985 LOC: Config.ClientDelay
7987 This option determines the client-side delay pool for the
7990 client_delay_access pool_ID allow|deny acl_name
7992 All client_delay_access options are checked in their pool ID
7993 order, starting with pool 1. The first checked pool with allowed
7994 request is selected for the request. If no ACL matches or there
7995 are no client_delay_access options, the request bandwidth is not
7998 The ACL-selected pool is then used to find the
7999 client_delay_parameters for the request. Client-side pools are
8000 not used to aggregate clients. Clients are always aggregated
8001 based on their source IP addresses (one bucket per source IP).
8003 This clause only supports fast acl types.
8004 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
8005 Additionally, only the client TCP connection details are available.
8006 ACLs testing HTTP properties will not work.
8008 Please see delay_access for more examples.
8011 client_delay_access 1 allow low_rate_network
8012 client_delay_access 2 allow vips_network
8015 See also client_delay_parameters and client_delay_pools.
8018 NAME: response_delay_pool
8019 TYPE: response_delay_pool_parameters
8021 IFDEF: USE_DELAY_POOLS
8022 LOC: Config.MessageDelay
8024 This option configures client response bandwidth limits using the
8027 response_delay_pool name [option=value] ...
8029 name the response delay pool name
8033 individual-restore The speed limit of an individual
8034 bucket(bytes/s). To be used in conjunction
8035 with 'individual-maximum'.
8037 individual-maximum The maximum number of bytes which can
8038 be placed into the individual bucket. To be used
8039 in conjunction with 'individual-restore'.
8041 aggregate-restore The speed limit for the aggregate
8042 bucket(bytes/s). To be used in conjunction with
8043 'aggregate-maximum'.
8045 aggregate-maximum The maximum number of bytes which can
8046 be placed into the aggregate bucket. To be used
8047 in conjunction with 'aggregate-restore'.
8049 initial-bucket-level The initial bucket size as a percentage
8050 of individual-maximum.
8052 Individual and(or) aggregate bucket options may not be specified,
8053 meaning no individual and(or) aggregate speed limitation.
8054 See also response_delay_pool_access and delay_parameters for
8055 terminology details.
8058 NAME: response_delay_pool_access
8059 TYPE: response_delay_pool_access
8061 DEFAULT_DOC: Deny use of the pool, unless allow rules exist in squid.conf for the pool.
8062 IFDEF: USE_DELAY_POOLS
8063 LOC: Config.MessageDelay
8065 Determines whether a specific named response delay pool is used
8066 for the transaction. The syntax for this directive is:
8068 response_delay_pool_access pool_name allow|deny acl_name
8070 All response_delay_pool_access options are checked in the order
8071 they appear in this configuration file. The first rule with a
8072 matching ACL wins. If (and only if) an "allow" rule won, Squid
8073 assigns the response to the corresponding named delay pool.
8077 WCCPv1 AND WCCPv2 CONFIGURATION OPTIONS
8078 -----------------------------------------------------------------------------
8083 LOC: Config.Wccp.router
8085 DEFAULT_DOC: WCCP disabled.
8088 Use this option to define your WCCP ``home'' router for
8091 wccp_router supports a single WCCP(v1) router
8093 wccp2_router supports multiple WCCPv2 routers
8095 only one of the two may be used at the same time and defines
8096 which version of WCCP to use.
8100 TYPE: IpAddress_list
8101 LOC: Config.Wccp2.router
8103 DEFAULT_DOC: WCCPv2 disabled.
8106 Use this option to define your WCCP ``home'' router for
8109 wccp_router supports a single WCCP(v1) router
8111 wccp2_router supports multiple WCCPv2 routers
8113 only one of the two may be used at the same time and defines
8114 which version of WCCP to use.
8119 LOC: Config.Wccp.version
8123 This directive is only relevant if you need to set up WCCP(v1)
8124 to some very old and end-of-life Cisco routers. In all other
8125 setups it must be left unset or at the default setting.
8126 It defines an internal version in the WCCP(v1) protocol,
8127 with version 4 being the officially documented protocol.
8129 According to some users, Cisco IOS 11.2 and earlier only
8130 support WCCP version 3. If you're using that or an earlier
8131 version of IOS, you may need to change this value to 3, otherwise
8132 do not specify this parameter.
8135 NAME: wccp2_rebuild_wait
8137 LOC: Config.Wccp2.rebuildwait
8141 If this is enabled Squid will wait for the cache dir rebuild to finish
8142 before sending the first wccp2 HereIAm packet
8145 NAME: wccp2_forwarding_method
8147 LOC: Config.Wccp2.forwarding_method
8151 WCCP2 allows the setting of forwarding methods between the
8152 router/switch and the cache. Valid values are as follows:
8154 gre - GRE encapsulation (forward the packet in a GRE/WCCP tunnel)
8155 l2 - L2 redirect (forward the packet using Layer 2/MAC rewriting)
8157 Currently (as of IOS 12.4) cisco routers only support GRE.
8158 Cisco switches only support the L2 redirect assignment method.
8161 NAME: wccp2_return_method
8163 LOC: Config.Wccp2.return_method
8167 WCCP2 allows the setting of return methods between the
8168 router/switch and the cache for packets that the cache
8169 decides not to handle. Valid values are as follows:
8171 gre - GRE encapsulation (forward the packet in a GRE/WCCP tunnel)
8172 l2 - L2 redirect (forward the packet using Layer 2/MAC rewriting)
8174 Currently (as of IOS 12.4) cisco routers only support GRE.
8175 Cisco switches only support the L2 redirect assignment.
8177 If the "ip wccp redirect exclude in" command has been
8178 enabled on the cache interface, then it is still safe for
8179 the proxy server to use a l2 redirect method even if this
8180 option is set to GRE.
8183 NAME: wccp2_assignment_method
8185 LOC: Config.Wccp2.assignment_method
8189 WCCP2 allows the setting of methods to assign the WCCP hash
8190 Valid values are as follows:
8192 hash - Hash assignment
8193 mask - Mask assignment
8195 As a general rule, cisco routers support the hash assignment method
8196 and cisco switches support the mask assignment method.
8201 LOC: Config.Wccp2.info
8202 DEFAULT_IF_NONE: standard 0
8203 DEFAULT_DOC: Use the 'web-cache' standard service.
8206 WCCP2 allows for multiple traffic services. There are two
8207 types: "standard" and "dynamic". The standard type defines
8208 one service id - http (id 0). The dynamic service ids can be from
8209 51 to 255 inclusive. In order to use a dynamic service id
8210 one must define the type of traffic to be redirected; this is done
8211 using the wccp2_service_info option.
8213 The "standard" type does not require a wccp2_service_info option,
8214 just specifying the service id will suffice.
8216 MD5 service authentication can be enabled by adding
8217 "password=<password>" to the end of this service declaration.
8221 wccp2_service standard 0 # for the 'web-cache' standard service
8222 wccp2_service dynamic 80 # a dynamic service type which will be
8223 # fleshed out with subsequent options.
8224 wccp2_service standard 0 password=foo
8227 NAME: wccp2_service_info
8228 TYPE: wccp2_service_info
8229 LOC: Config.Wccp2.info
8233 Dynamic WCCPv2 services require further information to define the
8234 traffic you wish to have diverted.
8238 wccp2_service_info <id> protocol=<protocol> flags=<flag>,<flag>..
8239 priority=<priority> ports=<port>,<port>..
8241 The relevant WCCPv2 flags:
8242 + src_ip_hash, dst_ip_hash
8243 + source_port_hash, dst_port_hash
8244 + src_ip_alt_hash, dst_ip_alt_hash
8245 + src_port_alt_hash, dst_port_alt_hash
8248 The port list can be one to eight entries.
8252 wccp2_service_info 80 protocol=tcp flags=src_ip_hash,ports_source
8253 priority=240 ports=80
8255 Note: the service id must have been defined by a previous
8256 'wccp2_service dynamic <id>' entry.
8261 LOC: Config.Wccp2.weight
8265 Each cache server gets assigned a set of the destination
8266 hash proportional to their weight.
8271 LOC: Config.Wccp.address
8273 DEFAULT_DOC: Address selected by the operating system.
8276 Use this option if you require WCCP(v1) to use a specific
8279 The default behavior is to not bind to any specific address.
8284 LOC: Config.Wccp2.address
8286 DEFAULT_DOC: Address selected by the operating system.
8289 Use this option if you require WCCPv2 to use a specific
8292 The default behavior is to not bind to any specific address.
8296 PERSISTENT CONNECTION HANDLING
8297 -----------------------------------------------------------------------------
8299 Also see "pconn_timeout" in the TIMEOUTS section
8302 NAME: client_persistent_connections
8304 LOC: Config.onoff.client_pconns
8307 Persistent connection support for clients.
8308 Squid uses persistent connections (when allowed). You can use
8309 this option to disable persistent connections with clients.
8312 NAME: server_persistent_connections
8314 LOC: Config.onoff.server_pconns
8317 Persistent connection support for servers.
8318 Squid uses persistent connections (when allowed). You can use
8319 this option to disable persistent connections with servers.
8322 NAME: persistent_connection_after_error
8324 LOC: Config.onoff.error_pconns
8327 With this directive the use of persistent connections after
8328 HTTP errors can be disabled. Useful if you have clients
8329 who fail to handle errors on persistent connections proper.
8332 NAME: detect_broken_pconn
8334 LOC: Config.onoff.detect_broken_server_pconns
8337 Some servers have been found to incorrectly signal the use
8338 of HTTP/1.0 persistent connections even on replies not
8339 compatible, causing significant delays. This server problem
8340 has mostly been seen on redirects.
8342 By enabling this directive Squid attempts to detect such
8343 broken replies and automatically assume the reply is finished
8344 after 10 seconds timeout.
8348 CACHE DIGEST OPTIONS
8349 -----------------------------------------------------------------------------
8352 NAME: digest_generation
8353 IFDEF: USE_CACHE_DIGESTS
8355 LOC: Config.onoff.digest_generation
8358 This controls whether the server will generate a Cache Digest
8359 of its contents. By default, Cache Digest generation is
8360 enabled if Squid is compiled with --enable-cache-digests defined.
