1 ## Copyright (C) 1996-2015 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 http://wiki.squid-cache.org/SquidFaq
20 http://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 (1024 bytes)
58 Values with spaces, quotes, and other special characters
60 Squid supports directive parameters with spaces, quotes, and other
61 special characters. Surround such parameters with "double quotes". Use
62 the configuration_includes_quoted_values directive to enable or
65 Squid supports reading configuration option parameters from external
66 files using the syntax:
67 parameters("/path/filename")
69 acl whitelist dstdomain parameters("/etc/squid/whitelist.txt")
71 Conditional configuration
73 If-statements can be used to make configuration directives
77 ... regular configuration directives ...
79 ... regular configuration directives ...]
82 The else part is optional. The keywords "if", "else", and "endif"
83 must be typed on their own lines, as if they were regular
84 configuration directives.
86 NOTE: An else-if condition is not supported.
88 These individual conditions types are supported:
91 Always evaluates to true.
93 Always evaluates to false.
95 Equality comparison of two integer numbers.
100 The following SMP-related preprocessor macros can be used.
102 ${process_name} expands to the current Squid process "name"
103 (e.g., squid1, squid2, or cache1).
105 ${process_number} expands to the current Squid process
106 identifier, which is an integer number (e.g., 1, 2, 3) unique
107 across all Squid processes of the current service instance.
109 ${service_name} expands into the current Squid service instance
110 name identifier which is provided by -n on the command line.
114 # options still not yet ported from 2.7 to 3.x
115 NAME: broken_vary_encoding
118 This option is not yet supported by Squid-3.
124 This option is not yet supported by Squid-3.
130 This option is not yet supported by Squid-3.
133 NAME: external_refresh_check
136 This option is not yet supported by Squid-3.
139 NAME: location_rewrite_program location_rewrite_access location_rewrite_children location_rewrite_concurrency
142 This option is not yet supported by Squid-3.
145 NAME: refresh_stale_hit
148 This option is not yet supported by Squid-3.
151 # Options removed in 3.6
152 NAME: cache_peer_domain cache_host_domain
155 Replace with dstdomain ACLs and cache_peer_access.
158 NAME: sslproxy_cafile
161 Remove this line. Use tls_outgoing_options cafile= instead.
164 NAME: sslproxy_capath
167 Remove this line. Use tls_outgoing_options capath= instead.
170 NAME: sslproxy_cipher
173 Remove this line. Use tls_outgoing_options cipher= instead.
176 NAME: sslproxy_client_certificate
179 Remove this line. Use tls_outgoing_options cert= instead.
182 NAME: sslproxy_client_key
185 Remove this line. Use tls_outgoing_options key= instead.
191 Remove this line. Use tls_outgoing_options flags= instead.
194 NAME: sslproxy_options
197 Remove this line. Use tls_outgoing_options options= instead.
200 NAME: sslproxy_version
203 Remove this line. Use tls_outgoing_options options= instead.
206 # Options removed in 3.5
207 NAME: hierarchy_stoplist
210 Remove this line. Use always_direct or cache_peer_access ACLs instead if you need to prevent cache_peer use.
213 # Options removed in 3.4
217 Remove this line. Use acls with access_log directives to control access logging
223 Remove this line. Use acls with icap_log directives to control icap logging
226 # Options Removed in 3.3
227 NAME: ignore_ims_on_miss
230 Remove this line. The HTTP/1.1 feature is now configured by 'cache_miss_revalidate'.
233 # Options Removed in 3.2
234 NAME: chunked_request_body_max_size
237 Remove this line. Squid is now HTTP/1.1 compliant.
240 NAME: dns_v4_fallback
243 Remove this line. Squid performs a 'Happy Eyeballs' algorithm, the 'fallback' algorithm is no longer relevant.
246 NAME: emulate_httpd_log
249 Replace this with an access_log directive using the format 'common' or 'combined'.
255 Use a regular access.log with ACL limiting it to MISS events.
261 Remove this line. Configure FTP page display using the CSS controls in errorpages.css instead.
264 NAME: ignore_expect_100
267 Remove this line. The HTTP/1.1 feature is now fully supported by default.
273 Remove this option from your config. To log FQDN use %>A in the log format.
276 NAME: log_ip_on_direct
279 Remove this option from your config. To log server or peer names use %<A in the log format.
282 NAME: maximum_single_addr_tries
285 Replaced by connect_retries. The behaviour has changed, please read the documentation before altering.
288 NAME: referer_log referrer_log
291 Replace this with an access_log directive using the format 'referrer'.
297 Remove this line. The feature is supported by default in storage types where update is implemented.
300 NAME: url_rewrite_concurrency
303 Remove this line. Set the 'concurrency=' option of url_rewrite_children instead.
309 Replace this with an access_log directive using the format 'useragent'.
312 # Options Removed in 3.1
316 Remove this line. DNS is no longer tested on startup.
319 NAME: extension_methods
322 Remove this line. All valid methods for HTTP are accepted by default.
325 # 2.7 Options Removed/Replaced in 3.2
330 # 2.7 Options Removed/Replaced in 3.1
338 Remove this line. HTTP/1.1 is supported by default.
341 NAME: upgrade_http0.9
344 Remove this line. ICY/1.0 streaming protocol is supported by default.
347 NAME: zph_local zph_mode zph_option zph_parent zph_sibling
350 Alter these entries. Use the qos_flows directive instead.
353 # Options Removed in 3.0
357 Since squid-3.0 replace with request_header_access or reply_header_access
358 depending on whether you wish to match client requests or server replies.
361 NAME: httpd_accel_no_pmtu_disc
364 Since squid-3.0 use the 'disable-pmtu-discovery' flag on http_port instead.
367 NAME: wais_relay_host
370 Replace this line with 'cache_peer' configuration.
373 NAME: wais_relay_port
376 Replace this line with 'cache_peer' configuration.
381 -----------------------------------------------------------------------------
388 DEFAULT_DOC: SMP support disabled.
390 Number of main Squid processes or "workers" to fork and maintain.
391 0: "no daemon" mode, like running "squid -N ..."
392 1: "no SMP" mode, start one main Squid process daemon (default)
393 N: start N main Squid process daemons (i.e., SMP mode)
395 In SMP mode, each worker does nearly all what a single Squid daemon
396 does (e.g., listen on http_port and forward HTTP requests).
399 NAME: cpu_affinity_map
401 LOC: Config.cpuAffinityMap
403 DEFAULT_DOC: Let operating system decide.
405 Usage: cpu_affinity_map process_numbers=P1,P2,... cores=C1,C2,...
407 Sets 1:1 mapping between Squid processes and CPU cores. For example,
409 cpu_affinity_map process_numbers=1,2,3,4 cores=1,3,5,7
411 affects processes 1 through 4 only and places them on the first
412 four even cores, starting with core #1.
414 CPU cores are numbered starting from 1. Requires support for
415 sched_getaffinity(2) and sched_setaffinity(2) system calls.
417 Multiple cpu_affinity_map options are merged.
423 OPTIONS FOR AUTHENTICATION
424 -----------------------------------------------------------------------------
433 This is used to define parameters for the various authentication
434 schemes supported by Squid.
436 format: auth_param scheme parameter [setting]
438 The order in which authentication schemes are presented to the client is
439 dependent on the order the scheme first appears in config file. IE
440 has a bug (it's not RFC 2617 compliant) in that it will use the basic
441 scheme if basic is the first entry presented, even if more secure
442 schemes are presented. For now use the order in the recommended
443 settings section below. If other browsers have difficulties (don't
444 recognize the schemes offered even if you are using basic) either
445 put basic first, or disable the other schemes (by commenting out their
448 Once an authentication scheme is fully configured, it can only be
449 shutdown by shutting squid down and restarting. Changes can be made on
450 the fly and activated with a reconfigure. I.E. You can change to a
451 different helper, but not unconfigure the helper completely.
453 Please note that while this directive defines how Squid processes
454 authentication it does not automatically activate authentication.
455 To use authentication you must in addition make use of ACLs based
456 on login name in http_access (proxy_auth, proxy_auth_regex or
457 external with %LOGIN used in the format tag). The browser will be
458 challenged for authentication on the first such acl encountered
459 in http_access processing and will also be re-challenged for new
460 login credentials if the request is being denied by a proxy_auth
463 WARNING: authentication can't be used in a transparently intercepting
464 proxy as the client then thinks it is talking to an origin server and
465 not the proxy. This is a limitation of bending the TCP/IP protocol to
466 transparently intercepting port 80, not a limitation in Squid.
467 Ports flagged 'transparent', 'intercept', or 'tproxy' have
468 authentication disabled.
470 === Parameters common to all schemes. ===
473 Specifies the command for the external authenticator.
475 By default, each authentication scheme is not used unless a
476 program is specified.
478 See http://wiki.squid-cache.org/Features/AddonHelpers for
479 more details on helper operations and creating your own.
482 Specifies a string to be append to request line format for
483 the authentication helper. "Quoted" format values may contain
484 spaces and logformat %macros. In theory, any logformat %macro
485 can be used. In practice, a %macro expands as a dash (-) if
486 the helper request is sent before the required macro
487 information is available to Squid.
489 By default, Squid uses request formats provided in
490 scheme-specific examples below (search for %credentials).
492 The expanded key_extras value is added to the Squid credentials
493 cache and, hence, will affect authentication. It can be used to
494 autenticate different users with identical user names (e.g.,
495 when user authentication depends on http_port).
497 Avoid adding frequently changing information to key_extras. For
498 example, if you add user source IP, and it changes frequently
499 in your environment, then max_user_ip ACL is going to treat
500 every user+IP combination as a unique "user", breaking the ACL
501 and wasting a lot of memory on those user records. It will also
502 force users to authenticate from scratch whenever their IP
506 Specifies the protection scope (aka realm name) which is to be
507 reported to the client for the authentication scheme. It is
508 commonly part of the text the user will see when prompted for
509 their username and password.
511 For Basic the default is "Squid proxy-caching web server".
512 For Digest there is no default, this parameter is mandatory.
513 For NTLM and Negotiate this parameter is ignored.
515 "children" numberofchildren [startup=N] [idle=N] [concurrency=N] [queue-size=N]
517 The maximum number of authenticator processes to spawn. If
518 you start too few Squid will have to wait for them to process
519 a backlog of credential verifications, slowing it down. When
520 password verifications are done via a (slow) network you are
521 likely to need lots of authenticator processes.
523 The startup= and idle= options permit some skew in the exact
524 amount run. A minimum of startup=N will begin during startup
525 and reconfigure. Squid will start more in groups of up to
526 idle=N in an attempt to meet traffic needs and to keep idle=N
527 free above those traffic needs up to the maximum.
529 The concurrency= option sets the number of concurrent requests
530 the helper can process. The default of 0 is used for helpers
531 who only supports one request at a time. Setting this to a
532 number greater than 0 changes the protocol used to include a
533 channel ID field first on the request/response line, allowing
534 multiple requests to be sent to the same helper in parallel
535 without waiting for the response.
537 Concurrency must not be set unless it's known the helper
538 supports the input format with channel-ID fields.
540 The queue-size= option sets the maximum number of queued
541 requests. If the queued requests exceed queue size for more
542 than 3 minutes then squid aborts its operation.
543 The default value is set to 2*numberofchildren/
545 NOTE: NTLM and Negotiate schemes do not support concurrency
546 in the Squid code module even though some helpers can.