8363 NAME: digest_bits_per_entry
8364 IFDEF: USE_CACHE_DIGESTS
8366 LOC: Config.digest.bits_per_entry
8369 This is the number of bits of the server's Cache Digest which
8370 will be associated with the Digest entry for a given HTTP
8371 Method and URL (public key) combination. The default is 5.
8374 NAME: digest_rebuild_period
8375 IFDEF: USE_CACHE_DIGESTS
8378 LOC: Config.digest.rebuild_period
8381 This is the wait time between Cache Digest rebuilds.
8384 NAME: digest_rewrite_period
8386 IFDEF: USE_CACHE_DIGESTS
8388 LOC: Config.digest.rewrite_period
8391 This is the wait time between Cache Digest writes to
8395 NAME: digest_swapout_chunk_size
8398 IFDEF: USE_CACHE_DIGESTS
8399 LOC: Config.digest.swapout_chunk_size
8402 This is the number of bytes of the Cache Digest to write to
8403 disk at a time. It defaults to 4096 bytes (4KB), the Squid
8407 NAME: digest_rebuild_chunk_percentage
8408 COMMENT: (percent, 0-100)
8409 IFDEF: USE_CACHE_DIGESTS
8411 LOC: Config.digest.rebuild_chunk_percentage
8414 This is the percentage of the Cache Digest to be scanned at a
8415 time. By default it is set to 10% of the Cache Digest.
8420 -----------------------------------------------------------------------------
8425 LOC: Config.Port.snmp
8427 DEFAULT_DOC: SNMP disabled.
8430 The port number where Squid listens for SNMP requests. To enable
8431 SNMP support set this to a suitable port number. Port number
8432 3401 is often used for the Squid SNMP agent. By default it's
8433 set to "0" (disabled)
8441 LOC: Config.accessList.snmp
8443 DEFAULT_DOC: Deny, unless rules exist in squid.conf.
8446 Allowing or denying access to the SNMP port.
8448 All access to the agent is denied by default.
8451 snmp_access allow|deny [!]aclname ...
8453 This clause only supports fast acl types.
8454 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
8457 snmp_access allow snmppublic localhost
8458 snmp_access deny all
8461 NAME: snmp_incoming_address
8463 LOC: Config.Addrs.snmp_incoming
8465 DEFAULT_DOC: Accept SNMP packets from all machine interfaces.
8468 Just like 'udp_incoming_address', but for the SNMP port.
8470 snmp_incoming_address is used for the SNMP socket receiving
8471 messages from SNMP agents.
8473 The default snmp_incoming_address is to listen on all
8474 available network interfaces.
8477 NAME: snmp_outgoing_address
8479 LOC: Config.Addrs.snmp_outgoing
8481 DEFAULT_DOC: Use snmp_incoming_address or an address selected by the operating system.
8484 Just like 'udp_outgoing_address', but for the SNMP port.
8486 snmp_outgoing_address is used for SNMP packets returned to SNMP
8489 If snmp_outgoing_address is not set it will use the same socket
8490 as snmp_incoming_address. Only change this if you want to have
8491 SNMP replies sent using another address than where this Squid
8492 listens for SNMP queries.
8494 NOTE, snmp_incoming_address and snmp_outgoing_address can not have
8495 the same value since they both use the same port.
8500 -----------------------------------------------------------------------------
8503 NAME: icp_port udp_port
8506 DEFAULT_DOC: ICP disabled.
8507 LOC: Config.Port.icp
8509 The port number where Squid sends and receives ICP queries to
8510 and from neighbor caches. The standard UDP port for ICP is 3130.
8520 DEFAULT_DOC: HTCP disabled.
8521 LOC: Config.Port.htcp
8523 The port number where Squid sends and receives HTCP queries to
8524 and from neighbor caches. To turn it on you want to set it to
8531 NAME: log_icp_queries
8535 LOC: Config.onoff.log_udp
8537 If set, ICP queries are logged to access.log. You may wish
8538 do disable this if your ICP load is VERY high to speed things
8539 up or to simplify log analysis.
8542 NAME: udp_incoming_address
8544 LOC:Config.Addrs.udp_incoming
8546 DEFAULT_DOC: Accept packets from all machine interfaces.
8548 udp_incoming_address is used for UDP packets received from other
8551 The default behavior is to not bind to any specific address.
8553 Only change this if you want to have all UDP queries received on
8554 a specific interface/address.
8556 NOTE: udp_incoming_address is used by the ICP, HTCP, and DNS
8557 modules. Altering it will affect all of them in the same manner.
8559 see also; udp_outgoing_address
8561 NOTE, udp_incoming_address and udp_outgoing_address can not
8562 have the same value since they both use the same port.
8565 NAME: udp_outgoing_address
8567 LOC: Config.Addrs.udp_outgoing
8569 DEFAULT_DOC: Use udp_incoming_address or an address selected by the operating system.
8571 udp_outgoing_address is used for UDP packets sent out to other
8574 The default behavior is to not bind to any specific address.
8576 Instead it will use the same socket as udp_incoming_address.
8577 Only change this if you want to have UDP queries sent using another
8578 address than where this Squid listens for UDP queries from other
8581 NOTE: udp_outgoing_address is used by the ICP, HTCP, and DNS
8582 modules. Altering it will affect all of them in the same manner.
8584 see also; udp_incoming_address
8586 NOTE, udp_incoming_address and udp_outgoing_address can not
8587 have the same value since they both use the same port.
8594 LOC: Config.onoff.icp_hit_stale
8596 If you want to return ICP_HIT for stale cache objects, set this
8597 option to 'on'. If you have sibling relationships with caches
8598 in other administrative domains, this should be 'off'. If you only
8599 have sibling relationships with caches under your control,
8600 it is probably okay to set this to 'on'.
8601 If set to 'on', your siblings should use the option "allow-miss"
8602 on their cache_peer lines for connecting to you.
8605 NAME: minimum_direct_hops
8608 LOC: Config.minDirectHops
8610 If using the ICMP pinging stuff, do direct fetches for sites
8611 which are no more than this many hops away.
8614 NAME: minimum_direct_rtt
8618 LOC: Config.minDirectRtt
8620 If using the ICMP pinging stuff, do direct fetches for sites
8621 which are no more than this many rtt milliseconds away.
8627 LOC: Config.Netdb.low
8629 The low water mark for the ICMP measurement database.
8631 Note: high watermark controlled by netdb_high directive.
8633 These watermarks are counts, not percents. The defaults are
8634 (low) 900 and (high) 1000. When the high water mark is
8635 reached, database entries will be deleted until the low
8642 LOC: Config.Netdb.high
8644 The high water mark for the ICMP measurement database.
8646 Note: low watermark controlled by netdb_low directive.
8648 These watermarks are counts, not percents. The defaults are
8649 (low) 900 and (high) 1000. When the high water mark is
8650 reached, database entries will be deleted until the low
8654 NAME: netdb_ping_period
8656 LOC: Config.Netdb.period
8659 The minimum period for measuring a site. There will be at
8660 least this much delay between successive pings to the same
8661 network. The default is five minutes.
8668 LOC: Config.onoff.query_icmp
8670 If you want to ask your peers to include ICMP data in their ICP
8671 replies, enable this option.
8673 If your peer has configured Squid (during compilation) with
8674 '--enable-icmp' that peer will send ICMP pings to origin server
8675 sites of the URLs it receives. If you enable this option the
8676 ICP replies from that peer will include the ICMP data (if available).
8677 Then, when choosing a parent cache, Squid will choose the parent with
8678 the minimal RTT to the origin server. When this happens, the
8679 hierarchy field of the access.log will be
8680 "CLOSEST_PARENT_MISS". This option is off by default.
8683 NAME: test_reachability
8687 LOC: Config.onoff.test_reachability
8689 When this is 'on', ICP MISS replies will be ICP_MISS_NOFETCH
8690 instead of ICP_MISS if the target host is NOT in the ICMP
8691 database, or has a zero RTT.
8694 NAME: icp_query_timeout
8697 DEFAULT_DOC: Dynamic detection.
8699 LOC: Config.Timeout.icp_query
8701 Normally Squid will automatically determine an optimal ICP
8702 query timeout value based on the round-trip-time of recent ICP
8703 queries. If you want to override the value determined by
8704 Squid, set this 'icp_query_timeout' to a non-zero value. This
8705 value is specified in MILLISECONDS, so, to use a 2-second
8706 timeout (the old default), you would write:
8708 icp_query_timeout 2000
8711 NAME: maximum_icp_query_timeout
8715 LOC: Config.Timeout.icp_query_max
8717 Normally the ICP query timeout is determined dynamically. But
8718 sometimes it can lead to very large values (say 5 seconds).
8719 Use this option to put an upper limit on the dynamic timeout
8720 value. Do NOT use this option to always use a fixed (instead
8721 of a dynamic) timeout value. To set a fixed timeout see the
8722 'icp_query_timeout' directive.
8725 NAME: minimum_icp_query_timeout
8729 LOC: Config.Timeout.icp_query_min
8731 Normally the ICP query timeout is determined dynamically. But
8732 sometimes it can lead to very small timeouts, even lower than
8733 the normal latency variance on your link due to traffic.
8734 Use this option to put an lower limit on the dynamic timeout
8735 value. Do NOT use this option to always use a fixed (instead
8736 of a dynamic) timeout value. To set a fixed timeout see the
8737 'icp_query_timeout' directive.
8740 NAME: background_ping_rate
8744 LOC: Config.backgroundPingRate
8746 Controls how often the ICP pings are sent to siblings that
8747 have background-ping set.
8751 MULTICAST ICP OPTIONS
8752 -----------------------------------------------------------------------------
8757 LOC: Config.mcast_group_list
8760 This tag specifies a list of multicast groups which your server
8761 should join to receive multicasted ICP queries.
8763 NOTE! Be very careful what you put here! Be sure you
8764 understand the difference between an ICP _query_ and an ICP
8765 _reply_. This option is to be set only if you want to RECEIVE
8766 multicast queries. Do NOT set this option to SEND multicast
8767 ICP (use cache_peer for that). ICP replies are always sent via
8768 unicast, so this option does not affect whether or not you will
8769 receive replies from multicast group members.
8771 You must be very careful to NOT use a multicast address which
8772 is already in use by another group of caches.
8774 If you are unsure about multicast, please read the Multicast
8775 chapter in the Squid FAQ (http://www.squid-cache.org/FAQ/).
8777 Usage: mcast_groups 239.128.16.128 224.0.1.20
8779 By default, Squid doesn't listen on any multicast groups.
8782 NAME: mcast_icp_query_timeout
8786 LOC: Config.Timeout.mcast_icp_query
8788 For multicast peers, Squid regularly sends out ICP "probes" to
8789 count how many other peers are listening on the given multicast
8790 address. This value specifies how long Squid should wait to
8791 count all the replies. The default is 2000 msec, or 2
8796 INTERNAL ICON OPTIONS
8797 -----------------------------------------------------------------------------
8800 NAME: icon_directory
8802 LOC: Config.icons.directory
8803 DEFAULT: @DEFAULT_ICON_DIR@
8805 Where the icons are stored. These are normally kept in
8809 NAME: global_internal_static
8811 LOC: Config.onoff.global_internal_static
8814 This directive controls is Squid should intercept all requests for
8815 /squid-internal-static/ no matter which host the URL is requesting
8816 (default on setting), or if nothing special should be done for
8817 such URLs (off setting). The purpose of this directive is to make
8818 icons etc work better in complex cache hierarchies where it may
8819 not always be possible for all corners in the cache mesh to reach
8820 the server generating a directory listing.
8823 NAME: short_icon_urls
8825 LOC: Config.icons.use_short_names
8828 If this is enabled Squid will use short URLs for icons.
8829 If disabled it will revert to the old behavior of including
8830 it's own name and port in the URL.
8832 If you run a complex cache hierarchy with a mix of Squid and
8833 other proxies you may need to disable this directive.
8838 -----------------------------------------------------------------------------
8841 NAME: error_directory
8843 LOC: Config.errorDirectory
8845 DEFAULT_DOC: Send error pages in the clients preferred language
8847 If you wish to create your own versions of the default
8848 error files to customize them to suit your company copy
8849 the error/template files to another directory and point
8852 WARNING: This option will disable multi-language support
8853 on error pages if used.
8855 The squid developers are interested in making squid available in
8856 a wide variety of languages. If you are making translations for a
8857 language that Squid does not currently provide please consider
8858 contributing your translation back to the project.
8859 https://wiki.squid-cache.org/Translations
8861 The squid developers working on translations are happy to supply drop-in
8862 translated error files in exchange for any new language contributions.
8865 NAME: error_default_language
8866 IFDEF: USE_ERR_LOCALES
8868 LOC: Config.errorDefaultLanguage
8870 DEFAULT_DOC: Generate English language pages.
8872 Set the default language which squid will send error pages in
8873 if no existing translation matches the clients language
8876 If unset (default) generic English will be used.
8878 The squid developers are interested in making squid available in
8879 a wide variety of languages. If you are interested in making
8880 translations for any language see the squid wiki for details.
8881 https://wiki.squid-cache.org/Translations
8884 NAME: error_log_languages
8885 IFDEF: USE_ERR_LOCALES
8887 LOC: Config.errorLogMissingLanguages
8890 Log to cache.log what languages users are attempting to
8891 auto-negotiate for translations.
8893 Successful negotiations are not logged. Only failures
8894 have meaning to indicate that Squid may need an upgrade
8895 of its error page translations.
8898 NAME: err_page_stylesheet
8900 LOC: Config.errorStylesheet
8901 DEFAULT: @DEFAULT_CONFIG_DIR@/errorpage.css
8903 CSS Stylesheet to pattern the display of Squid default error pages.
8905 For information on CSS see http://www.w3.org/Style/CSS/
8910 LOC: Config.errHtmlText
8913 HTML text to include in error messages. Make this a "mailto"
8914 URL to your admin address, or maybe just a link to your
8915 organizations Web page.
8917 To include this in your error messages, you must rewrite
8918 the error template files (found in the "errors" directory).
8919 Wherever you want the 'err_html_text' line to appear,
8920 insert a %L tag in the error template file.
8923 NAME: email_err_data
8926 LOC: Config.onoff.emailErrData
8929 If enabled, information about the occurred error will be
8930 included in the mailto links of the ERR pages (if %W is set)
8931 so that the email body contains the data.
8932 Syntax is <A HREF="mailto:%w%W">%w</A>
8935 Request headers and other included facts may contain
8936 sensitive information about transaction history, the
8937 Squid instance, and its environment which would be
8938 unavailable to error recipients otherwise.
8943 LOC: Config.denyInfoList
8946 Usage: deny_info err_page_name acl
8947 or deny_info http://... acl
8948 or deny_info TCP_RESET acl
8950 This can be used to return a ERR_ page for requests which
8951 do not pass the 'http_access' rules. Squid remembers the last
8952 acl it evaluated in http_access, and if a 'deny_info' line exists
8953 for that ACL Squid returns a corresponding error page.
8955 The acl is typically the last acl on the http_access deny line which
8956 denied access. The exceptions to this rule are:
8957 - When Squid needs to request authentication credentials. It's then
8958 the first authentication related acl encountered
8959 - When none of the http_access lines matches. It's then the last
8960 acl processed on the last http_access line.
8961 - When the decision to deny access was made by an adaptation service,
8962 the acl name is the corresponding eCAP or ICAP service_name.
8964 NP: If providing your own custom error pages with error_directory
8965 you may also specify them by your custom file name:
8966 Example: deny_info ERR_CUSTOM_ACCESS_DENIED bad_guys
8968 By default Squid will send "403 Forbidden". A different 4xx or 5xx
8969 may be specified by prefixing the file name with the code and a colon.
8970 e.g. 404:ERR_CUSTOM_ACCESS_DENIED
8972 Alternatively you can tell Squid to reset the TCP connection
8973 by specifying TCP_RESET.
8975 Or you can specify an error URL or URL pattern. The browsers will
8976 get redirected to the specified URL after formatting tags have
8977 been replaced. Redirect will be done with 302 or 307 according to
8978 HTTP/1.1 specs. A different 3xx code may be specified by prefixing
8979 the URL. e.g. 303:http://example.com/
8982 %a - username (if available. Password NOT included)
8983 %A - Local listening IP address the client connection was connected to
8986 %E - Error description
8988 %H - Request domain name
8989 %i - Client IP Address
8991 %O - Unescaped message result from external ACL helper
8992 %o - Message result from external ACL helper
8993 %p - Request Port number
8994 %P - Request Protocol name
8995 %R - Request URL path
8996 %T - Timestamp in RFC 1123 format
8997 %U - Full canonical URL from client
8998 (HTTPS URLs terminate with *)
8999 %u - Full canonical URL from client
9000 %w - Admin email from squid.conf
9002 %% - Literal percent (%) code
9007 OPTIONS INFLUENCING REQUEST FORWARDING
9008 -----------------------------------------------------------------------------
9011 NAME: nonhierarchical_direct
9013 LOC: Config.onoff.nonhierarchical_direct
9016 By default, Squid will send any non-hierarchical requests
9017 (not cacheable request type) direct to origin servers.
9019 When this is set to "off", Squid will prefer to send these
9020 requests to parents.
9022 Note that in most configurations, by turning this off you will only
9023 add latency to these request without any improvement in global hit
9026 This option only sets a preference. If the parent is unavailable a
9027 direct connection to the origin server may still be attempted. To
9028 completely prevent direct connections use never_direct.
9033 LOC: Config.onoff.prefer_direct
9036 Normally Squid tries to use parents for most requests. If you for some
9037 reason like it to first try going direct and only use a parent if
9038 going direct fails set this to on.
9040 By combining nonhierarchical_direct off and prefer_direct on you
9041 can set up Squid to use a parent as a backup path if going direct
9044 Note: If you want Squid to use parents for all requests see
9045 the never_direct directive. prefer_direct only modifies how Squid
9046 acts on cacheable requests.
9049 NAME: cache_miss_revalidate
9053 LOC: Config.onoff.cache_miss_revalidate
9055 RFC 7232 defines a conditional request mechanism to prevent
9056 response objects being unnecessarily transferred over the network.
9057 If that mechanism is used by the client and a cache MISS occurs
9058 it can prevent new cache entries being created.
9060 This option determines whether Squid on cache MISS will pass the
9061 client revalidation request to the server or tries to fetch new
9062 content for caching. It can be useful while the cache is mostly
9063 empty to more quickly have the cache populated by generating
9064 non-conditional GETs.
9066 When set to 'on' (default), Squid will pass all client If-* headers
9067 to the server. This permits server responses without a cacheable
9068 payload to be delivered and on MISS no new cache entry is created.
9070 When set to 'off' and if the request is cacheable, Squid will
9071 remove the clients If-Modified-Since and If-None-Match headers from
9072 the request sent to the server. This requests a 200 status response
9073 from the server to create a new cache entry with.
9078 LOC: Config.accessList.AlwaysDirect
9080 DEFAULT_DOC: Prevent any cache_peer being used for this request.
9082 Usage: always_direct allow|deny [!]aclname ...
9084 Here you can use ACL elements to specify requests which should
9085 ALWAYS be forwarded by Squid to the origin servers without using
9086 any peers. For example, to always directly forward requests for
9087 local servers ignoring any parents or siblings you may have use
9090 acl local-servers dstdomain my.domain.net
9091 always_direct allow local-servers
9093 To always forward FTP requests directly, use
9096 always_direct allow FTP
9098 NOTE: There is a similar, but opposite option named
9099 'never_direct'. You need to be aware that "always_direct deny
9100 foo" is NOT the same thing as "never_direct allow foo". You
9101 may need to use a deny rule to exclude a more-specific case of
9102 some other rule. Example:
9104 acl local-external dstdomain external.foo.net
9105 acl local-servers dstdomain .foo.net
9106 always_direct deny local-external
9107 always_direct allow local-servers
9109 NOTE: If your goal is to make the client forward the request
9110 directly to the origin server bypassing Squid then this needs
9111 to be done in the client configuration. Squid configuration
9112 can only tell Squid how Squid should fetch the object.
9114 NOTE: This directive is not related to caching. The replies
9115 is cached as usual even if you use always_direct. To not cache
9116 the replies see the 'cache' directive.
9118 This clause supports both fast and slow acl types.
9119 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
9124 LOC: Config.accessList.NeverDirect
9126 DEFAULT_DOC: Allow DNS results to be used for this request.
9128 Usage: never_direct allow|deny [!]aclname ...
9130 never_direct is the opposite of always_direct. Please read
9131 the description for always_direct if you have not already.
9133 With 'never_direct' you can use ACL elements to specify
9134 requests which should NEVER be forwarded directly to origin
9135 servers. For example, to force the use of a proxy for all
9136 requests, except those in your local domain use something like:
9138 acl local-servers dstdomain .foo.net
9139 never_direct deny local-servers
9140 never_direct allow all
9142 or if Squid is inside a firewall and there are local intranet
9143 servers inside the firewall use something like:
9145 acl local-intranet dstdomain .foo.net
9146 acl local-external dstdomain external.foo.net
9147 always_direct deny local-external
9148 always_direct allow local-intranet
9149 never_direct allow all
9151 This clause supports both fast and slow acl types.
9152 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
9156 ADVANCED NETWORKING OPTIONS
9157 -----------------------------------------------------------------------------
9160 NAME: incoming_udp_average incoming_icp_average
9163 LOC: Config.comm_incoming.udp.average
9165 Heavy voodoo here. I can't even believe you are reading this.
9166 Are you crazy? Don't even think about adjusting these unless
9167 you understand the algorithms in comm_select.c first!
9170 NAME: incoming_tcp_average incoming_http_average
9173 LOC: Config.comm_incoming.tcp.average
9175 Heavy voodoo here. I can't even believe you are reading this.
9176 Are you crazy? Don't even think about adjusting these unless
9177 you understand the algorithms in comm_select.c first!
9180 NAME: incoming_dns_average
9183 LOC: Config.comm_incoming.dns.average
9185 Heavy voodoo here. I can't even believe you are reading this.
9186 Are you crazy? Don't even think about adjusting these unless
9187 you understand the algorithms in comm_select.c first!