549 IF HAVE_AUTH_MODULE_BASIC
550 === Basic authentication parameters ===
553 HTTP uses iso-latin-1 as character set, while some
554 authentication backends such as LDAP expects UTF-8. If this is
555 set to on Squid will translate the HTTP iso-latin-1 charset to
556 UTF-8 before sending the username and password to the helper.
558 "credentialsttl" timetolive
559 Specifies how long squid assumes an externally validated
560 username:password pair is valid for - in other words how
561 often the helper program is called for that user. Set this
562 low to force revalidation with short lived passwords.
564 NOTE: setting this high does not impact your susceptibility
565 to replay attacks unless you are using an one-time password
566 system (such as SecureID). If you are using such a system,
567 you will be vulnerable to replay attacks unless you also
568 use the max_user_ip ACL in an http_access rule.
570 "casesensitive" on|off
571 Specifies if usernames are case sensitive. Most user databases
572 are case insensitive allowing the same username to be spelled
573 using both lower and upper case letters, but some are case
574 sensitive. This makes a big difference for user_max_ip ACL
575 processing and similar.
578 IF HAVE_AUTH_MODULE_DIGEST
579 === Digest authentication parameters ===
582 HTTP uses iso-latin-1 as character set, while some
583 authentication backends such as LDAP expects UTF-8. If this is
584 set to on Squid will translate the HTTP iso-latin-1 charset to
585 UTF-8 before sending the username and password to the helper.
587 "nonce_garbage_interval" timeinterval
588 Specifies the interval that nonces that have been issued
589 to client_agent's are checked for validity.
591 "nonce_max_duration" timeinterval
592 Specifies the maximum length of time a given nonce will be
595 "nonce_max_count" number
596 Specifies the maximum number of times a given nonce can be
599 "nonce_strictness" on|off
600 Determines if squid requires strict increment-by-1 behavior
601 for nonce counts, or just incrementing (off - for use when
602 user agents generate nonce counts that occasionally miss 1
603 (ie, 1,2,4,6)). Default off.
605 "check_nonce_count" on|off
606 This directive if set to off can disable the nonce count check
607 completely to work around buggy digest qop implementations in
608 certain mainstream browser versions. Default on to check the
609 nonce count to protect from authentication replay attacks.
611 "post_workaround" on|off
612 This is a workaround to certain buggy browsers who send an
613 incorrect request digest in POST requests when reusing the
614 same nonce as acquired earlier on a GET request.
617 IF HAVE_AUTH_MODULE_NEGOTIATE
618 === Negotiate authentication parameters ===
621 If you experience problems with PUT/POST requests when using
622 the this authentication scheme then you can try setting this
623 to off. This will cause Squid to forcibly close the connection
624 on the initial request where the browser asks which schemes
625 are supported by the proxy.
628 IF HAVE_AUTH_MODULE_NTLM
629 === NTLM authentication parameters ===
632 If you experience problems with PUT/POST requests when using
633 the this authentication scheme then you can try setting this
634 to off. This will cause Squid to forcibly close the connection
635 on the initial request where the browser asks which schemes
636 are supported by the proxy.
639 === Example Configuration ===
641 This configuration displays the recommended authentication scheme
642 order from most to least secure with recommended minimum configuration
643 settings for each scheme:
645 #auth_param negotiate program <uncomment and complete this line to activate>
646 #auth_param negotiate children 20 startup=0 idle=1
647 #auth_param negotiate keep_alive on
649 #auth_param digest program <uncomment and complete this line to activate>
650 #auth_param digest children 20 startup=0 idle=1
651 #auth_param digest realm Squid proxy-caching web server
652 #auth_param digest nonce_garbage_interval 5 minutes
653 #auth_param digest nonce_max_duration 30 minutes
654 #auth_param digest nonce_max_count 50
656 #auth_param ntlm program <uncomment and complete this line to activate>
657 #auth_param ntlm children 20 startup=0 idle=1
658 #auth_param ntlm keep_alive on
660 #auth_param basic program <uncomment and complete this line>
661 #auth_param basic children 5 startup=5 idle=1
662 #auth_param basic realm Squid proxy-caching web server
663 #auth_param basic credentialsttl 2 hours
666 NAME: authenticate_cache_garbage_interval
669 LOC: Config.authenticateGCInterval
671 The time period between garbage collection across the username cache.
672 This is a trade-off between memory utilization (long intervals - say
673 2 days) and CPU (short intervals - say 1 minute). Only change if you
677 NAME: authenticate_ttl
680 LOC: Config.authenticateTTL
682 The time a user & their credentials stay in the logged in
683 user cache since their last request. When the garbage
684 interval passes, all user credentials that have passed their
685 TTL are removed from memory.
688 NAME: authenticate_ip_ttl
690 LOC: Config.authenticateIpTTL
693 If you use proxy authentication and the 'max_user_ip' ACL,
694 this directive controls how long Squid remembers the IP
695 addresses associated with each user. Use a small value
696 (e.g., 60 seconds) if your users might change addresses
697 quickly, as is the case with dialup. You might be safe
698 using a larger value (e.g., 2 hours) in a corporate LAN
699 environment with relatively static address assignments.
704 -----------------------------------------------------------------------------
707 NAME: external_acl_type
708 TYPE: externalAclHelper
709 LOC: Config.externalAclHelperList
712 This option defines external acl classes using a helper program
713 to look up the status
715 external_acl_type name [options] FORMAT.. /path/to/helper [helper arguments..]
719 ttl=n TTL in seconds for cached results (defaults to 3600
723 TTL for cached negative lookups (default same
726 grace=n Percentage remaining of TTL where a refresh of a
727 cached entry should be initiated without needing to
728 wait for a new reply. (default is for no grace period)
730 cache=n The maximum number of entries in the result cache. The
731 default limit is 262144 entries. Each cache entry usually
732 consumes at least 256 bytes. Squid currently does not remove
733 expired cache entries until the limit is reached, so a proxy
734 will sooner or later reach the limit. The expanded FORMAT
735 value is used as the cache key, so if the details in FORMAT
736 are highly variable, a larger cache may be needed to produce
737 reduction in helper load.
740 Maximum number of acl helper processes spawned to service
741 external acl lookups of this type. (default 20)
744 Minimum number of acl helper processes to spawn during
745 startup and reconfigure to service external acl lookups
746 of this type. (default 0)
749 Number of acl helper processes to keep ahead of traffic
750 loads. Squid will spawn this many at once whenever load
751 rises above the capabilities of existing processes.
752 Up to the value of children-max. (default 1)
754 concurrency=n concurrency level per process. Only used with helpers
755 capable of processing more than one query at a time.
757 queue-size=N The queue-size= option sets the maximum number of queued
758 requests. If the queued requests exceed queue size
760 The default value is set to 2*children-max.
762 protocol=2.5 Compatibility mode for Squid-2.5 external acl helpers.
764 ipv4 / ipv6 IP protocol used to communicate with this helper.
765 The default is to auto-detect IPv6 and use it when available.
768 FORMAT is a series of %macro codes. See logformat directive for a full list
769 of the accepted codes. Although note that at the time of any external ACL
770 being tested data may not be available and thus some %macro expand to '-'.
772 In addition to the logformat codes; when processing external ACLs these
773 additional macros are made available:
775 %ACL The name of the ACL being tested.
777 %DATA The ACL arguments. If a logformat encoding modifier
778 is used it will encode the whole set of arguments
781 If not used; then any arguments are automatically
782 added at the end of the line sent to the helper
783 as separately URL-encoded fields.
785 If SSL is enabled, the following formating codes become available:
787 %USER_CERT SSL User certificate in PEM format
788 %USER_CERTCHAIN SSL User certificate chain in PEM format
789 %USER_CERT_xx SSL User certificate subject attribute xx
790 %USER_CA_CERT_xx SSL User certificate issuer attribute xx
793 NOTE: all other format codes accepted by older Squid versions
797 General request syntax:
799 [channel-ID] FORMAT-values [acl-values ...]
802 FORMAT-values consists of transaction details expanded with
803 whitespace separation per the config file FORMAT specification
804 using the FORMAT macros listed above.
806 acl-values consists of any string specified in the referencing
807 config 'acl ... external' line. see the "acl external" directive.
809 Request values sent to the helper are URL escaped to protect
810 each value in requests against whitespaces.
812 If using protocol=2.5 then the request sent to the helper is not
813 URL escaped to protect against whitespace.
815 NOTE: protocol=3.0 is deprecated as no longer necessary.
817 When using the concurrency= option the protocol is changed by
818 introducing a query channel tag in front of the request/response.
819 The query channel tag is a number between 0 and concurrency-1.
820 This value must be echoed back unchanged to Squid as the first part
821 of the response relating to its request.
824 The helper receives lines expanded per the above format specification
825 and for each input line returns 1 line starting with OK/ERR/BH result
826 code and optionally followed by additional keywords with more details.
829 General result syntax:
831 [channel-ID] result keyword=value ...
833 Result consists of one of the codes:
836 the ACL test produced a match.
839 the ACL test does not produce a match.
842 An internal error occurred in the helper, preventing
843 a result being identified.
845 The meaning of 'a match' is determined by your squid.conf
846 access control configuration. See the Squid wiki for details.
850 user= The users name (login)
852 password= The users password (for login= cache_peer option)
854 message= Message describing the reason for this response.
855 Available as %o in error pages.
856 Useful on (ERR and BH results).
858 tag= Apply a tag to a request. Only sets a tag once,
859 does not alter existing tags.
861 log= String to be logged in access.log. Available as
862 %ea in logformat specifications.
864 clt_conn_tag= Associates a TAG with the client TCP connection.
865 Please see url_rewrite_program related documentation
868 Any keywords may be sent on any response whether OK, ERR or BH.
870 All response keyword values need to be a single token with URL
871 escaping, or enclosed in double quotes (") and escaped using \ on
872 any double quotes or \ characters within the value. The wrapping
873 double quotes are removed before the value is interpreted by Squid.
874 \r and \n are also replace by CR and LF.
876 Some example key values:
880 user="J. \"Bob\" Smith"
887 DEFAULT: ssl::certHasExpired ssl_error X509_V_ERR_CERT_HAS_EXPIRED
888 DEFAULT: ssl::certNotYetValid ssl_error X509_V_ERR_CERT_NOT_YET_VALID
889 DEFAULT: ssl::certDomainMismatch ssl_error SQUID_X509_V_ERR_DOMAIN_MISMATCH
890 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
891 DEFAULT: ssl::certSelfSigned ssl_error X509_V_ERR_DEPTH_ZERO_SELF_SIGNED_CERT
894 DEFAULT: manager url_regex -i ^cache_object:// +i ^https?://[^/]+/squid-internal-mgr/
895 DEFAULT: localhost src 127.0.0.1/32 ::1
896 DEFAULT: to_localhost dst 127.0.0.0/8 0.0.0.0/32 ::1
897 DEFAULT_DOC: ACLs all, manager, localhost, and to_localhost are predefined.
899 Defining an Access List
901 Every access list definition must begin with an aclname and acltype,
902 followed by either type-specific arguments or a quoted filename that
905 acl aclname acltype argument ...
906 acl aclname acltype "file" ...
908 When using "file", the file should contain one item per line.
913 Some acl types supports options which changes their default behaviour:
915 -i,+i By default, regular expressions are CASE-SENSITIVE. To make them
916 case-insensitive, use the -i option. To return case-sensitive
917 use the +i option between patterns, or make a new ACL line
920 -n Disable lookups and address type conversions. If lookup or
921 conversion is required because the parameter type (IP or
922 domain name) does not match the message address type (domain
923 name or IP), then the ACL would immediately declare a mismatch
924 without any warnings or lookups.
927 Perform a list membership test, interpreting values as
928 comma-separated token lists and matching against individual
929 tokens instead of whole values.
930 The optional "delimiters" parameter specifies one or more
931 alternative non-alphanumeric delimiter characters.