9190 NAME: min_udp_poll_cnt min_icp_poll_cnt
9193 LOC: Config.comm_incoming.udp.min_poll
9195 Heavy voodoo here. I can't even believe you are reading this.
9196 Are you crazy? Don't even think about adjusting these unless
9197 you understand the algorithms in comm_select.c first!
9200 NAME: min_dns_poll_cnt
9203 LOC: Config.comm_incoming.dns.min_poll
9205 Heavy voodoo here. I can't even believe you are reading this.
9206 Are you crazy? Don't even think about adjusting these unless
9207 you understand the algorithms in comm_select.c first!
9210 NAME: min_tcp_poll_cnt min_http_poll_cnt
9213 LOC: Config.comm_incoming.tcp.min_poll
9215 Heavy voodoo here. I can't even believe you are reading this.
9216 Are you crazy? Don't even think about adjusting these unless
9217 you understand the algorithms in comm_select.c first!
9223 LOC: Config.accept_filter
9227 The name of an accept(2) filter to install on Squid's
9228 listen socket(s). This feature is perhaps specific to
9229 FreeBSD and requires support in the kernel.
9231 The 'httpready' filter delays delivering new connections
9232 to Squid until a full HTTP request has been received.
9233 See the accf_http(9) man page for details.
9235 The 'dataready' filter delays delivering new connections
9236 to Squid until there is some data to process.
9237 See the accf_dataready(9) man page for details.
9241 The 'data' filter delays delivering of new connections
9242 to Squid until there is some data to process by TCP_ACCEPT_DEFER.
9243 You may optionally specify a number of seconds to wait by
9244 'data=N' where N is the number of seconds. Defaults to 30
9245 if not specified. See the tcp(7) man page for details.
9248 accept_filter httpready
9253 NAME: client_ip_max_connections
9255 LOC: Config.client_ip_max_connections
9257 DEFAULT_DOC: No limit.
9259 Set an absolute limit on the number of connections a single
9260 client IP can use. Any more than this and Squid will begin to drop
9261 new connections from the client until it closes some links.
9263 Note that this is a global limit. It affects all HTTP, HTCP, and FTP
9264 connections from the client. For finer control use the ACL access controls.
9266 Requires client_db to be enabled (the default).
9268 WARNING: This may noticeably slow down traffic received via external proxies
9269 or NAT devices and cause them to rebound error messages back to their clients.
9272 NAME: tcp_recv_bufsize
9276 DEFAULT_DOC: Use operating system TCP defaults.
9277 LOC: Config.tcpRcvBufsz
9279 Size of receive buffer to set for TCP sockets. Probably just
9280 as easy to change your kernel's default.
9281 Omit from squid.conf to use the default buffer size.
9286 -----------------------------------------------------------------------------
9293 LOC: Adaptation::Icap::TheConfig.onoff
9296 If you want to enable the ICAP module support, set this to on.
9299 NAME: icap_connect_timeout
9302 LOC: Adaptation::Icap::TheConfig.connect_timeout_raw
9305 This parameter specifies how long to wait for the TCP connect to
9306 the requested ICAP server to complete before giving up and either
9307 terminating the HTTP transaction or bypassing the failure.
9309 The default for optional services is peer_connect_timeout.
9310 The default for essential services is connect_timeout.
9311 If this option is explicitly set, its value applies to all services.
9314 NAME: icap_io_timeout
9318 DEFAULT_DOC: Use read_timeout.
9319 LOC: Adaptation::Icap::TheConfig.io_timeout_raw
9322 This parameter specifies how long to wait for an I/O activity on
9323 an established, active ICAP connection before giving up and
9324 either terminating the HTTP transaction or bypassing the
9328 NAME: icap_service_failure_limit
9329 COMMENT: limit [in memory-depth time-units]
9330 TYPE: icap_service_failure_limit
9332 LOC: Adaptation::Icap::TheConfig
9335 The limit specifies the number of failures that Squid tolerates
9336 when establishing a new TCP connection with an ICAP service. If
9337 the number of failures exceeds the limit, the ICAP service is
9338 not used for new ICAP requests until it is time to refresh its
9341 A negative value disables the limit. Without the limit, an ICAP
9342 service will not be considered down due to connectivity failures
9343 between ICAP OPTIONS requests.
9345 Squid forgets ICAP service failures older than the specified
9346 value of memory-depth. The memory fading algorithm
9347 is approximate because Squid does not remember individual
9348 errors but groups them instead, splitting the option
9349 value into ten time slots of equal length.
9351 When memory-depth is 0 and by default this option has no
9352 effect on service failure expiration.
9354 Squid always forgets failures when updating service settings
9355 using an ICAP OPTIONS transaction, regardless of this option
9359 # suspend service usage after 10 failures in 5 seconds:
9360 icap_service_failure_limit 10 in 5 seconds
9363 NAME: icap_service_revival_delay
9366 LOC: Adaptation::Icap::TheConfig.service_revival_delay
9369 The delay specifies the number of seconds to wait after an ICAP
9370 OPTIONS request failure before requesting the options again. The
9371 failed ICAP service is considered "down" until fresh OPTIONS are
9374 The actual delay cannot be smaller than the hardcoded minimum
9375 delay of 30 seconds.
9378 NAME: icap_preview_enable
9382 LOC: Adaptation::Icap::TheConfig.preview_enable
9385 The ICAP Preview feature allows the ICAP server to handle the
9386 HTTP message by looking only at the beginning of the message body
9387 or even without receiving the body at all. In some environments,
9388 previews greatly speedup ICAP processing.
9390 During an ICAP OPTIONS transaction, the server may tell Squid what
9391 HTTP messages should be previewed and how big the preview should be.
9392 Squid will not use Preview if the server did not request one.
9394 To disable ICAP Preview for all ICAP services, regardless of
9395 individual ICAP server OPTIONS responses, set this option to "off".
9397 icap_preview_enable off
9400 NAME: icap_preview_size
9403 LOC: Adaptation::Icap::TheConfig.preview_size
9405 DEFAULT_DOC: No preview sent.
9407 The default size of preview data to be sent to the ICAP server.
9408 This value might be overwritten on a per server basis by OPTIONS requests.
9411 NAME: icap_206_enable
9415 LOC: Adaptation::Icap::TheConfig.allow206_enable
9418 206 (Partial Content) responses is an ICAP extension that allows the
9419 ICAP agents to optionally combine adapted and original HTTP message
9420 content. The decision to combine is postponed until the end of the
9421 ICAP response. Squid supports Partial Content extension by default.
9423 Activation of the Partial Content extension is negotiated with each
9424 ICAP service during OPTIONS exchange. Most ICAP servers should handle
9425 negotiation correctly even if they do not support the extension, but
9426 some might fail. To disable Partial Content support for all ICAP
9427 services and to avoid any negotiation, set this option to "off".
9433 NAME: icap_default_options_ttl
9436 LOC: Adaptation::Icap::TheConfig.default_options_ttl
9439 The default TTL value for ICAP OPTIONS responses that don't have
9440 an Options-TTL header.
9443 NAME: icap_persistent_connections
9447 LOC: Adaptation::Icap::TheConfig.reuse_connections
9450 Whether or not Squid should use persistent connections to
9454 NAME: adaptation_send_client_ip icap_send_client_ip
9456 IFDEF: USE_ADAPTATION
9458 LOC: Adaptation::Config::send_client_ip
9461 If enabled, Squid shares HTTP client IP information with adaptation
9462 services. For ICAP, Squid adds the X-Client-IP header to ICAP requests.
9463 For eCAP, Squid sets the libecap::metaClientIp transaction option.
9465 See also: adaptation_uses_indirect_client
9468 NAME: adaptation_send_username icap_send_client_username
9470 IFDEF: USE_ADAPTATION
9472 LOC: Adaptation::Config::send_username
9475 This sends authenticated HTTP client username (if available) to
9476 the adaptation service.
9478 For ICAP, the username value is encoded based on the
9479 icap_client_username_encode option and is sent using the header
9480 specified by the icap_client_username_header option.
9483 NAME: icap_client_username_header
9486 LOC: Adaptation::Icap::TheConfig.client_username_header
9487 DEFAULT: X-Client-Username
9489 ICAP request header name to use for adaptation_send_username.
9492 NAME: icap_client_username_encode
9496 LOC: Adaptation::Icap::TheConfig.client_username_encode
9499 Whether to base64 encode the authenticated client username.
9503 TYPE: icap_service_type
9505 LOC: Adaptation::Icap::TheConfig
9508 Defines a single ICAP service using the following format:
9510 icap_service id vectoring_point uri [option ...]
9513 an opaque identifier or name which is used to direct traffic to
9514 this specific service. Must be unique among all adaptation
9515 services in squid.conf.
9517 vectoring_point: reqmod_precache|reqmod_postcache|respmod_precache|respmod_postcache
9518 This specifies at which point of transaction processing the
9519 ICAP service should be activated. *_postcache vectoring points
9520 are not yet supported.
9522 uri: icap://servername:port/servicepath
9523 ICAP server and service location.
9524 icaps://servername:port/servicepath
9525 The "icap:" URI scheme is used for traditional ICAP server and
9526 service location (default port is 1344, connections are not
9527 encrypted). The "icaps:" URI scheme is for Secure ICAP
9528 services that use SSL/TLS-encrypted ICAP connections (by
9529 default, on port 11344).
9531 ICAP does not allow a single service to handle both REQMOD and RESPMOD
9532 transactions. Squid does not enforce that requirement. You can specify
9533 services with the same service_url and different vectoring_points. You
9534 can even specify multiple identical services as long as their
9535 service_names differ.
9537 To activate a service, use the adaptation_access directive. To group
9538 services, use adaptation_service_chain and adaptation_service_set.
9540 Service options are separated by white space. ICAP services support
9541 the following name=value options:
9544 If set to 'on' or '1', the ICAP service is treated as
9545 optional. If the service cannot be reached or malfunctions,
9546 Squid will try to ignore any errors and process the message as
9547 if the service was not enabled. No all ICAP errors can be
9548 bypassed. If set to 0, the ICAP service is treated as
9549 essential and all ICAP errors will result in an error page
9550 returned to the HTTP client.
9552 Bypass is off by default: services are treated as essential.
9555 If set to 'on' or '1', the ICAP service is allowed to
9556 dynamically change the current message adaptation plan by
9557 returning a chain of services to be used next. The services
9558 are specified using the X-Next-Services ICAP response header
9559 value, formatted as a comma-separated list of service names.
9560 Each named service should be configured in squid.conf. Other
9561 services are ignored. An empty X-Next-Services value results
9562 in an empty plan which ends the current adaptation.
9564 Dynamic adaptation plan may cross or cover multiple supported
9565 vectoring points in their natural processing order.
9567 Routing is not allowed by default: the ICAP X-Next-Services
9568 response header is ignored.
9571 Only has effect on split-stack systems. The default on those systems
9572 is to use IPv4-only connections. When set to 'on' this option will
9573 make Squid use IPv6-only connections to contact this ICAP service.
9575 on-overload=block|bypass|wait|force
9576 If the service Max-Connections limit has been reached, do
9577 one of the following for each new ICAP transaction:
9578 * block: send an HTTP error response to the client
9579 * bypass: ignore the "over-connected" ICAP service
9580 * wait: wait (in a FIFO queue) for an ICAP connection slot
9581 * force: proceed, ignoring the Max-Connections limit
9583 In SMP mode with N workers, each worker assumes the service
9584 connection limit is Max-Connections/N, even though not all
9585 workers may use a given service.
9587 The default value is "bypass" if service is bypassable,
9588 otherwise it is set to "wait".
9592 Use the given number as the Max-Connections limit, regardless
9593 of the Max-Connections value given by the service, if any.
9595 connection-encryption=on|off
9596 Determines the ICAP service effect on the connections_encrypted
9599 The default is "on" for Secure ICAP services (i.e., those
9600 with the icaps:// service URIs scheme) and "off" for plain ICAP
9603 Does not affect ICAP connections (e.g., does not turn Secure
9606 ==== ICAPS / TLS OPTIONS ====
9608 These options are used for Secure ICAP (icaps://....) services only.
9610 tls-cert=/path/to/ssl/certificate
9611 A client X.509 certificate to use when connecting to
9614 tls-key=/path/to/ssl/key
9615 The private key corresponding to the previous
9618 If tls-key= is not specified tls-cert= is assumed to
9619 reference a PEM file containing both the certificate
9622 tls-cipher=... The list of valid TLS/SSL ciphers to use when connecting
9623 to this icap server.
9626 The minimum TLS protocol version to permit. To control
9627 SSLv3 use the tls-options= parameter.
9628 Supported Values: 1.0 (default), 1.1, 1.2
9630 tls-options=... Specify various OpenSSL library options:
9632 NO_SSLv3 Disallow the use of SSLv3
9635 Always create a new key when using
9636 temporary/ephemeral DH key exchanges
9638 ALL Enable various bug workarounds
9639 suggested as "harmless" by OpenSSL
9640 Be warned that this reduces SSL/TLS
9641 strength to some attacks.
9643 See the OpenSSL SSL_CTX_set_options documentation for a
9644 more complete list. Options relevant only to SSLv2 are
9647 tls-cafile= PEM file containing CA certificates to use when verifying
9648 the icap server certificate.
9649 Use to specify intermediate CA certificate(s) if not sent
9650 by the server. Or the full CA chain for the server when
9651 using the tls-default-ca=off flag.
9652 May be repeated to load multiple files.
9654 tls-capath=... A directory containing additional CA certificates to
9655 use when verifying the icap server certificate.
9656 Requires OpenSSL or LibreSSL.
9658 tls-crlfile=... A certificate revocation list file to use when
9659 verifying the icap server certificate.
9661 tls-flags=... Specify various flags modifying the Squid TLS implementation:
9664 Accept certificates even if they fail to
9667 Don't verify the icap server certificate
9668 matches the server name
9670 tls-default-ca[=off]
9671 Whether to use the system Trusted CAs. Default is ON.
9673 tls-domain= The icap server name as advertised in it's certificate.
9674 Used for verifying the correctness of the received icap
9675 server certificate. If not specified the icap server
9676 hostname extracted from ICAP URI will be used.
9678 Older icap_service format without optional named parameters is
9679 deprecated but supported for backward compatibility.
9682 icap_service svcBlocker reqmod_precache icap://icap1.mydomain.net:1344/reqmod bypass=0
9683 icap_service svcLogger reqmod_precache icaps://icap2.mydomain.net:11344/reqmod routing=on
9687 TYPE: icap_class_type
9692 This deprecated option was documented to define an ICAP service
9693 chain, even though it actually defined a set of similar, redundant
9694 services, and the chains were not supported.
9696 To define a set of redundant services, please use the
9697 adaptation_service_set directive. For service chains, use
9698 adaptation_service_chain.
9702 TYPE: icap_access_type
9707 This option is deprecated. Please use adaptation_access, which
9708 has the same ICAP functionality, but comes with better
9709 documentation, and eCAP support.
9714 -----------------------------------------------------------------------------
9721 LOC: Adaptation::Ecap::TheConfig.onoff
9724 Controls whether eCAP support is enabled.
9728 TYPE: ecap_service_type
9730 LOC: Adaptation::Ecap::TheConfig
9733 Defines a single eCAP service
9735 ecap_service id vectoring_point uri [option ...]
9738 an opaque identifier or name which is used to direct traffic to
9739 this specific service. Must be unique among all adaptation
9740 services in squid.conf.
9742 vectoring_point: reqmod_precache|reqmod_postcache|respmod_precache|respmod_postcache
9743 This specifies at which point of transaction processing the
9744 eCAP service should be activated. *_postcache vectoring points
9745 are not yet supported.
9747 uri: ecap://vendor/service_name?custom&cgi=style¶meters=optional
9748 Squid uses the eCAP service URI to match this configuration
9749 line with one of the dynamically loaded services. Each loaded
9750 eCAP service must have a unique URI. Obtain the right URI from
9751 the service provider.
9753 To activate a service, use the adaptation_access directive. To group
9754 services, use adaptation_service_chain and adaptation_service_set.
9756 Service options are separated by white space. eCAP services support
9757 the following name=value options:
9760 If set to 'on' or '1', the eCAP service is treated as optional.
9761 If the service cannot be reached or malfunctions, Squid will try
9762 to ignore any errors and process the message as if the service
9763 was not enabled. No all eCAP errors can be bypassed.
9764 If set to 'off' or '0', the eCAP service is treated as essential
9765 and all eCAP errors will result in an error page returned to the
9768 Bypass is off by default: services are treated as essential.
9771 If set to 'on' or '1', the eCAP service is allowed to
9772 dynamically change the current message adaptation plan by
9773 returning a chain of services to be used next.
9775 Dynamic adaptation plan may cross or cover multiple supported
9776 vectoring points in their natural processing order.
9778 Routing is not allowed by default.
9780 connection-encryption=on|off
9781 Determines the eCAP service effect on the connections_encrypted
9784 Defaults to "on", which does not taint the master transaction
9787 Does not affect eCAP API calls.
9789 Older ecap_service format without optional named parameters is
9790 deprecated but supported for backward compatibility.
9794 ecap_service s1 reqmod_precache ecap://filters.R.us/leakDetector?on_error=block bypass=off
9795 ecap_service s2 respmod_precache ecap://filters.R.us/virusFilter config=/etc/vf.cfg bypass=on
9798 NAME: loadable_modules
9800 IFDEF: USE_LOADABLE_MODULES
9801 LOC: Config.loadable_module_names
9804 Instructs Squid to load the specified dynamic module(s) or activate
9805 preloaded module(s).
9807 loadable_modules @DEFAULT_PREFIX@/lib/MinimalAdapter.so
9811 MESSAGE ADAPTATION OPTIONS
9812 -----------------------------------------------------------------------------
9815 NAME: adaptation_service_set
9816 TYPE: adaptation_service_set_type
9817 IFDEF: USE_ADAPTATION
9822 Configures an ordered set of similar, redundant services. This is
9823 useful when hot standby or backup adaptation servers are available.
9825 adaptation_service_set set_name service_name1 service_name2 ...
9827 The named services are used in the set declaration order. The first
9828 applicable adaptation service from the set is used first. The next
9829 applicable service is tried if and only if the transaction with the
9830 previous service fails and the message waiting to be adapted is still
9833 When adaptation starts, broken services are ignored as if they were
9834 not a part of the set. A broken service is a down optional service.
9836 The services in a set must be attached to the same vectoring point
9837 (e.g., pre-cache) and use the same adaptation method (e.g., REQMOD).
9839 If all services in a set are optional then adaptation failures are
9840 bypassable. If all services in the set are essential, then a
9841 transaction failure with one service may still be retried using
9842 another service from the set, but when all services fail, the master
9843 transaction fails as well.
9845 A set may contain a mix of optional and essential services, but that
9846 is likely to lead to surprising results because broken services become
9847 ignored (see above), making previously bypassable failures fatal.
9848 Technically, it is the bypassability of the last failed service that
9851 See also: adaptation_access adaptation_service_chain
9854 adaptation_service_set svcBlocker urlFilterPrimary urlFilterBackup
9855 adaptation service_set svcLogger loggerLocal loggerRemote
9858 NAME: adaptation_service_chain
9859 TYPE: adaptation_service_chain_type
9860 IFDEF: USE_ADAPTATION
9865 Configures a list of complementary services that will be applied
9866 one-by-one, forming an adaptation chain or pipeline. This is useful
9867 when Squid must perform different adaptations on the same message.
9869 adaptation_service_chain chain_name service_name1 svc_name2 ...
9871 The named services are used in the chain declaration order. The first
9872 applicable adaptation service from the chain is used first. The next
9873 applicable service is applied to the successful adaptation results of
9874 the previous service in the chain.
9876 When adaptation starts, broken services are ignored as if they were
9877 not a part of the chain. A broken service is a down optional service.
9879 Request satisfaction terminates the adaptation chain because Squid
9880 does not currently allow declaration of RESPMOD services at the
9881 "reqmod_precache" vectoring point (see icap_service or ecap_service).
9883 The services in a chain must be attached to the same vectoring point
9884 (e.g., pre-cache) and use the same adaptation method (e.g., REQMOD).
9886 A chain may contain a mix of optional and essential services. If an
9887 essential adaptation fails (or the failure cannot be bypassed for
9888 other reasons), the master transaction fails. Otherwise, the failure
9889 is bypassed as if the failed adaptation service was not in the chain.
9891 See also: adaptation_access adaptation_service_set
9894 adaptation_service_chain svcRequest requestLogger urlFilter leakDetector
9897 NAME: adaptation_access
9898 TYPE: adaptation_access_type
9899 IFDEF: USE_ADAPTATION
9902 DEFAULT_DOC: Allow, unless rules exist in squid.conf.
9904 Sends an HTTP transaction to an ICAP or eCAP adaptation service.
9906 adaptation_access service_name allow|deny [!]aclname...
9907 adaptation_access set_name allow|deny [!]aclname...
9909 At each supported vectoring point, the adaptation_access
9910 statements are processed in the order they appear in this
9911 configuration file. Statements pointing to the following services
9912 are ignored (i.e., skipped without checking their ACL):
9914 - services serving different vectoring points
9915 - "broken-but-bypassable" services
9916 - "up" services configured to ignore such transactions
9917 (e.g., based on the ICAP Transfer-Ignore header).