932 non-alphanumeric delimiter characters.
934 -- Used to stop processing all options, in the case the first acl
935 value has '-' character as first character (for example the '-'
936 is a valid domain name)
938 Some acl types require suspending the current request in order
939 to access some external data source.
940 Those which do are marked with the tag [slow], those which
941 don't are marked as [fast].
942 See http://wiki.squid-cache.org/SquidFaq/SquidAcl
943 for further information
945 ***** ACL TYPES AVAILABLE *****
947 acl aclname src ip-address/mask ... # clients IP address [fast]
948 acl aclname src addr1-addr2/mask ... # range of addresses [fast]
949 acl aclname dst [-n] ip-address/mask ... # URL host's IP address [slow]
950 acl aclname localip ip-address/mask ... # IP address the client connected to [fast]
952 acl aclname arp mac-address ... (xx:xx:xx:xx:xx:xx notation)
954 # The 'arp' ACL code is not portable to all operating systems.
955 # It works on Linux, Solaris, Windows, FreeBSD, and some other
958 # NOTE: Squid can only determine the MAC/EUI address for IPv4
959 # clients that are on the same subnet. If the client is on a
960 # different subnet, then Squid cannot find out its address.
962 # NOTE 2: IPv6 protocol does not contain ARP. MAC/EUI is either
963 # encoded directly in the IPv6 address or not available.
965 acl aclname srcdomain .foo.com ...
966 # reverse lookup, from client IP [slow]
967 acl aclname dstdomain [-n] .foo.com ...
968 # Destination server from URL [fast]
969 acl aclname srcdom_regex [-i] \.foo\.com ...
970 # regex matching client name [slow]
971 acl aclname dstdom_regex [-n] [-i] \.foo\.com ...
972 # regex matching server [fast]
974 # For dstdomain and dstdom_regex a reverse lookup is tried if a IP
975 # based URL is used and no match is found. The name "none" is used
976 # if the reverse lookup fails.
978 acl aclname src_as number ...
979 acl aclname dst_as number ...
981 # Except for access control, AS numbers can be used for
982 # routing of requests to specific caches. Here's an
983 # example for routing all requests for AS#1241 and only
984 # those to mycache.mydomain.net:
985 # acl asexample dst_as 1241
986 # cache_peer_access mycache.mydomain.net allow asexample
987 # cache_peer_access mycache_mydomain.net deny all
989 acl aclname peername myPeer ...
991 # match against a named cache_peer entry
992 # set unique name= on cache_peer lines for reliable use.
994 acl aclname time [day-abbrevs] [h1:m1-h2:m2]
1004 # h1:m1 must be less than h2:m2
1006 acl aclname url_regex [-i] ^http:// ...
1007 # regex matching on whole URL [fast]
1008 acl aclname urllogin [-i] [^a-zA-Z0-9] ...
1009 # regex matching on URL login field
1010 acl aclname urlpath_regex [-i] \.gif$ ...
1011 # regex matching on URL path [fast]
1013 acl aclname port 80 70 21 0-1024... # destination TCP port [fast]
1015 acl aclname localport 3128 ... # TCP port the client connected to [fast]
1016 # NP: for interception mode this is usually '80'
1018 acl aclname myportname 3128 ... # *_port name [fast]
1020 acl aclname proto HTTP FTP ... # request protocol [fast]
1022 acl aclname method GET POST ... # HTTP request method [fast]
1024 acl aclname http_status 200 301 500- 400-403 ...
1025 # status code in reply [fast]
1027 acl aclname browser [-i] regexp ...
1028 # pattern match on User-Agent header (see also req_header below) [fast]
1030 acl aclname referer_regex [-i] regexp ...
1031 # pattern match on Referer header [fast]
1032 # Referer is highly unreliable, so use with care
1034 acl aclname ident username ...
1035 acl aclname ident_regex [-i] pattern ...
1036 # string match on ident output [slow]
1037 # use REQUIRED to accept any non-null ident.
1039 acl aclname proxy_auth [-i] username ...
1040 acl aclname proxy_auth_regex [-i] pattern ...
1041 # perform http authentication challenge to the client and match against
1042 # supplied credentials [slow]
1044 # takes a list of allowed usernames.
1045 # use REQUIRED to accept any valid username.
1047 # Will use proxy authentication in forward-proxy scenarios, and plain
1048 # http authenticaiton in reverse-proxy scenarios
1050 # NOTE: when a Proxy-Authentication header is sent but it is not
1051 # needed during ACL checking the username is NOT logged
1054 # NOTE: proxy_auth requires a EXTERNAL authentication program
1055 # to check username/password combinations (see
1056 # auth_param directive).
1058 # NOTE: proxy_auth can't be used in a transparent/intercepting proxy
1059 # as the browser needs to be configured for using a proxy in order
1060 # to respond to proxy authentication.
1062 acl aclname snmp_community string ...
1063 # A community string to limit access to your SNMP Agent [fast]
1066 # acl snmppublic snmp_community public
1068 acl aclname maxconn number
1069 # This will be matched when the client's IP address has
1070 # more than <number> TCP connections established. [fast]
1071 # NOTE: This only measures direct TCP links so X-Forwarded-For
1072 # indirect clients are not counted.
1074 acl aclname max_user_ip [-s] number
1075 # This will be matched when the user attempts to log in from more
1076 # than <number> different ip addresses. The authenticate_ip_ttl
1077 # parameter controls the timeout on the ip entries. [fast]
1078 # If -s is specified the limit is strict, denying browsing
1079 # from any further IP addresses until the ttl has expired. Without
1080 # -s Squid will just annoy the user by "randomly" denying requests.
1081 # (the counter is reset each time the limit is reached and a
1082 # request is denied)
1083 # NOTE: in acceleration mode or where there is mesh of child proxies,
1084 # clients may appear to come from multiple addresses if they are
1085 # going through proxy farms, so a limit of 1 may cause user problems.
1087 acl aclname random probability
1088 # Pseudo-randomly match requests. Based on the probability given.
1089 # Probability may be written as a decimal (0.333), fraction (1/3)
1090 # or ratio of matches:non-matches (3:5).
1092 acl aclname req_mime_type [-i] mime-type ...
1093 # regex match against the mime type of the request generated
1094 # by the client. Can be used to detect file upload or some
1095 # types HTTP tunneling requests [fast]
1096 # NOTE: This does NOT match the reply. You cannot use this
1097 # to match the returned file type.
1099 acl aclname req_header header-name [-i] any\.regex\.here
1100 # regex match against any of the known request headers. May be
1101 # thought of as a superset of "browser", "referer" and "mime-type"
1104 acl aclname rep_mime_type [-i] mime-type ...
1105 # regex match against the mime type of the reply received by
1106 # squid. Can be used to detect file download or some
1107 # types HTTP tunneling requests. [fast]
1108 # NOTE: This has no effect in http_access rules. It only has
1109 # effect in rules that affect the reply data stream such as
1110 # http_reply_access.
1112 acl aclname rep_header header-name [-i] any\.regex\.here
1113 # regex match against any of the known reply headers. May be
1114 # thought of as a superset of "browser", "referer" and "mime-type"
1117 acl aclname external class_name [arguments...]
1118 # external ACL lookup via a helper class defined by the
1119 # external_acl_type directive [slow]
1121 acl aclname user_cert attribute values...
1122 # match against attributes in a user SSL certificate
1123 # attribute is one of DN/C/O/CN/L/ST or a numerical OID [fast]
1125 acl aclname ca_cert attribute values...
1126 # match against attributes a users issuing CA SSL certificate
1127 # attribute is one of DN/C/O/CN/L/ST or a numerical OID [fast]
1129 acl aclname ext_user username ...
1130 acl aclname ext_user_regex [-i] pattern ...
1131 # string match on username returned by external acl helper [slow]
1132 # use REQUIRED to accept any non-null user name.
1134 acl aclname tag tagvalue ...
1135 # string match on tag returned by external acl helper [fast]
1136 # DEPRECATED. Only the first tag will match with this ACL.
1137 # Use the 'note' ACL instead for handling multiple tag values.
1139 acl aclname hier_code codename ...
1140 # string match against squid hierarchy code(s); [fast]
1141 # e.g., DIRECT, PARENT_HIT, NONE, etc.
1143 # NOTE: This has no effect in http_access rules. It only has
1144 # effect in rules that affect the reply data stream such as
1145 # http_reply_access.
1147 acl aclname note [-m[=delimiters]] name [value ...]
1148 # match transaction annotation [fast]
1149 # Without values, matches any annotation with a given name.
1150 # With value(s), matches any annotation with a given name that
1151 # also has one of the given values.
1152 # If the -m flag is used, then the value of the named
1153 # annotation is interpreted as a list of tokens, and the ACL
1154 # matches individual name=token pairs rather than whole
1155 # name=value pairs. See "ACL Options" above for more info.
1156 # Annotation sources include note and adaptation_meta directives
1157 # as well as helper and eCAP responses.
1159 acl aclname adaptation_service service ...
1160 # Matches the name of any icap_service, ecap_service,
1161 # adaptation_service_set, or adaptation_service_chain that Squid
1162 # has used (or attempted to use) for the master transaction.
1163 # This ACL must be defined after the corresponding adaptation
1164 # service is named in squid.conf. This ACL is usable with
1165 # adaptation_meta because it starts matching immediately after
1166 # the service has been selected for adaptation.
1169 acl aclname ssl_error errorname
1170 # match against SSL certificate validation error [fast]
1172 # For valid error names see in @DEFAULT_ERROR_DIR@/templates/error-details.txt
1175 # The following can be used as shortcuts for certificate properties:
1176 # [ssl::]certHasExpired: the "not after" field is in the past
1177 # [ssl::]certNotYetValid: the "not before" field is in the future
1178 # [ssl::]certUntrusted: The certificate issuer is not to be trusted.
1179 # [ssl::]certSelfSigned: The certificate is self signed.
1180 # [ssl::]certDomainMismatch: The certificate CN domain does not
1181 # match the name the name of the host we are connecting to.
1183 # The ssl::certHasExpired, ssl::certNotYetValid, ssl::certDomainMismatch,
1184 # ssl::certUntrusted, and ssl::certSelfSigned can also be used as
1185 # predefined ACLs, just like the 'all' ACL.
1187 # NOTE: The ssl_error ACL is only supported with sslproxy_cert_error,
1188 # sslproxy_cert_sign, and sslproxy_cert_adapt options.
1190 acl aclname server_cert_fingerprint [-sha1] fingerprint
1191 # match against server SSL certificate fingerprint [fast]
1193 # The fingerprint is the digest of the DER encoded version
1194 # of the whole certificate. The user should use the form: XX:XX:...
1195 # Optional argument specifies the digest algorithm to use.
1196 # The SHA1 digest algorithm is the default and is currently
1197 # the only algorithm supported (-sha1).
1199 acl aclname at_step step
1200 # match against the current step during ssl_bump evaluation [fast]
1201 # Never matches and should not be used outside the ssl_bump context.
1203 # At each SslBump step, Squid evaluates ssl_bump directives to find
1204 # the next bumping action (e.g., peek or splice). Valid SslBump step
1205 # values and the corresponding ssl_bump evaluation moments are:
1206 # SslBump1: After getting TCP-level and HTTP CONNECT info.
1207 # SslBump2: After getting SSL Client Hello info.
1208 # SslBump3: After getting SSL Server Hello info.
1210 acl aclname ssl::server_name .foo.com ...
1211 # matches server name obtained from various sources [fast]
1213 # The server name is obtained during Ssl-Bump steps from such sources
1214 # as CONNECT request URI, client SNI, and SSL server certificate CN.
1215 # During each Ssl-Bump step, Squid may improve its understanding of a
1216 # "true server name". Unlike dstdomain, this ACL does not perform
1219 acl aclname ssl::server_name_regex [-i] \.foo\.com ...
1220 # regex matches server name obtained from various sources [fast]
1222 acl aclname any-of acl1 acl2 ...
1223 # match any one of the acls [fast or slow]
1224 # The first matching ACL stops further ACL evaluation.
1226 # ACLs from multiple any-of lines with the same name are ORed.
1227 # For example, A = (a1 or a2) or (a3 or a4) can be written as
1228 # acl A any-of a1 a2
1229 # acl A any-of a3 a4
1231 # This group ACL is fast if all evaluated ACLs in the group are fast
1232 # and slow otherwise.
1234 acl aclname all-of acl1 acl2 ...
1235 # match all of the acls [fast or slow]
1236 # The first mismatching ACL stops further ACL evaluation.
1238 # ACLs from multiple all-of lines with the same name are ORed.
1239 # For example, B = (b1 and b2) or (b3 and b4) can be written as
1240 # acl B all-of b1 b2
1241 # acl B all-of b3 b4
1243 # This group ACL is fast if all evaluated ACLs in the group are fast
1244 # and slow otherwise.
1247 acl macaddress arp 09:00:2b:23:45:67
1248 acl myexample dst_as 1241
1249 acl password proxy_auth REQUIRED
1250 acl fileupload req_mime_type -i ^multipart/form-data$
1251 acl javascript rep_mime_type -i ^application/x-javascript$
1255 # Recommended minimum configuration:
1258 # Example rule allowing access from your local networks.
1259 # Adapt to list your (internal) IP networks from where browsing
1261 acl localnet src 0.0.0.1-0.255.255.255 # RFC 1122 "this" network (LAN)
1262 acl localnet src 10.0.0.0/8 # RFC 1918 local private network (LAN)
1263 acl localnet src 100.64.0.0/10 # RFC 6598 shared address space (CGN)
1264 acl localhet src 169.254.0.0/16 # RFC 3927 link-local (directly plugged) machines
1265 acl localnet src 172.16.0.0/12 # RFC 1918 local private network (LAN)
1266 acl localnet src 192.168.0.0/16 # RFC 1918 local private network (LAN)
1267 acl localnet src fc00::/7 # RFC 4193 local private network range
1268 acl localnet src fe80::/10 # RFC 4291 link-local (directly plugged) machines
1270 acl SSL_ports port 443
1271 acl Safe_ports port 80 # http
1272 acl Safe_ports port 21 # ftp
1273 acl Safe_ports port 443 # https
1274 acl Safe_ports port 70 # gopher
1275 acl Safe_ports port 210 # wais
1276 acl Safe_ports port 1025-65535 # unregistered ports
1277 acl Safe_ports port 280 # http-mgmt
1278 acl Safe_ports port 488 # gss-http
1279 acl Safe_ports port 591 # filemaker
1280 acl Safe_ports port 777 # multiling http
1281 acl CONNECT method CONNECT
1285 NAME: proxy_protocol_access
1287 LOC: Config.accessList.proxyProtocol
1289 DEFAULT_DOC: all TCP connections to ports with require-proxy-header will be denied
1291 Determine which client proxies can be trusted to provide correct
1292 information regarding real client IP address using PROXY protocol.
1294 Requests may pass through a chain of several other proxies
1295 before reaching us. The original source details may by sent in:
1296 * HTTP message Forwarded header, or
1297 * HTTP message X-Forwarded-For header, or
1298 * PROXY protocol connection header.
1300 This directive is solely for validating new PROXY protocol
1301 connections received from a port flagged with require-proxy-header.
1302 It is checked only once after TCP connection setup.
1304 A deny match results in TCP connection closure.
1306 An allow match is required for Squid to permit the corresponding
1307 TCP connection, before Squid even looks for HTTP request headers.
1308 If there is an allow match, Squid starts using PROXY header information
1309 to determine the source address of the connection for all future ACL
1310 checks, logging, etc.
1312 SECURITY CONSIDERATIONS:
1314 Any host from which we accept client IP details can place
1315 incorrect information in the relevant header, and Squid
1316 will use the incorrect information as if it were the
1317 source address of the request. This may enable remote
1318 hosts to bypass any access control restrictions that are
1319 based on the client's source addresses.
1321 This clause only supports fast acl types.
1322 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1325 NAME: follow_x_forwarded_for
1327 IFDEF: FOLLOW_X_FORWARDED_FOR
1328 LOC: Config.accessList.followXFF
1329 DEFAULT_IF_NONE: deny all
1330 DEFAULT_DOC: X-Forwarded-For header will be ignored.
1332 Determine which client proxies can be trusted to provide correct
1333 information regarding real client IP address.
1335 Requests may pass through a chain of several other proxies
1336 before reaching us. The original source details may by sent in:
1337 * HTTP message Forwarded header, or
1338 * HTTP message X-Forwarded-For header, or
1339 * PROXY protocol connection header.
1341 PROXY protocol connections are controlled by the proxy_protocol_access
1342 directive which is checked before this.
1344 If a request reaches us from a source that is allowed by this
1345 directive, then we trust the information it provides regarding
1346 the IP of the client it received from (if any).
1348 For the purpose of ACLs used in this directive the src ACL type always
1349 matches the address we are testing and srcdomain matches its rDNS.
1351 On each HTTP request Squid checks for X-Forwarded-For header fields.
1352 If found the header values are iterated in reverse order and an allow
1353 match is required for Squid to continue on to the next value.
1354 The verification ends when a value receives a deny match, cannot be
1355 tested, or there are no more values to test.
1356 NOTE: Squid does not yet follow the Forwarded HTTP header.
1358 The end result of this process is an IP address that we will
1359 refer to as the indirect client address. This address may
1360 be treated as the client address for access control, ICAP, delay
1361 pools and logging, depending on the acl_uses_indirect_client,
1362 icap_uses_indirect_client, delay_pool_uses_indirect_client,
1363 log_uses_indirect_client and tproxy_uses_indirect_client options.
1365 This clause only supports fast acl types.
1366 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1368 SECURITY CONSIDERATIONS:
1370 Any host from which we accept client IP details can place
1371 incorrect information in the relevant header, and Squid
1372 will use the incorrect information as if it were the
1373 source address of the request. This may enable remote
1374 hosts to bypass any access control restrictions that are
1375 based on the client's source addresses.
1379 acl localhost src 127.0.0.1
1380 acl my_other_proxy srcdomain .proxy.example.com
1381 follow_x_forwarded_for allow localhost
1382 follow_x_forwarded_for allow my_other_proxy
1385 NAME: acl_uses_indirect_client
1388 IFDEF: FOLLOW_X_FORWARDED_FOR
1390 LOC: Config.onoff.acl_uses_indirect_client
1392 Controls whether the indirect client address
1393 (see follow_x_forwarded_for) is used instead of the
1394 direct client address in acl matching.
1396 NOTE: maxconn ACL considers direct TCP links and indirect
1397 clients will always have zero. So no match.
1400 NAME: delay_pool_uses_indirect_client
1403 IFDEF: FOLLOW_X_FORWARDED_FOR&&USE_DELAY_POOLS
1405 LOC: Config.onoff.delay_pool_uses_indirect_client
1407 Controls whether the indirect client address
1408 (see follow_x_forwarded_for) is used instead of the
1409 direct client address in delay pools.
1412 NAME: log_uses_indirect_client
1415 IFDEF: FOLLOW_X_FORWARDED_FOR
1417 LOC: Config.onoff.log_uses_indirect_client
1419 Controls whether the indirect client address
1420 (see follow_x_forwarded_for) is used instead of the
1421 direct client address in the access log.
1424 NAME: tproxy_uses_indirect_client
1427 IFDEF: FOLLOW_X_FORWARDED_FOR&&LINUX_NETFILTER
1429 LOC: Config.onoff.tproxy_uses_indirect_client
1431 Controls whether the indirect client address
1432 (see follow_x_forwarded_for) is used instead of the
1433 direct client address when spoofing the outgoing client.
1435 This has no effect on requests arriving in non-tproxy
1438 SECURITY WARNING: Usage of this option is dangerous
1439 and should not be used trivially. Correct configuration
1440 of follow_x_forwarded_for with a limited set of trusted
1441 sources is required to prevent abuse of your proxy.
1444 NAME: spoof_client_ip
1446 LOC: Config.accessList.spoof_client_ip
1448 DEFAULT_DOC: Allow spoofing on all TPROXY traffic.
1450 Control client IP address spoofing of TPROXY traffic based on
1451 defined access lists.
1453 spoof_client_ip allow|deny [!]aclname ...
1455 If there are no "spoof_client_ip" lines present, the default
1456 is to "allow" spoofing of any suitable request.
1458 Note that the cache_peer "no-tproxy" option overrides this ACL.
1460 This clause supports fast acl types.
1461 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1466 LOC: Config.accessList.http
1467 DEFAULT_IF_NONE: deny all
1468 DEFAULT_DOC: Deny, unless rules exist in squid.conf.
1470 Allowing or Denying access based on defined access lists
1472 To allow or deny a message received on an HTTP, HTTPS, or FTP port:
1473 http_access allow|deny [!]aclname ...
1475 NOTE on default values:
1477 If there are no "access" lines present, the default is to deny
1480 If none of the "access" lines cause a match, the default is the
1481 opposite of the last line in the list. If the last line was
1482 deny, the default is allow. Conversely, if the last line
1483 is allow, the default will be deny. For these reasons, it is a
1484 good idea to have an "deny all" entry at the end of your access
1485 lists to avoid potential confusion.
1487 This clause supports both fast and slow acl types.
1488 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1493 # Recommended minimum Access Permission configuration:
1495 # Deny requests to certain unsafe ports
1496 http_access deny !Safe_ports
1498 # Deny CONNECT to other than secure SSL ports
1499 http_access deny CONNECT !SSL_ports
1501 # Only allow cachemgr access from localhost
1502 http_access allow localhost manager
1503 http_access deny manager
1505 # We strongly recommend the following be uncommented to protect innocent
1506 # web applications running on the proxy server who think the only
1507 # one who can access services on "localhost" is a local user
1508 #http_access deny to_localhost
1511 # INSERT YOUR OWN RULE(S) HERE TO ALLOW ACCESS FROM YOUR CLIENTS
1514 # Example rule allowing access from your local networks.
1515 # Adapt localnet in the ACL section to list your (internal) IP networks
1516 # from where browsing should be allowed
1517 http_access allow localnet
1518 http_access allow localhost
1520 # And finally deny all other access to this proxy
1521 http_access deny all
1525 NAME: adapted_http_access http_access2
1527 LOC: Config.accessList.adapted_http
1529 DEFAULT_DOC: Allow, unless rules exist in squid.conf.
1531 Allowing or Denying access based on defined access lists
1533 Essentially identical to http_access, but runs after redirectors
1534 and ICAP/eCAP adaptation. Allowing access control based on their
1537 If not set then only http_access is used.
1540 NAME: http_reply_access
1542 LOC: Config.accessList.reply
1544 DEFAULT_DOC: Allow, unless rules exist in squid.conf.
1546 Allow replies to client requests. This is complementary to http_access.
1548 http_reply_access allow|deny [!] aclname ...
1550 NOTE: if there are no access lines present, the default is to allow
1553 If none of the access lines cause a match the opposite of the
1554 last line will apply. Thus it is good practice to end the rules
1555 with an "allow all" or "deny all" entry.