9919 When a set_name is used, all services in the set are checked
9920 using the same rules, to find the first applicable one. See
9921 adaptation_service_set for details.
9923 If an access list is checked and there is a match, the
9924 processing stops: For an "allow" rule, the corresponding
9925 adaptation service is used for the transaction. For a "deny"
9926 rule, no adaptation service is activated.
9928 It is currently not possible to apply more than one adaptation
9929 service at the same vectoring point to the same HTTP transaction.
9931 See also: icap_service and ecap_service
9934 adaptation_access service_1 allow all
9937 NAME: adaptation_service_iteration_limit
9939 IFDEF: USE_ADAPTATION
9940 LOC: Adaptation::Config::service_iteration_limit
9943 Limits the number of iterations allowed when applying adaptation
9944 services to a message. If your longest adaptation set or chain
9945 may have more than 16 services, increase the limit beyond its
9946 default value of 16. If detecting infinite iteration loops sooner
9947 is critical, make the iteration limit match the actual number
9948 of services in your longest adaptation set or chain.
9950 Infinite adaptation loops are most likely with routing services.
9952 See also: icap_service routing=1
9955 NAME: adaptation_masterx_shared_names
9957 IFDEF: USE_ADAPTATION
9958 LOC: Adaptation::Config::masterx_shared_name
9961 For each master transaction (i.e., the HTTP request and response
9962 sequence, including all related ICAP and eCAP exchanges), Squid
9963 maintains a table of metadata. The table entries are (name, value)
9964 pairs shared among eCAP and ICAP exchanges. The table is destroyed
9965 with the master transaction.
9967 This option specifies the table entry names that Squid must accept
9968 from and forward to the adaptation transactions.
9970 An ICAP REQMOD or RESPMOD transaction may set an entry in the
9971 shared table by returning an ICAP header field with a name
9972 specified in adaptation_masterx_shared_names.
9974 An eCAP REQMOD or RESPMOD transaction may set an entry in the
9975 shared table by implementing the libecap::visitEachOption() API
9976 to provide an option with a name specified in
9977 adaptation_masterx_shared_names.
9979 Squid will store and forward the set entry to subsequent adaptation
9980 transactions within the same master transaction scope.
9982 Only one shared entry name is supported at this time.
9985 # share authentication information among ICAP services
9986 adaptation_masterx_shared_names X-Subscriber-ID
9989 NAME: adaptation_meta
9991 IFDEF: USE_ADAPTATION
9992 LOC: Adaptation::Config::metaHeaders()
9995 This option allows Squid administrator to add custom ICAP request
9996 headers or eCAP options to Squid ICAP requests or eCAP transactions.
9997 Use it to pass custom authentication tokens and other
9998 transaction-state related meta information to an ICAP/eCAP service.
10000 The addition of a meta header is ACL-driven:
10001 adaptation_meta name value [!]aclname ...
10003 Processing for a given header name stops after the first ACL list match.
10004 Thus, it is impossible to add two headers with the same name. If no ACL
10005 lists match for a given header name, no such header is added. For
10008 # do not debug transactions except for those that need debugging
10009 adaptation_meta X-Debug 1 needs_debugging
10011 # log all transactions except for those that must remain secret
10012 adaptation_meta X-Log 1 !keep_secret
10014 # mark transactions from users in the "G 1" group
10015 adaptation_meta X-Authenticated-Groups "G 1" authed_as_G1
10017 The "value" parameter may be a regular squid.conf token or a "double
10018 quoted string". Within the quoted string, use backslash (\) to escape
10019 any character, which is currently only useful for escaping backslashes
10020 and double quotes. For example,
10021 "this string has one backslash (\\) and two \"quotes\""
10023 Used adaptation_meta header values may be logged via %note
10024 logformat code. If multiple adaptation_meta headers with the same name
10025 are used during master transaction lifetime, the header values are
10026 logged in the order they were used and duplicate values are ignored
10027 (only the first repeated value will be logged).
10033 LOC: Adaptation::Icap::TheConfig.repeat
10034 DEFAULT_IF_NONE: deny all
10036 This ACL determines which retriable ICAP transactions are
10037 retried. Transactions that received a complete ICAP response
10038 and did not have to consume or produce HTTP bodies to receive
10039 that response are usually retriable.
10041 icap_retry allow|deny [!]aclname ...
10043 Squid automatically retries some ICAP I/O timeouts and errors
10044 due to persistent connection race conditions.
10046 See also: icap_retry_limit
10049 NAME: icap_retry_limit
10052 LOC: Adaptation::Icap::TheConfig.repeat_limit
10054 DEFAULT_DOC: No retries are allowed.
10056 Limits the number of retries allowed.
10058 Communication errors due to persistent connection race
10059 conditions are unavoidable, automatically retried, and do not
10060 count against this limit.
10062 See also: icap_retry
10068 -----------------------------------------------------------------------------
10071 NAME: check_hostnames
10074 LOC: Config.onoff.check_hostnames
10076 For security and stability reasons Squid can check
10077 hostnames for Internet standard RFC compliance. If you want
10078 Squid to perform these checks turn this directive on.
10081 NAME: allow_underscore
10084 LOC: Config.onoff.allow_underscore
10086 Underscore characters is not strictly allowed in Internet hostnames
10087 but nevertheless used by many sites. Set this to off if you want
10088 Squid to be strict about the standard.
10089 This check is performed only when check_hostnames is set to on.
10092 NAME: dns_retransmit_interval
10095 LOC: Config.Timeout.idns_retransmit
10097 Initial retransmit interval for DNS queries. The interval is
10098 doubled each time all configured DNS servers have been tried.
10103 DEFAULT: 30 seconds
10104 LOC: Config.Timeout.idns_query
10106 DNS Query timeout. If no response is received to a DNS query
10107 within this time all DNS servers for the queried domain
10108 are assumed to be unavailable.
10111 NAME: dns_packet_max
10113 DEFAULT_DOC: EDNS disabled
10115 LOC: Config.dns.packet_max
10117 Maximum number of bytes packet size to advertise via EDNS.
10118 Set to "none" to disable EDNS large packet support.
10120 For legacy reasons DNS UDP replies will default to 512 bytes which
10121 is too small for many responses. EDNS provides a means for Squid to
10122 negotiate receiving larger responses back immediately without having
10123 to failover with repeat requests. Responses larger than this limit
10124 will retain the old behaviour of failover to TCP DNS.
10126 Squid has no real fixed limit internally, but allowing packet sizes
10127 over 1500 bytes requires network jumbogram support and is usually not
10130 WARNING: The RFC also indicates that some older resolvers will reply
10131 with failure of the whole request if the extension is added. Some
10132 resolvers have already been identified which will reply with mangled
10133 EDNS response on occasion. Usually in response to many-KB jumbogram
10134 sizes being advertised by Squid.
10135 Squid will currently treat these both as an unable-to-resolve domain
10136 even if it would be resolvable without EDNS.
10143 DEFAULT_DOC: Search for single-label domain names is disabled.
10144 LOC: Config.onoff.res_defnames
10146 Normally the RES_DEFNAMES resolver option is disabled
10147 (see res_init(3)). This prevents caches in a hierarchy
10148 from interpreting single-component hostnames locally. To allow
10149 Squid to handle single-component names, enable this option.
10152 NAME: dns_multicast_local
10156 DEFAULT_DOC: Search for .local and .arpa names is disabled.
10157 LOC: Config.onoff.dns_mdns
10159 When set to on, Squid sends multicast DNS lookups on the local
10160 network for domains ending in .local and .arpa.
10161 This enables local servers and devices to be contacted in an
10162 ad-hoc or zero-configuration network environment.
10165 NAME: dns_nameservers
10168 DEFAULT_DOC: Use operating system definitions
10169 LOC: Config.dns.nameservers
10171 Use this if you want to specify a list of DNS name servers
10172 (IP addresses) to use instead of those given in your
10173 /etc/resolv.conf file.
10175 On Windows platforms, if no value is specified here or in
10176 the /etc/resolv.conf file, the list of DNS name servers are
10177 taken from the Windows registry, both static and dynamic DHCP
10178 configurations are supported.
10180 Example: dns_nameservers 10.0.0.1 192.172.0.4
10185 DEFAULT: @DEFAULT_HOSTS@
10186 LOC: Config.etcHostsPath
10188 Location of the host-local IP name-address associations
10189 database. Most Operating Systems have such a file on different
10191 - Un*X & Linux: /etc/hosts
10192 - Windows NT/2000: %SystemRoot%\system32\drivers\etc\hosts
10193 (%SystemRoot% value install default is c:\winnt)
10194 - Windows XP/2003: %SystemRoot%\system32\drivers\etc\hosts
10195 (%SystemRoot% value install default is c:\windows)
10196 - Windows 9x/Me: %windir%\hosts
10197 (%windir% value is usually c:\windows)
10198 - Cygwin: /etc/hosts
10200 The file contains newline-separated definitions, in the
10201 form ip_address_in_dotted_form name [name ...] names are
10202 whitespace-separated. Lines beginning with an hash (#)
10203 character are comments.
10205 The file is checked at startup and upon configuration.
10206 If set to 'none', it won't be checked.
10207 If append_domain is used, that domain will be added to
10208 domain-local (i.e. not containing any dot character) host
10212 NAME: append_domain
10214 LOC: Config.appendDomain
10216 DEFAULT_DOC: Use operating system definitions
10218 Appends local domain name to hostnames without any dots in
10219 them. append_domain must begin with a period.
10221 Be warned there are now Internet names with no dots in
10222 them using only top-domain names, so setting this may
10223 cause some Internet sites to become unavailable.
10226 append_domain .yourdomain.com
10229 NAME: ignore_unknown_nameservers
10231 LOC: Config.onoff.ignore_unknown_nameservers
10234 By default Squid checks that DNS responses are received
10235 from the same IP addresses they are sent to. If they
10236 don't match, Squid ignores the response and writes a warning
10237 message to cache.log. You can allow responses from unknown
10238 nameservers by setting this option to 'off'.
10242 COMMENT: (number of entries)
10245 LOC: Config.ipcache.size
10247 Maximum number of DNS IP cache entries.
10254 LOC: Config.ipcache.low
10261 LOC: Config.ipcache.high
10263 The size, low-, and high-water marks for the IP cache.
10266 NAME: fqdncache_size
10267 COMMENT: (number of entries)
10270 LOC: Config.fqdncache.size
10272 Maximum number of FQDN cache entries.
10277 -----------------------------------------------------------------------------
10280 NAME: configuration_includes_quoted_values
10282 TYPE: configuration_includes_quoted_values
10284 LOC: ConfigParser::RecognizeQuotedValues
10286 If set, Squid will recognize each "quoted string" after a configuration
10287 directive as a single parameter. The quotes are stripped before the
10288 parameter value is interpreted or used.
10289 See "Values with spaces, quotes, and other special characters"
10290 section for more details.
10297 LOC: Config.onoff.mem_pools
10299 If set, Squid will keep pools of allocated (but unused) memory
10300 available for future use. If memory is a premium on your
10301 system and you believe your malloc library outperforms Squid
10302 routines, disable this.
10305 NAME: memory_pools_limit
10309 LOC: Config.MemPools.limit
10311 Used only with memory_pools on:
10312 memory_pools_limit 50 MB
10314 If set to a non-zero value, Squid will keep at most the specified
10315 limit of allocated (but unused) memory in memory pools. All free()
10316 requests that exceed this limit will be handled by your malloc
10317 library. Squid does not pre-allocate any memory, just safe-keeps
10318 objects that otherwise would be free()d. Thus, it is safe to set
10319 memory_pools_limit to a reasonably high value even if your
10320 configuration will use less memory.
10322 If set to none, Squid will keep all memory it can. That is, there
10323 will be no limit on the total amount of memory used for safe-keeping.
10325 To disable memory allocation optimization, do not set
10326 memory_pools_limit to 0 or none. Set memory_pools to "off" instead.
10328 An overhead for maintaining memory pools is not taken into account
10329 when the limit is checked. This overhead is close to four bytes per
10330 object kept. However, pools may actually _save_ memory because of
10331 reduced memory thrashing in your malloc library.
10334 NAME: forwarded_for
10335 COMMENT: on|off|transparent|truncate|delete
10338 LOC: opt_forwarded_for
10340 If set to "on", Squid will append your client's IP address
10341 in the HTTP requests it forwards. By default it looks like:
10343 X-Forwarded-For: 192.1.2.3
10345 If set to "off", it will appear as
10347 X-Forwarded-For: unknown
10349 If set to "transparent", Squid will not alter the
10350 X-Forwarded-For header in any way.
10352 If set to "delete", Squid will delete the entire
10353 X-Forwarded-For header.
10355 If set to "truncate", Squid will remove all existing
10356 X-Forwarded-For entries, and place the client IP as the sole entry.
10359 NAME: cachemgr_passwd
10360 TYPE: cachemgrpasswd
10362 DEFAULT_DOC: No password. Actions which require password are denied.
10363 LOC: Config.passwd_list
10365 Specify passwords for cachemgr operations.
10367 Usage: cachemgr_passwd password action action ...
10369 Some valid actions are (see cache manager menu for a full list):
10407 * Indicates actions which will not be performed without a
10408 valid password, others can be performed if not listed here.
10410 To disable an action, set the password to "disable".
10411 To allow performing an action without a password, set the
10412 password to "none".
10414 Use the keyword "all" to set the same password for all actions.
10417 cachemgr_passwd secret shutdown
10418 cachemgr_passwd lesssssssecret info stats/objects
10419 cachemgr_passwd disable all
10426 LOC: Config.onoff.client_db
10428 If you want to disable collecting per-client statistics,
10429 turn off client_db here.
10432 NAME: refresh_all_ims
10436 LOC: Config.onoff.refresh_all_ims
10438 When you enable this option, squid will always check
10439 the origin server for an update when a client sends an
10440 If-Modified-Since request. Many browsers use IMS
10441 requests when the user requests a reload, and this
10442 ensures those clients receive the latest version.
10444 By default (off), squid may return a Not Modified response
10445 based on the age of the cached version.
10448 NAME: reload_into_ims
10449 IFDEF: USE_HTTP_VIOLATIONS
10453 LOC: Config.onoff.reload_into_ims
10455 When you enable this option, client no-cache or ``reload''
10456 requests will be changed to If-Modified-Since requests.
10457 Doing this VIOLATES the HTTP standard. Enabling this
10458 feature could make you liable for problems which it
10461 see also refresh_pattern for a more selective approach.
10464 NAME: connect_retries
10466 LOC: Config.connect_retries
10468 DEFAULT_DOC: Do not retry failed connections.
10470 Limits the number of reopening attempts when establishing a single
10471 TCP connection. All these attempts must still complete before the
10472 applicable connection opening timeout expires.
10474 By default and when connect_retries is set to zero, Squid does not
10475 retry failed connection opening attempts.
10477 The (not recommended) maximum is 10 tries. An attempt to configure a
10478 higher value results in the value of 10 being used (with a warning).
10480 Squid may open connections to retry various high-level forwarding
10481 failures. For an outside observer, that activity may look like a
10482 low-level connection reopening attempt, but those high-level retries
10483 are governed by forward_max_tries instead.
10485 See also: connect_timeout, forward_timeout, icap_connect_timeout,
10486 and forward_max_tries.
10489 NAME: retry_on_error
10491 LOC: Config.retry.onerror
10494 If set to ON Squid will automatically retry requests when
10495 receiving an error response with status 403 (Forbidden),
10496 500 (Internal Error), 501 or 503 (Service not available).
10497 Status 502 and 504 (Gateway errors) are always retried.
10499 This is mainly useful if you are in a complex cache hierarchy to
10500 work around access control errors.
10502 NOTE: This retry will attempt to find another working destination.
10503 Which is different from the server which just failed.
10506 NAME: as_whois_server
10508 LOC: Config.as_whois_server
10509 DEFAULT: whois.ra.net
10511 WHOIS server to query for AS numbers. NOTE: AS numbers are
10512 queried only when Squid starts up, not for every request.
10517 LOC: Config.onoff.offline
10520 Enable this option and Squid will never try to validate cached
10524 NAME: uri_whitespace
10525 TYPE: uri_whitespace
10526 LOC: Config.uri_whitespace
10529 What to do with requests that have whitespace characters in the
10532 strip: The whitespace characters are stripped out of the URL.
10533 This is the behavior recommended by RFC2396 and RFC3986
10534 for tolerant handling of generic URI.
10535 NOTE: This is one difference between generic URI and HTTP URLs.
10537 deny: The request is denied. The user receives an "Invalid
10539 This is the behaviour recommended by RFC2616 for safe
10540 handling of HTTP request URL.
10542 allow: The request is allowed and the URI is not changed. The
10543 whitespace characters remain in the URI. Note the
10544 whitespace is passed to redirector processes if they
10546 Note this may be considered a violation of RFC2616
10547 request parsing where whitespace is prohibited in the
10550 encode: The request is allowed and the whitespace characters are
10551 encoded according to RFC1738.
10553 chop: The request is allowed and the URI is chopped at the
10557 NOTE the current Squid implementation of encode and chop violates
10558 RFC2616 by not using a 301 redirect after altering the URL.
10563 LOC: Config.chroot_dir
10566 Specifies a directory where Squid should do a chroot() while
10567 initializing. This also causes Squid to fully drop root
10568 privileges after initializing. This means, for example, if you
10569 use a HTTP port less than 1024 and try to reconfigure, you may
10570 get an error saying that Squid can not open the port.
10573 NAME: pipeline_prefetch
10574 TYPE: pipelinePrefetch
10575 LOC: Config.pipeline_max_prefetch
10577 DEFAULT_DOC: Do not pre-parse pipelined requests.
10579 HTTP clients may send a pipeline of 1+N requests to Squid using a
10580 single connection, without waiting for Squid to respond to the first
10581 of those requests. This option limits the number of concurrent
10582 requests Squid will try to handle in parallel. If set to N, Squid
10583 will try to receive and process up to 1+N requests on the same
10584 connection concurrently.
10586 Defaults to 0 (off) for bandwidth management and access logging
10589 NOTE: pipelining requires persistent connections to clients.
10591 WARNING: pipelining breaks NTLM and Negotiate/Kerberos authentication.
10594 NAME: high_response_time_warning
10597 LOC: Config.warnings.high_rptm
10599 DEFAULT_DOC: disabled.
10601 If the one-minute median response time exceeds this value,
10602 Squid prints a WARNING with debug level 0 to get the
10603 administrators attention. The value is in milliseconds.
10606 NAME: high_page_fault_warning
10608 LOC: Config.warnings.high_pf
10610 DEFAULT_DOC: disabled.
10612 If the one-minute average page fault rate exceeds this
10613 value, Squid prints a WARNING with debug level 0 to get
10614 the administrators attention. The value is in page faults
10618 NAME: high_memory_warning
10620 LOC: Config.warnings.high_memory
10621 IFDEF: HAVE_MSTATS&&HAVE_GNUMALLOC_H
10623 DEFAULT_DOC: disabled.
10625 If the memory usage (as determined by gnumalloc, if available and used)
10626 exceeds this amount, Squid prints a WARNING with debug level 0 to get
10627 the administrators attention.
10629 # TODO: link high_memory_warning to mempools?
10631 NAME: sleep_after_fork
10632 COMMENT: (microseconds)
10634 LOC: Config.sleep_after_fork
10637 When this is set to a non-zero value, the main Squid process
10638 sleeps the specified number of microseconds after a fork()
10639 system call. This sleep may help the situation where your
10640 system reports fork() failures due to lack of (virtual)
10641 memory. Note, however, if you have a lot of child
10642 processes, these sleep delays will add up and your
10643 Squid will not service requests for some amount of time
10644 until all the child processes have been started.
10647 NAME: windows_ipaddrchangemonitor
10648 IFDEF: _SQUID_WINDOWS_
10652 LOC: Config.onoff.WIN32_IpAddrChangeMonitor
10654 On Windows Squid by default will monitor IP address changes and will
10655 reconfigure itself after any detected event. This is very useful for
10656 proxies connected to internet with dial-up interfaces.
10657 In some cases (a Proxy server acting as VPN gateway is one) it could be
10658 desiderable to disable this behaviour setting this to 'off'.
10659 Note: after changing this, Squid service must be restarted.
10664 IFDEF: USE_SQUID_EUI
10666 LOC: Eui::TheConfig.euiLookup
10668 Whether to lookup the EUI or MAC address of a connected client.
10671 NAME: max_filedescriptors max_filedesc
10674 DEFAULT_DOC: Use operating system soft limit set by ulimit.
10675 LOC: Config.max_filedescriptors
10677 Set the maximum number of filedescriptors, either below the
10678 operating system default or up to the hard limit.
10680 Remove from squid.conf to inherit the current ulimit soft
10683 Note: Changing this requires a restart of Squid. Also
10684 not all I/O types supports large values (eg on Windows).
10687 NAME: force_request_body_continuation
10689 LOC: Config.accessList.forceRequestBodyContinuation
10691 DEFAULT_DOC: Deny, unless rules exist in squid.conf.
10693 This option controls how Squid handles data upload requests from HTTP
10694 and FTP agents that require a "Please Continue" control message response
10695 to actually send the request body to Squid. It is mostly useful in
10696 adaptation environments.