1557 This clause supports both fast and slow acl types.
1558 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1563 LOC: Config.accessList.icp
1565 DEFAULT_DOC: Deny, unless rules exist in squid.conf.
1567 Allowing or Denying access to the ICP port based on defined
1570 icp_access allow|deny [!]aclname ...
1572 NOTE: The default if no icp_access lines are present is to
1573 deny all traffic. This default may cause problems with peers
1576 This clause only supports fast acl types.
1577 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1579 # Allow ICP queries from local networks only
1580 #icp_access allow localnet
1581 #icp_access deny all
1587 LOC: Config.accessList.htcp
1589 DEFAULT_DOC: Deny, unless rules exist in squid.conf.
1591 Allowing or Denying access to the HTCP port based on defined
1594 htcp_access allow|deny [!]aclname ...
1596 See also htcp_clr_access for details on access control for
1597 cache purge (CLR) HTCP messages.
1599 NOTE: The default if no htcp_access lines are present is to
1600 deny all traffic. This default may cause problems with peers
1601 using the htcp option.
1603 This clause only supports fast acl types.
1604 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1606 # Allow HTCP queries from local networks only
1607 #htcp_access allow localnet
1608 #htcp_access deny all
1611 NAME: htcp_clr_access
1614 LOC: Config.accessList.htcp_clr
1616 DEFAULT_DOC: Deny, unless rules exist in squid.conf.
1618 Allowing or Denying access to purge content using HTCP based
1619 on defined access lists.
1620 See htcp_access for details on general HTCP access control.
1622 htcp_clr_access allow|deny [!]aclname ...
1624 This clause only supports fast acl types.
1625 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1627 # Allow HTCP CLR requests from trusted peers
1628 acl htcp_clr_peer src 192.0.2.2 2001:DB8::2
1629 htcp_clr_access allow htcp_clr_peer
1630 htcp_clr_access deny all
1635 LOC: Config.accessList.miss
1637 DEFAULT_DOC: Allow, unless rules exist in squid.conf.
1639 Determines whether network access is permitted when satisfying a request.
1642 to force your neighbors to use you as a sibling instead of
1645 acl localclients src 192.0.2.0/24 2001:DB8::a:0/64
1646 miss_access deny !localclients
1647 miss_access allow all
1649 This means only your local clients are allowed to fetch relayed/MISS
1650 replies from the network and all other clients can only fetch cached
1653 The default for this setting allows all clients who passed the
1654 http_access rules to relay via this proxy.
1656 This clause only supports fast acl types.
1657 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1660 NAME: ident_lookup_access
1664 DEFAULT_DOC: Unless rules exist in squid.conf, IDENT is not fetched.
1665 LOC: Ident::TheConfig.identLookup
1667 A list of ACL elements which, if matched, cause an ident
1668 (RFC 931) lookup to be performed for this request. For
1669 example, you might choose to always perform ident lookups
1670 for your main multi-user Unix boxes, but not for your Macs
1671 and PCs. By default, ident lookups are not performed for
1674 To enable ident lookups for specific client addresses, you
1675 can follow this example:
1677 acl ident_aware_hosts src 198.168.1.0/24
1678 ident_lookup_access allow ident_aware_hosts
1679 ident_lookup_access deny all
1681 Only src type ACL checks are fully supported. A srcdomain
1682 ACL might work at times, but it will not always provide
1685 This clause only supports fast acl types.
1686 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
1689 NAME: reply_body_max_size
1690 COMMENT: size [acl acl...]
1693 DEFAULT_DOC: No limit is applied.
1694 LOC: Config.ReplyBodySize
1696 This option specifies the maximum size of a reply body. It can be
1697 used to prevent users from downloading very large files, such as
1698 MP3's and movies. When the reply headers are received, the
1699 reply_body_max_size lines are processed, and the first line where
1700 all (if any) listed ACLs are true is used as the maximum body size
1703 This size is checked twice. First when we get the reply headers,
1704 we check the content-length value. If the content length value exists
1705 and is larger than the allowed size, the request is denied and the
1706 user receives an error message that says "the request or reply
1707 is too large." If there is no content-length, and the reply
1708 size exceeds this limit, the client's connection is just closed
1709 and they will receive a partial reply.
1711 WARNING: downstream caches probably can not detect a partial reply
1712 if there is no content-length header, so they will cache
1713 partial responses and give them out as hits. You should NOT
1714 use this option if you have downstream caches.
1716 WARNING: A maximum size smaller than the size of squid's error messages
1717 will cause an infinite loop and crash squid. Ensure that the smallest
1718 non-zero value you use is greater that the maximum header size plus
1719 the size of your largest error page.
1721 If you set this parameter none (the default), there will be
1724 Configuration Format is:
1725 reply_body_max_size SIZE UNITS [acl ...]
1727 reply_body_max_size 10 MB
1731 NAME: on_unsupported_protocol
1732 TYPE: on_unsupported_protocol
1733 LOC: Config.accessList.on_unsupported_protocol
1735 DEFAULT_DOC: Respond with an error message to unidentifiable traffic
1737 Determines Squid behavior when encountering strange requests at the
1738 beginning of an accepted TCP connection or the beginning of a bumped
1739 CONNECT tunnel. Controlling Squid reaction to unexpected traffic is
1740 especially useful in interception environments where Squid is likely
1741 to see connections for unsupported protocols that Squid should either
1742 terminate or tunnel at TCP level.
1744 on_unsupported_protocol <action> [!]acl ...
1746 The first matching action wins. Only fast ACLs are supported.
1748 Supported actions are:
1750 tunnel: Establish a TCP connection with the intended server and
1751 blindly shovel TCP packets between the client and server.
1753 respond: Respond with an error message, using the transfer protocol
1754 for the Squid port that received the request (e.g., HTTP
1755 for connections intercepted at the http_port). This is the
1758 Squid expects the following traffic patterns:
1760 http_port: a plain HTTP request
1761 https_port: SSL/TLS handshake followed by an [encrypted] HTTP request
1762 ftp_port: a plain FTP command (no on_unsupported_protocol support yet!)
1763 CONNECT tunnel on http_port: same as https_port
1764 CONNECT tunnel on https_port: same as https_port
1766 Currently, this directive has effect on intercepted connections and
1767 bumped tunnels only. Other cases are not supported because Squid
1768 cannot know the intended destination of other traffic.
1771 # define what Squid errors indicate receiving non-HTTP traffic:
1772 acl foreignProtocol squid_error ERR_PROTOCOL_UNKNOWN ERR_TOO_BIG
1773 # define what Squid errors indicate receiving nothing:
1774 acl serverTalksFirstProtocol squid_error ERR_REQUEST_START_TIMEOUT
1775 # tunnel everything that does not look like HTTP:
1776 on_unsupported_protocol tunnel foreignProtocol
1777 # tunnel if we think the client waits for the server to talk first:
1778 on_unsupported_protocol tunnel serverTalksFirstProtocol
1779 # in all other error cases, just send an HTTP "error page" response:
1780 on_unsupported_protocol respond all
1782 See also: squid_error ACL
1787 -----------------------------------------------------------------------------
1790 NAME: http_port ascii_port
1795 Usage: port [mode] [options]
1796 hostname:port [mode] [options]
1797 1.2.3.4:port [mode] [options]
1799 The socket addresses where Squid will listen for HTTP client
1800 requests. You may specify multiple socket addresses.
1801 There are three forms: port alone, hostname with port, and
1802 IP address with port. If you specify a hostname or IP
1803 address, Squid binds the socket to that specific
1804 address. Most likely, you do not need to bind to a specific
1805 address, so you can use the port number alone.
1807 If you are running Squid in accelerator mode, you
1808 probably want to listen on port 80 also, or instead.
1810 The -a command line option may be used to specify additional
1811 port(s) where Squid listens for proxy request. Such ports will
1812 be plain proxy ports with no options.
1814 You may specify multiple socket addresses on multiple lines.
1818 intercept Support for IP-Layer NAT interception delivering
1819 traffic to this Squid port.
1820 NP: disables authentication on the port.
1822 tproxy Support Linux TPROXY (or BSD divert-to) with spoofing
1823 of outgoing connections using the client IP address.
1824 NP: disables authentication on the port.
1826 accel Accelerator / reverse proxy mode
1828 ssl-bump For each CONNECT request allowed by ssl_bump ACLs,
1829 establish secure connection with the client and with
1830 the server, decrypt HTTPS messages as they pass through
1831 Squid, and treat them as unencrypted HTTP messages,
1832 becoming the man-in-the-middle.
1834 The ssl_bump option is required to fully enable
1835 bumping of CONNECT requests.
1837 Omitting the mode flag causes default forward proxy mode to be used.
1840 Accelerator Mode Options:
1842 defaultsite=domainname
1843 What to use for the Host: header if it is not present
1844 in a request. Determines what site (not origin server)
1845 accelerators should consider the default.
1847 no-vhost Disable using HTTP/1.1 Host header for virtual domain support.
1849 protocol= Protocol to reconstruct accelerated and intercepted
1850 requests with. Defaults to HTTP/1.1 for http_port and
1851 HTTPS/1.1 for https_port.
1852 When an unsupported value is configured Squid will
1853 produce a FATAL error.
1854 Values: HTTP or HTTP/1.1, HTTPS or HTTPS/1.1
1856 vport Virtual host port support. Using the http_port number
1857 instead of the port passed on Host: headers.
1859 vport=NN Virtual host port support. Using the specified port
1860 number instead of the port passed on Host: headers.
1863 Act as if this Squid is the origin server.
1864 This currently means generate new Date: and Expires:
1865 headers on HIT instead of adding Age:.
1867 ignore-cc Ignore request Cache-Control headers.
1869 WARNING: This option violates HTTP specifications if
1870 used in non-accelerator setups.
1872 allow-direct Allow direct forwarding in accelerator mode. Normally
1873 accelerated requests are denied direct forwarding as if
1874 never_direct was used.
1876 WARNING: this option opens accelerator mode to security
1877 vulnerabilities usually only affecting in interception
1878 mode. Make sure to protect forwarding with suitable
1879 http_access rules when using this.
1882 SSL Bump Mode Options:
1883 In addition to these options ssl-bump requires TLS/SSL options.
1885 generate-host-certificates[=<on|off>]
1886 Dynamically create SSL server certificates for the
1887 destination hosts of bumped CONNECT requests.When
1888 enabled, the cert and key options are used to sign
1889 generated certificates. Otherwise generated
1890 certificate will be selfsigned.
1891 If there is a CA certificate lifetime of the generated
1892 certificate equals lifetime of the CA certificate. If
1893 generated certificate is selfsigned lifetime is three
1895 This option is enabled by default when ssl-bump is used.
1896 See the ssl-bump option above for more information.
1898 dynamic_cert_mem_cache_size=SIZE
1899 Approximate total RAM size spent on cached generated
1900 certificates. If set to zero, caching is disabled. The
1901 default value is 4MB.
1905 cert= Path to SSL certificate (PEM format).
1907 key= Path to SSL private key file (PEM format)
1908 if not specified, the certificate file is
1909 assumed to be a combined certificate and
1912 cipher= Colon separated list of supported ciphers.
1913 NOTE: some ciphers such as EDH ciphers depend on
1914 additional settings. If those settings are
1915 omitted the ciphers may be silently ignored
1916 by the OpenSSL library.
1918 options= Various SSL implementation options. The most important
1921 NO_SSLv3 Disallow the use of SSLv3
1923 NO_TLSv1 Disallow the use of TLSv1.0
1925 NO_TLSv1_1 Disallow the use of TLSv1.1
1927 NO_TLSv1_2 Disallow the use of TLSv1.2
1930 Always create a new key when using
1931 temporary/ephemeral DH key exchanges
1934 Enable ephemeral ECDH key exchange.
1935 The adopted curve should be specified
1936 using the tls-dh option.