10698 When Squid receives an HTTP request with an "Expect: 100-continue"
10699 header or an FTP upload command (e.g., STOR), Squid normally sends the
10700 request headers or FTP command information to an adaptation service (or
10701 peer) and waits for a response. Most adaptation services (and some
10702 broken peers) may not respond to Squid at that stage because they may
10703 decide to wait for the HTTP request body or FTP data transfer. However,
10704 that request body or data transfer may never come because Squid has not
10705 responded with the HTTP 100 or FTP 150 (Please Continue) control message
10706 to the request sender yet!
10708 An allow match tells Squid to respond with the HTTP 100 or FTP 150
10709 (Please Continue) control message on its own, before forwarding the
10710 request to an adaptation service or peer. Such a response usually forces
10711 the request sender to proceed with sending the body. A deny match tells
10712 Squid to delay that control response until the origin server confirms
10713 that the request body is needed. Delaying is the default behavior.
10716 NAME: http_upgrade_request_protocols
10717 TYPE: http_upgrade_request_protocols
10718 LOC: Config.http_upgrade_request_protocols
10720 DEFAULT_DOC: Upgrade header dropped, effectively blocking an upgrade attempt.
10722 Controls client-initiated and server-confirmed switching from HTTP to
10723 another protocol (or to several protocols) using HTTP Upgrade mechanism
10724 defined in RFC 7230 Section 6.7. Squid itself does not understand the
10725 protocols being upgraded to and participates in the upgraded
10726 communication only as a dumb TCP proxy. Admins should not allow
10727 upgrading to protocols that require a more meaningful proxy
10730 Usage: http_upgrade_request_protocols <protocol> allow|deny [!]acl ...
10732 The required "protocol" parameter is either an all-caps word OTHER or an
10733 explicit protocol name (e.g. "WebSocket") optionally followed by a slash
10734 and a version token (e.g. "HTTP/3"). Explicit protocol names and
10735 versions are case sensitive.
10737 When an HTTP client sends an Upgrade request header, Squid iterates over
10738 the client-offered protocols and, for each protocol P (with an optional
10739 version V), evaluates the first non-empty set of
10740 http_upgrade_request_protocols rules (if any) from the following list:
10742 * All rules with an explicit protocol name equal to P.
10743 * All rules that use OTHER instead of a protocol name.
10745 In other words, rules using OTHER are considered for protocol P if and
10746 only if there are no rules mentioning P by name.
10748 If both of the above sets are empty, then Squid removes protocol P from
10751 If the client sent a versioned protocol offer P/X, then explicit rules
10752 referring to the same-name but different-version protocol P/Y are
10753 declared inapplicable. Inapplicable rules are not evaluated (i.e. are
10754 ignored). However, inapplicable rules still belong to the first set of
10757 Within the applicable rule subset, individual rules are evaluated in
10758 their configuration order. If all ACLs of an applicable "allow" rule
10759 match, then the protocol offered by the client is forwarded to the next
10760 hop as is. If all ACLs of an applicable "deny" rule match, then the
10761 offer is dropped. If no applicable rules have matching ACLs, then the
10762 offer is also dropped. The first matching rule also ends rules
10763 evaluation for the offered protocol.
10765 If all client-offered protocols are removed, then Squid forwards the
10766 client request without the Upgrade header. Squid never sends an empty
10767 Upgrade request header.
10769 An Upgrade request header with a value violating HTTP syntax is dropped
10770 and ignored without an attempt to use extractable individual protocol
10773 Upon receiving an HTTP 101 (Switching Protocols) control message, Squid
10774 checks that the server listed at least one protocol name and sent a
10775 Connection:upgrade response header. Squid does not understand individual
10776 protocol naming and versioning concepts enough to implement stricter
10777 checks, but an admin can restrict HTTP 101 (Switching Protocols)
10778 responses further using http_reply_access. Responses denied by
10779 http_reply_access rules and responses flagged by the internal Upgrade
10780 checks result in HTTP 502 (Bad Gateway) ERR_INVALID_RESP errors and
10781 Squid-to-server connection closures.
10783 If Squid sends an Upgrade request header, and the next hop (e.g., the
10784 origin server) responds with an acceptable HTTP 101 (Switching
10785 Protocols), then Squid forwards that message to the client and becomes
10788 The presence of an Upgrade request header alone does not preclude cache
10789 lookups. In other words, an Upgrade request might be satisfied from the
10790 cache, using regular HTTP caching rules.
10792 This clause only supports fast acl types.
10793 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
10795 Each of the following groups of configuration lines represents a
10796 separate configuration example:
10798 # never upgrade to protocol Foo; all others are OK
10799 http_upgrade_request_protocols Foo deny all
10800 http_upgrade_request_protocols OTHER allow all
10802 # only allow upgrades to protocol Bar (except for its first version)
10803 http_upgrade_request_protocols Bar/1 deny all
10804 http_upgrade_request_protocols Bar allow all
10805 http_upgrade_request_protocols OTHER deny all # this rule is optional
10807 # only allow upgrades to protocol Baz, and only if Baz is the only offer
10808 acl UpgradeHeaderHasMultipleOffers ...
10809 http_upgrade_request_protocols Baz deny UpgradeHeaderHasMultipleOffers
10810 http_upgrade_request_protocols Baz allow all
10813 NAME: server_pconn_for_nonretriable
10816 DEFAULT_DOC: Open new connections for forwarding requests Squid cannot retry safely.
10817 LOC: Config.accessList.serverPconnForNonretriable
10819 This option provides fine-grained control over persistent connection
10820 reuse when forwarding HTTP requests that Squid cannot retry. It is useful
10821 in environments where opening new connections is very expensive
10822 (e.g., all connections are secured with TLS with complex client and server
10823 certificate validation) and race conditions associated with persistent
10824 connections are very rare and/or only cause minor problems.
10826 HTTP prohibits retrying unsafe and non-idempotent requests (e.g., POST).
10827 Squid limitations also prohibit retrying all requests with bodies (e.g., PUT).
10828 By default, when forwarding such "risky" requests, Squid opens a new
10829 connection to the server or cache_peer, even if there is an idle persistent
10830 connection available. When Squid is configured to risk sending a non-retriable
10831 request on a previously used persistent connection, and the server closes
10832 the connection before seeing that risky request, the user gets an error response
10833 from Squid. In most cases, that error response will be HTTP 502 (Bad Gateway)
10834 with ERR_ZERO_SIZE_OBJECT or ERR_WRITE_ERROR (peer connection reset) error detail.
10836 If an allow rule matches, Squid reuses an available idle persistent connection
10837 (if any) for the request that Squid cannot retry. If a deny rule matches, then
10838 Squid opens a new connection for the request that Squid cannot retry.
10840 This option does not affect requests that Squid can retry. They will reuse idle
10841 persistent connections (if any).
10843 This clause only supports fast acl types.
10844 See https://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
10847 acl SpeedIsWorthTheRisk method POST
10848 server_pconn_for_nonretriable allow SpeedIsWorthTheRisk
10851 NAME: happy_eyeballs_connect_timeout
10855 LOC: Config.happyEyeballs.connect_timeout
10857 This Happy Eyeballs (RFC 8305) tuning directive specifies the minimum
10858 delay between opening a primary to-server connection and opening a
10859 spare to-server connection for the same master transaction. This delay
10860 is similar to the Connection Attempt Delay in RFC 8305, but it is only
10861 applied to the first spare connection attempt. Subsequent spare
10862 connection attempts use happy_eyeballs_connect_gap, and primary
10863 connection attempts are not artificially delayed at all.
10865 Terminology: The "primary" and "spare" designations are determined by
10866 the order of DNS answers received by Squid: If Squid DNS AAAA query
10867 was answered first, then primary connections are connections to IPv6
10868 peer addresses (while spare connections use IPv4 addresses).
10869 Similarly, if Squid DNS A query was answered first, then primary
10870 connections are connections to IPv4 peer addresses (while spare
10871 connections use IPv6 addresses).
10873 Shorter happy_eyeballs_connect_timeout values reduce master
10874 transaction response time, potentially improving user-perceived
10875 response times (i.e., making user eyeballs happier). Longer delays
10876 reduce both concurrent connection level and server bombardment with
10877 connection requests, potentially improving overall Squid performance
10878 and reducing the chance of being blocked by servers for opening too
10879 many unused connections.
10881 RFC 8305 prohibits happy_eyeballs_connect_timeout values smaller than
10882 10 (milliseconds) to "avoid congestion collapse in the presence of
10883 high packet-loss rates".
10885 The following Happy Eyeballs directives place additional connection
10886 opening restrictions: happy_eyeballs_connect_gap and
10887 happy_eyeballs_connect_limit.
10890 NAME: happy_eyeballs_connect_gap
10894 DEFAULT_DOC: no artificial delays between spare attempts
10895 LOC: Config.happyEyeballs.connect_gap
10897 This Happy Eyeballs (RFC 8305) tuning directive specifies the
10898 minimum delay between opening spare to-server connections (to any
10899 server; i.e. across all concurrent master transactions in a Squid
10900 instance). Each SMP worker currently multiplies the configured gap
10901 by the total number of workers so that the combined spare connection
10902 opening rate of a Squid instance obeys the configured limit. The
10903 workers do not coordinate connection openings yet; a micro burst
10904 of spare connection openings may violate the configured gap.
10906 This directive has similar trade-offs as
10907 happy_eyeballs_connect_timeout, but its focus is on limiting traffic
10908 amplification effects for Squid as a whole, while
10909 happy_eyeballs_connect_timeout works on an individual master
10912 The following Happy Eyeballs directives place additional connection
10913 opening restrictions: happy_eyeballs_connect_timeout and
10914 happy_eyeballs_connect_limit. See the former for related terminology.
10917 NAME: happy_eyeballs_connect_limit
10920 DEFAULT_DOC: no artificial limit on the number of concurrent spare attempts
10921 LOC: Config.happyEyeballs.connect_limit
10923 This Happy Eyeballs (RFC 8305) tuning directive specifies the
10924 maximum number of spare to-server connections (to any server; i.e.
10925 across all concurrent master transactions in a Squid instance).
10926 Each SMP worker gets an equal share of the total limit. However,
10927 the workers do not share the actual connection counts yet, so one
10928 (busier) worker cannot "borrow" spare connection slots from another
10929 (less loaded) worker.
10931 Setting this limit to zero disables concurrent use of primary and
10932 spare TCP connections: Spare connection attempts are made only after
10933 all primary attempts fail. However, Squid would still use the
10934 DNS-related optimizations of the Happy Eyeballs approach.
10936 This directive has similar trade-offs as happy_eyeballs_connect_gap,
10937 but its focus is on limiting Squid overheads, while
10938 happy_eyeballs_connect_gap focuses on the origin server and peer
10941 The following Happy Eyeballs directives place additional connection
10942 opening restrictions: happy_eyeballs_connect_timeout and
10943 happy_eyeballs_connect_gap. See the former for related terminology.
projects / thirdparty / squid.git / blob