1939 Disable use of RFC5077 session tickets.
1940 Some servers may have problems
1941 understanding the TLS extension due
1942 to ambiguous specification in RFC4507.
1944 ALL Enable various bug workarounds
1945 suggested as "harmless" by OpenSSL
1946 Be warned that this reduces SSL/TLS
1947 strength to some attacks.
1949 See the OpenSSL SSL_CTX_set_options documentation for a
1952 clientca= File containing the list of CAs to use when
1953 requesting a client certificate.
1955 tls-cafile= PEM file containing CA certificates to use when verifying
1956 client certificates. If not configured clientca will be
1957 used. May be repeated to load multiple files.
1959 capath= Directory containing additional CA certificates
1960 and CRL lists to use when verifying client certificates.
1961 Requires OpenSSL or LibreSSL.
1963 crlfile= File of additional CRL lists to use when verifying
1964 the client certificate, in addition to CRLs stored in
1965 the capath. Implies VERIFY_CRL flag below.
1968 File containing DH parameters for temporary/ephemeral DH key
1969 exchanges, optionally prefixed by a curve for ephemeral ECDH
1971 See OpenSSL documentation for details on how to create the
1972 DH parameter file. Supported curves for ECDH can be listed
1973 using the "openssl ecparam -list_curves" command.
1974 WARNING: EDH and EECDH ciphers will be silently disabled if
1975 this option is not set.
1977 sslflags= Various flags modifying the use of SSL:
1979 Don't request client certificates
1980 immediately, but wait until acl processing
1981 requires a certificate (not yet implemented).
1983 Don't allow for session reuse. Each connection
1984 will result in a new SSL session.
1986 Verify CRL lists when accepting client
1989 Verify CRL lists for all certificates in the
1990 client certificate chain.
1993 Do not use the system default Trusted CA.
1995 tls-no-npn Do not use the TLS NPN extension to advertise HTTP/1.1.
1997 sslcontext= SSL session ID context identifier.
2001 connection-auth[=on|off]
2002 use connection-auth=off to tell Squid to prevent
2003 forwarding Microsoft connection oriented authentication
2004 (NTLM, Negotiate and Kerberos)
2006 disable-pmtu-discovery=
2007 Control Path-MTU discovery usage:
2008 off lets OS decide on what to do (default).
2009 transparent disable PMTU discovery when transparent
2011 always disable always PMTU discovery.
2013 In many setups of transparently intercepting proxies
2014 Path-MTU discovery can not work on traffic towards the
2015 clients. This is the case when the intercepting device
2016 does not fully track connections and fails to forward
2017 ICMP must fragment messages to the cache server. If you
2018 have such setup and experience that certain clients
2019 sporadically hang or never complete requests set
2020 disable-pmtu-discovery option to 'transparent'.
2022 name= Specifies a internal name for the port. Defaults to
2023 the port specification (port or addr:port)
2025 tcpkeepalive[=idle,interval,timeout]
2026 Enable TCP keepalive probes of idle connections.
2027 In seconds; idle is the initial time before TCP starts
2028 probing the connection, interval how often to probe, and
2029 timeout the time before giving up.
2031 require-proxy-header
2032 Require PROXY protocol version 1 or 2 connections.
2033 The proxy_protocol_access is required to whitelist
2034 downstream proxies which can be trusted.
2036 If you run Squid on a dual-homed machine with an internal
2037 and an external interface we recommend you to specify the
2038 internal address:port in http_port. This way Squid will only be
2039 visible on the internal address.
2043 # Squid normally listens to port 3128
2044 http_port @DEFAULT_HTTP_PORT@
2049 IFDEF: USE_GNUTLS||USE_OPENSSL
2054 Usage: [ip:]port [mode] cert=certificate.pem [options]
2056 The socket address where Squid will listen for client requests made
2057 over TLS or SSL connections. Commonly referred to as HTTPS.
2059 This is most useful for situations where you are running squid in
2060 accelerator mode and you want to do the TLS work at the accelerator level.
2062 You may specify multiple socket addresses on multiple lines,
2063 each with their own certificate and/or options.
2065 The TLS cert= option is mandatory on HTTPS ports.
2067 See http_port for a list of modes and options.
2075 Enables Native FTP proxy by specifying the socket address where Squid
2076 listens for FTP client requests. See http_port directive for various
2077 ways to specify the listening address and mode.
2079 Usage: ftp_port address [mode] [options]
2081 WARNING: This is a new, experimental, complex feature that has seen
2082 limited production exposure. Some Squid modules (e.g., caching) do not
2083 currently work with native FTP proxying, and many features have not
2084 even been tested for compatibility. Test well before deploying!
2086 Native FTP proxying differs substantially from proxying HTTP requests
2087 with ftp:// URIs because Squid works as an FTP server and receives
2088 actual FTP commands (rather than HTTP requests with FTP URLs).
2090 Native FTP commands accepted at ftp_port are internally converted or
2091 wrapped into HTTP-like messages. The same happens to Native FTP
2092 responses received from FTP origin servers. Those HTTP-like messages
2093 are shoveled through regular access control and adaptation layers
2094 between the FTP client and the FTP origin server. This allows Squid to
2095 examine, adapt, block, and log FTP exchanges. Squid reuses most HTTP
2096 mechanisms when shoveling wrapped FTP messages. For example,
2097 http_access and adaptation_access directives are used.
2101 intercept Same as http_port intercept. The FTP origin address is
2102 determined based on the intended destination of the
2103 intercepted connection.
2105 tproxy Support Linux TPROXY for spoofing outgoing
2106 connections using the client IP address.
2107 NP: disables authentication and maybe IPv6 on the port.
2109 By default (i.e., without an explicit mode option), Squid extracts the
2110 FTP origin address from the login@origin parameter of the FTP USER
2111 command. Many popular FTP clients support such native FTP proxying.
2115 name=token Specifies an internal name for the port. Defaults to
2116 the port address. Usable with myportname ACL.
2119 Enables tracking of FTP directories by injecting extra
2120 PWD commands and adjusting Request-URI (in wrapping
2121 HTTP requests) to reflect the current FTP server
2122 directory. Tracking is disabled by default.
2124 protocol=FTP Protocol to reconstruct accelerated and intercepted
2125 requests with. Defaults to FTP. No other accepted
2126 values have been tested with. An unsupported value
2127 results in a FATAL error. Accepted values are FTP,
2128 HTTP (or HTTP/1.1), and HTTPS (or HTTPS/1.1).
2130 Other http_port modes and options that are not specific to HTTP and
2131 HTTPS may also work.
2134 NAME: tcp_outgoing_tos tcp_outgoing_ds tcp_outgoing_dscp
2137 LOC: Ip::Qos::TheConfig.tosToServer
2139 Allows you to select a TOS/Diffserv value for packets outgoing
2140 on the server side, based on an ACL.
2142 tcp_outgoing_tos ds-field [!]aclname ...
2144 Example where normal_service_net uses the TOS value 0x00
2145 and good_service_net uses 0x20
2147 acl normal_service_net src 10.0.0.0/24
2148 acl good_service_net src 10.0.1.0/24
2149 tcp_outgoing_tos 0x00 normal_service_net
2150 tcp_outgoing_tos 0x20 good_service_net
2152 TOS/DSCP values really only have local significance - so you should
2153 know what you're specifying. For more information, see RFC2474,
2154 RFC2475, and RFC3260.
2156 The TOS/DSCP byte must be exactly that - a octet value 0 - 255, or
2157 "default" to use whatever default your host has.
2158 Note that only multiples of 4 are usable as the two rightmost bits have
2159 been redefined for use by ECN (RFC 3168 section 23.1).
2160 The squid parser will enforce this by masking away the ECN bits.
2162 Processing proceeds in the order specified, and stops at first fully
2165 Only fast ACLs are supported.
2168 NAME: clientside_tos
2171 LOC: Ip::Qos::TheConfig.tosToClient
2173 Allows you to select a TOS/DSCP value for packets being transmitted
2174 on the client-side, based on an ACL.
2176 clientside_tos ds-field [!]aclname ...
2178 Example where normal_service_net uses the TOS value 0x00
2179 and good_service_net uses 0x20
2181 acl normal_service_net src 10.0.0.0/24
2182 acl good_service_net src 10.0.1.0/24
2183 clientside_tos 0x00 normal_service_net
2184 clientside_tos 0x20 good_service_net
2186 Note: This feature is incompatible with qos_flows. Any TOS values set here
2187 will be overwritten by TOS values in qos_flows.
2189 The TOS/DSCP byte must be exactly that - a octet value 0 - 255, or
2190 "default" to use whatever default your host has.
2191 Note that only multiples of 4 are usable as the two rightmost bits have
2192 been redefined for use by ECN (RFC 3168 section 23.1).
2193 The squid parser will enforce this by masking away the ECN bits.
2197 NAME: tcp_outgoing_mark
2199 IFDEF: SO_MARK&&USE_LIBCAP
2201 LOC: Ip::Qos::TheConfig.nfmarkToServer
2203 Allows you to apply a Netfilter mark value to outgoing packets
2204 on the server side, based on an ACL.
2206 tcp_outgoing_mark mark-value [!]aclname ...
2208 Example where normal_service_net uses the mark value 0x00
2209 and good_service_net uses 0x20
2211 acl normal_service_net src 10.0.0.0/24
2212 acl good_service_net src 10.0.1.0/24
2213 tcp_outgoing_mark 0x00 normal_service_net
2214 tcp_outgoing_mark 0x20 good_service_net
2216 Only fast ACLs are supported.
2219 NAME: clientside_mark
2221 IFDEF: SO_MARK&&USE_LIBCAP
2223 LOC: Ip::Qos::TheConfig.nfmarkToClient
2225 Allows you to apply a Netfilter mark value to packets being transmitted
2226 on the client-side, based on an ACL.
2228 clientside_mark mark-value [!]aclname ...
2230 Example where normal_service_net uses the mark value 0x00
2231 and good_service_net uses 0x20
2233 acl normal_service_net src 10.0.0.0/24
2234 acl good_service_net src 10.0.1.0/24
2235 clientside_mark 0x00 normal_service_net
2236 clientside_mark 0x20 good_service_net
2238 Note: This feature is incompatible with qos_flows. Any mark values set here
2239 will be overwritten by mark values in qos_flows.
2246 LOC: Ip::Qos::TheConfig
2248 Allows you to select a TOS/DSCP value to mark outgoing
2249 connections to the client, based on where the reply was sourced.
2250 For platforms using netfilter, allows you to set a netfilter mark
2251 value instead of, or in addition to, a TOS value.
2253 By default this functionality is disabled. To enable it with the default
2254 settings simply use "qos_flows mark" or "qos_flows tos". Default
2255 settings will result in the netfilter mark or TOS value being copied
2256 from the upstream connection to the client. Note that it is the connection
2257 CONNMARK value not the packet MARK value that is copied.
2259 It is not currently possible to copy the mark or TOS value from the
2260 client to the upstream connection request.
2262 TOS values really only have local significance - so you should
2263 know what you're specifying. For more information, see RFC2474,
2264 RFC2475, and RFC3260.
2266 The TOS/DSCP byte must be exactly that - a octet value 0 - 255.
2267 Note that only multiples of 4 are usable as the two rightmost bits have
2268 been redefined for use by ECN (RFC 3168 section 23.1).
2269 The squid parser will enforce this by masking away the ECN bits.
2271 Mark values can be any unsigned 32-bit integer value.
2273 This setting is configured by setting the following values:
2275 tos|mark Whether to set TOS or netfilter mark values
2277 local-hit=0xFF Value to mark local cache hits.
2279 sibling-hit=0xFF Value to mark hits from sibling peers.
2281 parent-hit=0xFF Value to mark hits from parent peers.
2283 miss=0xFF[/mask] Value to mark cache misses. Takes precedence
2284 over the preserve-miss feature (see below), unless
2285 mask is specified, in which case only the bits
2286 specified in the mask are written.
2288 The TOS variant of the following features are only possible on Linux
2289 and require your kernel to be patched with the TOS preserving ZPH
2290 patch, available from http://zph.bratcheda.org
2291 No patch is needed to preserve the netfilter mark, which will work
2292 with all variants of netfilter.
2294 disable-preserve-miss
2295 This option disables the preservation of the TOS or netfilter
2296 mark. By default, the existing TOS or netfilter mark value of
2297 the response coming from the remote server will be retained
2298 and masked with miss-mark.
2299 NOTE: in the case of a netfilter mark, the mark must be set on
2300 the connection (using the CONNMARK target) not on the packet
2304 Allows you to mask certain bits in the TOS or mark value
2305 received from the remote server, before copying the value to
2306 the TOS sent towards clients.
2307 Default for tos: 0xFF (TOS from server is not changed).
2308 Default for mark: 0xFFFFFFFF (mark from server is not changed).
2310 All of these features require the --enable-zph-qos compilation flag
2311 (enabled by default). Netfilter marking also requires the
2312 libnetfilter_conntrack libraries (--with-netfilter-conntrack) and
2313 libcap 2.09+ (--with-libcap).
2317 NAME: tcp_outgoing_address
2320 DEFAULT_DOC: Address selection is performed by the operating system.
2321 LOC: Config.accessList.outgoing_address
2323 Allows you to map requests to different outgoing IP addresses
2324 based on the username or source address of the user making
2327 tcp_outgoing_address ipaddr [[!]aclname] ...
2330 Forwarding clients with dedicated IPs for certain subnets.
2332 acl normal_service_net src 10.0.0.0/24
2333 acl good_service_net src 10.0.2.0/24
2335 tcp_outgoing_address 2001:db8::c001 good_service_net
2336 tcp_outgoing_address 10.1.0.2 good_service_net
2338 tcp_outgoing_address 2001:db8::beef normal_service_net
2339 tcp_outgoing_address 10.1.0.1 normal_service_net
2341 tcp_outgoing_address 2001:db8::1
2342 tcp_outgoing_address 10.1.0.3
2344 Processing proceeds in the order specified, and stops at first fully
2347 Squid will add an implicit IP version test to each line.
2348 Requests going to IPv4 websites will use the outgoing 10.1.0.* addresses.
2349 Requests going to IPv6 websites will use the outgoing 2001:db8:* addresses.
2352 NOTE: The use of this directive using client dependent ACLs is
2353 incompatible with the use of server side persistent connections. To
2354 ensure correct results it is best to set server_persistent_connections
2355 to off when using this directive in such configurations.
2357 NOTE: The use of this directive to set a local IP on outgoing TCP links
2358 is incompatible with using TPROXY to set client IP out outbound TCP links.
2359 When needing to contact peers use the no-tproxy cache_peer option and the
2360 client_dst_passthru directive re-enable normal forwarding such as this.
2364 NAME: host_verify_strict
2367 LOC: Config.onoff.hostStrictVerify
2369 Regardless of this option setting, when dealing with intercepted
2370 traffic, Squid always verifies that the destination IP address matches
2371 the Host header domain or IP (called 'authority form URL').
2373 This enforcement is performed to satisfy a MUST-level requirement in
2374 RFC 2616 section 14.23: "The Host field value MUST represent the naming
2375 authority of the origin server or gateway given by the original URL".
2378 Squid always responds with an HTTP 409 (Conflict) error
2379 page and logs a security warning if there is no match.
2381 Squid verifies that the destination IP address matches
2382 the Host header for forward-proxy and reverse-proxy traffic
2383 as well. For those traffic types, Squid also enables the
2384 following checks, comparing the corresponding Host header
2385 and Request-URI components:
2387 * The host names (domain or IP) must be identical,
2388 but valueless or missing Host header disables all checks.
2389 For the two host names to match, both must be either IP
2392 * Port numbers must be identical, but if a port is missing
2393 the scheme-default port is assumed.
2396 When set to OFF (the default):
2397 Squid allows suspicious requests to continue but logs a
2398 security warning and blocks caching of the response.
2400 * Forward-proxy traffic is not checked at all.
2402 * Reverse-proxy traffic is not checked at all.
2404 * Intercepted traffic which passes verification is handled
2405 according to client_dst_passthru.
2407 * Intercepted requests which fail verification are sent
2408 to the client original destination instead of DIRECT.
2409 This overrides 'client_dst_passthru off'.
2411 For now suspicious intercepted CONNECT requests are always
2412 responded to with an HTTP 409 (Conflict) error page.
2417 As described in CVE-2009-0801 when the Host: header alone is used
2418 to determine the destination of a request it becomes trivial for
2419 malicious scripts on remote websites to bypass browser same-origin
2420 security policy and sandboxing protections.
2422 The cause of this is that such applets are allowed to perform their
2423 own HTTP stack, in which case the same-origin policy of the browser
2424 sandbox only verifies that the applet tries to contact the same IP
2425 as from where it was loaded at the IP level. The Host: header may
2426 be different from the connected IP and approved origin.
2430 NAME: client_dst_passthru
2433 LOC: Config.onoff.client_dst_passthru
2435 With NAT or TPROXY intercepted traffic Squid may pass the request
2436 directly to the original client destination IP or seek a faster
2437 source using the HTTP Host header.
2439 Using Host to locate alternative servers can provide faster
2440 connectivity with a range of failure recovery options.
2441 But can also lead to connectivity trouble when the client and
2442 server are attempting stateful interactions unaware of the proxy.
2444 This option (on by default) prevents alternative DNS entries being
2445 located to send intercepted traffic DIRECT to an origin server.
2446 The clients original destination IP and port will be used instead.
2448 Regardless of this option setting, when dealing with intercepted
2449 traffic Squid will verify the Host: header and any traffic which
2450 fails Host verification will be treated as if this option were ON.
2452 see host_verify_strict for details on the verification process.
2457 -----------------------------------------------------------------------------
2460 NAME: tls_outgoing_options
2461 IFDEF: USE_GNUTLS||USE_OPENSSL
2462 TYPE: securePeerOptions
2463 DEFAULT: min-version=1.0
2464 LOC: Security::ProxyOutgoingConfig
2466 disable Do not support https:// URLs.
2468 cert=/path/to/client/certificate
2469 A client TLS certificate to use when connecting.
2471 key=/path/to/client/private_key
2472 The private TLS key corresponding to the cert= above.
2473 If key= is not specified cert= is assumed to reference
2474 a PEM file containing both the certificate and the key.
2476 cipher=... The list of valid TLS ciphers to use.
2479 The minimum TLS protocol version to permit.
2480 To control SSLv3 use the options= parameter.
2481 Supported Values: 1.0 (default), 1.1, 1.2
2483 options=... Specify various TLS/SSL implementation options:
2485 NO_SSLv3 Disallow the use of SSLv3
2487 NO_TLSv1 Disallow the use of TLSv1.0
2489 NO_TLSv1_1 Disallow the use of TLSv1.1
2491 NO_TLSv1_2 Disallow the use of TLSv1.2
2494 Always create a new key when using
2495 temporary/ephemeral DH key exchanges
2498 Disable use of RFC5077 session tickets.
2499 Some servers may have problems
2500 understanding the TLS extension due
2501 to ambiguous specification in RFC4507.
2503 ALL Enable various bug workarounds
2504 suggested as "harmless" by OpenSSL
2505 Be warned that this reduces SSL/TLS
2506 strength to some attacks.
2508 See the OpenSSL SSL_CTX_set_options documentation for a
2511 cafile= PEM file containing CA certificates to use when verifying
2512 the peer certificate. May be repeated to load multiple files.
2514 capath= A directory containing additional CA certificates to
2515 use when verifying the peer certificate.
2516 Requires OpenSSL or LibreSSL.
2518 crlfile=... A certificate revocation list file to use when
2519 verifying the peer certificate.
2521 flags=... Specify various flags modifying the TLS implementation:
2524 Accept certificates even if they fail to
2527 Don't verify the peer certificate
2528 matches the server name
2530 no-default-ca Do not use the system default Trusted CA.
2532 domain= The peer name as advertised in its certificate.
2533 Used for verifying the correctness of the received peer
2534 certificate. If not specified the peer hostname will be
2540 -----------------------------------------------------------------------------
2543 NAME: ssl_unclean_shutdown
2547 LOC: Config.SSL.unclean_shutdown
2549 Some browsers (especially MSIE) bugs out on SSL shutdown
2556 LOC: Config.SSL.ssl_engine
2559 The OpenSSL engine to use. You will need to set this if you
2560 would like to use hardware SSL acceleration for example.
2563 NAME: sslproxy_session_ttl
2566 LOC: Config.SSL.session_ttl
2569 Sets the timeout value for SSL sessions
2572 NAME: sslproxy_session_cache_size
2575 LOC: Config.SSL.sessionCacheSize
2578 Sets the cache size to use for ssl session
2581 NAME: sslproxy_foreign_intermediate_certs
2584 LOC: Config.ssl_client.foreignIntermediateCertsPath
2587 Many origin servers fail to send their full server certificate
2588 chain for verification, assuming the client already has or can
2589 easily locate any missing intermediate certificates.
2591 Squid uses the certificates from the specified file to fill in
2592 these missing chains when trying to validate origin server
2595 The file is expected to contain zero or more PEM-encoded
2596 intermediate certificates. These certificates are not treated
2597 as trusted root certificates, and any self-signed certificate in
2598 this file will be ignored.
2600 This directive may be repeated to load multiple files.
2603 NAME: sslproxy_cert_sign_hash
2606 LOC: Config.SSL.certSignHash
2609 Sets the hashing algorithm to use when signing generated certificates.
2610 Valid algorithm names depend on the OpenSSL library used. The following
2611 names are usually available: sha1, sha256, sha512, and md5. Please see
2612 your OpenSSL library manual for the available hashes. By default, Squids
2613 that support this option use sha256 hashes.
2615 Squid does not forcefully purge cached certificates that were generated
2616 with an algorithm other than the currently configured one. They remain
2617 in the cache, subject to the regular cache eviction policy, and become
2618 useful if the algorithm changes again.
2623 TYPE: sslproxy_ssl_bump
2624 LOC: Config.accessList.ssl_bump
2625 DEFAULT_DOC: Become a TCP tunnel without decrypting proxied traffic.
2628 This option is consulted when a CONNECT request is received on
2629 an http_port (or a new connection is intercepted at an
2630 https_port), provided that port was configured with an ssl-bump
2631 flag. The subsequent data on the connection is either treated as
2632 HTTPS and decrypted OR tunneled at TCP level without decryption,
2633 depending on the first matching bumping "action".
2635 ssl_bump <action> [!]acl ...
2637 The following bumping actions are currently supported:
2640 Become a TCP tunnel without decrypting proxied traffic.
2641 This is the default action.
2644 Establish a secure connection with the server and, using a
2645 mimicked server certificate, with the client.
2648 Receive client (step SslBump1) or server (step SslBump2)
2649 certificate while preserving the possibility of splicing the
2650 connection. Peeking at the server certificate (during step 2)
2651 usually precludes bumping of the connection at step 3.
2654 Receive client (step SslBump1) or server (step SslBump2)
2655 certificate while preserving the possibility of bumping the
2656 connection. Staring at the server certificate (during step 2)
2657 usually precludes splicing of the connection at step 3.
2660 Close client and server connections.
2662 Backward compatibility actions available at step SslBump1:
2665 Bump the connection. Establish a secure connection with the
2666 client first, then connect to the server. This old mode does
2667 not allow Squid to mimic server SSL certificate and does not
2668 work with intercepted SSL connections.
2671 Bump the connection. Establish a secure connection with the
2672 server first, then establish a secure connection with the
2673 client, using a mimicked server certificate. Works with both
2674 CONNECT requests and intercepted SSL connections, but does
2675 not allow to make decisions based on SSL handshake info.
2678 Decide whether to bump or splice the connection based on
2679 client-to-squid and server-to-squid SSL hello messages.
2683 Same as the "splice" action.
2685 All ssl_bump rules are evaluated at each of the supported bumping
2686 steps. Rules with actions that are impossible at the current step are
2687 ignored. The first matching ssl_bump action wins and is applied at the
2688 end of the current step. If no rules match, the splice action is used.
2689 See the at_step ACL for a list of the supported SslBump steps.
2691 This clause supports both fast and slow acl types.
2692 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
2694 See also: http_port ssl-bump, https_port ssl-bump, and acl at_step.
2697 # Example: Bump all requests except those originating from
2698 # localhost or those going to example.com.
2700 acl broken_sites dstdomain .example.com
2701 ssl_bump splice localhost
2702 ssl_bump splice broken_sites
2706 NAME: sslproxy_cert_error
2709 DEFAULT_DOC: Server certificate errors terminate the transaction.
2710 LOC: Config.ssl_client.cert_error
2713 Use this ACL to bypass server certificate validation errors.
2715 For example, the following lines will bypass all validation errors
2716 when talking to servers for example.com. All other
2717 validation errors will result in ERR_SECURE_CONNECT_FAIL error.
2719 acl BrokenButTrustedServers dstdomain example.com
2720 sslproxy_cert_error allow BrokenButTrustedServers
2721 sslproxy_cert_error deny all
2723 This clause only supports fast acl types.
2724 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
2725 Using slow acl types may result in server crashes
2727 Without this option, all server certificate validation errors
2728 terminate the transaction to protect Squid and the client.
2730 SQUID_X509_V_ERR_INFINITE_VALIDATION error cannot be bypassed
2731 but should not happen unless your OpenSSL library is buggy.
2734 Bypassing validation errors is dangerous because an
2735 error usually implies that the server cannot be trusted
2736 and the connection may be insecure.
2738 See also: sslproxy_flags and DONT_VERIFY_PEER.
2741 NAME: sslproxy_cert_sign
2744 POSTSCRIPTUM: signUntrusted ssl::certUntrusted
2745 POSTSCRIPTUM: signSelf ssl::certSelfSigned
2746 POSTSCRIPTUM: signTrusted all
2747 TYPE: sslproxy_cert_sign
2748 LOC: Config.ssl_client.cert_sign
2751 sslproxy_cert_sign <signing algorithm> acl ...
2753 The following certificate signing algorithms are supported:
2756 Sign using the configured CA certificate which is usually
2757 placed in and trusted by end-user browsers. This is the
2758 default for trusted origin server certificates.
2761 Sign to guarantee an X509_V_ERR_CERT_UNTRUSTED browser error.
2762 This is the default for untrusted origin server certificates
2763 that are not self-signed (see ssl::certUntrusted).
2766 Sign using a self-signed certificate with the right CN to
2767 generate a X509_V_ERR_DEPTH_ZERO_SELF_SIGNED_CERT error in the
2768 browser. This is the default for self-signed origin server
2769 certificates (see ssl::certSelfSigned).
2771 This clause only supports fast acl types.
2773 When sslproxy_cert_sign acl(s) match, Squid uses the corresponding
2774 signing algorithm to generate the certificate and ignores all
2775 subsequent sslproxy_cert_sign options (the first match wins). If no
2776 acl(s) match, the default signing algorithm is determined by errors
2777 detected when obtaining and validating the origin server certificate.
2779 WARNING: SQUID_X509_V_ERR_DOMAIN_MISMATCH and ssl:certDomainMismatch can
2780 be used with sslproxy_cert_adapt, but if and only if Squid is bumping a
2781 CONNECT request that carries a domain name. In all other cases (CONNECT
2782 to an IP address or an intercepted SSL connection), Squid cannot detect
2783 the domain mismatch at certificate generation time when
2784 bump-server-first is used.
2787 NAME: sslproxy_cert_adapt
2790 TYPE: sslproxy_cert_adapt
2791 LOC: Config.ssl_client.cert_adapt
2794 sslproxy_cert_adapt <adaptation algorithm> acl ...
2796 The following certificate adaptation algorithms are supported:
2799 Sets the "Not After" property to the "Not After" property of
2800 the CA certificate used to sign generated certificates.
2803 Sets the "Not Before" property to the "Not Before" property of
2804 the CA certificate used to sign generated certificates.
2806 setCommonName or setCommonName{CN}
2807 Sets Subject.CN property to the host name specified as a
2808 CN parameter or, if no explicit CN parameter was specified,
2809 extracted from the CONNECT request. It is a misconfiguration
2810 to use setCommonName without an explicit parameter for
2811 intercepted or tproxied SSL connections.
2813 This clause only supports fast acl types.
2815 Squid first groups sslproxy_cert_adapt options by adaptation algorithm.
2816 Within a group, when sslproxy_cert_adapt acl(s) match, Squid uses the
2817 corresponding adaptation algorithm to generate the certificate and
2818 ignores all subsequent sslproxy_cert_adapt options in that algorithm's
2819 group (i.e., the first match wins within each algorithm group). If no
2820 acl(s) match, the default mimicking action takes place.
2822 WARNING: SQUID_X509_V_ERR_DOMAIN_MISMATCH and ssl:certDomainMismatch can
2823 be used with sslproxy_cert_adapt, but if and only if Squid is bumping a
2824 CONNECT request that carries a domain name. In all other cases (CONNECT
2825 to an IP address or an intercepted SSL connection), Squid cannot detect
2826 the domain mismatch at certificate generation time when
2827 bump-server-first is used.
2830 NAME: sslpassword_program
2833 LOC: Config.Program.ssl_password
2836 Specify a program used for entering SSL key passphrases
2837 when using encrypted SSL certificate keys. If not specified
2838 keys must either be unencrypted, or Squid started with the -N
2839 option to allow it to query interactively for the passphrase.
2841 The key file name is given as argument to the program allowing
2842 selection of the right password if you have multiple encrypted
2847 OPTIONS RELATING TO EXTERNAL SSL_CRTD
2848 -----------------------------------------------------------------------------
2851 NAME: sslcrtd_program
2854 DEFAULT: @DEFAULT_SSL_CRTD@ -s @DEFAULT_SSL_DB_DIR@ -M 4MB
2855 LOC: Ssl::TheConfig.ssl_crtd
2857 Specify the location and options of the executable for ssl_crtd process.
2858 @DEFAULT_SSL_CRTD@ program requires -s and -M parameters
2859 For more information use:
2860 @DEFAULT_SSL_CRTD@ -h
2863 NAME: sslcrtd_children
2864 TYPE: HelperChildConfig
2866 DEFAULT: 32 startup=5 idle=1
2867 LOC: Ssl::TheConfig.ssl_crtdChildren
2869 The maximum number of processes spawn to service ssl server.
2870 The maximum this may be safely set to is 32.
2872 The startup= and idle= options allow some measure of skew in your
2877 Sets the minimum number of processes to spawn when Squid
2878 starts or reconfigures. When set to zero the first request will
2879 cause spawning of the first child process to handle it.
2881 Starting too few children temporary slows Squid under load while it
2882 tries to spawn enough additional processes to cope with traffic.
2886 Sets a minimum of how many processes Squid is to try and keep available
2887 at all times. When traffic begins to rise above what the existing
2888 processes can handle this many more will be spawned up to the maximum
2889 configured. A minimum setting of 1 is required.
2893 Sets the maximum number of queued requests.
2894 If the queued requests exceed queue size for more than 3 minutes
2895 squid aborts its operation.
2896 The default value is set to 2*numberofchildren.
2898 You must have at least one ssl_crtd process.
2901 NAME: sslcrtvalidator_program
2905 LOC: Ssl::TheConfig.ssl_crt_validator
2907 Specify the location and options of the executable for ssl_crt_validator
2910 Usage: sslcrtvalidator_program [ttl=n] [cache=n] path ...
2913 ttl=n TTL in seconds for cached results. The default is 60 secs
2914 cache=n limit the result cache size. The default value is 2048
2917 NAME: sslcrtvalidator_children
2918 TYPE: HelperChildConfig
2920 DEFAULT: 32 startup=5 idle=1 concurrency=1
2921 LOC: Ssl::TheConfig.ssl_crt_validator_Children
2923 The maximum number of processes spawn to service SSL server.
2924 The maximum this may be safely set to is 32.
2926 The startup= and idle= options allow some measure of skew in your
2931 Sets the minimum number of processes to spawn when Squid
2932 starts or reconfigures. When set to zero the first request will
2933 cause spawning of the first child process to handle it.
2935 Starting too few children temporary slows Squid under load while it
2936 tries to spawn enough additional processes to cope with traffic.
2940 Sets a minimum of how many processes Squid is to try and keep available
2941 at all times. When traffic begins to rise above what the existing
2942 processes can handle this many more will be spawned up to the maximum
2943 configured. A minimum setting of 1 is required.
2947 The number of requests each certificate validator helper can handle in
2948 parallel. A value of 0 indicates the certficate validator does not
2949 support concurrency. Defaults to 1.
2951 When this directive is set to a value >= 1 then the protocol
2952 used to communicate with the helper is modified to include
2953 a request ID in front of the request/response. The request
2954 ID from the request must be echoed back with the response
2959 Sets the maximum number of queued requests.
2960 If the queued requests exceed queue size for more than 3 minutes
2961 squid aborts its operation.
2962 The default value is set to 2*numberofchildren.
2964 You must have at least one ssl_crt_validator process.
2968 OPTIONS WHICH AFFECT THE NEIGHBOR SELECTION ALGORITHM
2969 -----------------------------------------------------------------------------
2977 To specify other caches in a hierarchy, use the format:
2979 cache_peer hostname type http-port icp-port [options]
2984 # hostname type port port options
2985 # -------------------- -------- ----- ----- -----------
2986 cache_peer parent.foo.net parent 3128 3130 default
2987 cache_peer sib1.foo.net sibling 3128 3130 proxy-only
2988 cache_peer sib2.foo.net sibling 3128 3130 proxy-only
2989 cache_peer example.com parent 80 0 default
2990 cache_peer cdn.example.com sibling 3128 0
2992 type: either 'parent', 'sibling', or 'multicast'.
2994 proxy-port: The port number where the peer accept HTTP requests.
2995 For other Squid proxies this is usually 3128
2996 For web servers this is usually 80
2998 icp-port: Used for querying neighbor caches about objects.
2999 Set to 0 if the peer does not support ICP or HTCP.
3000 See ICP and HTCP options below for additional details.
3003 ==== ICP OPTIONS ====
3005 You MUST also set icp_port and icp_access explicitly when using these options.
3006 The defaults will prevent peer traffic using ICP.
3009 no-query Disable ICP queries to this neighbor.
3012 Indicates the named peer is a member of a multicast group.
3013 ICP queries will not be sent directly to the peer, but ICP
3014 replies will be accepted from it.
3016 closest-only Indicates that, for ICP_OP_MISS replies, we'll only forward
3017 CLOSEST_PARENT_MISSes and never FIRST_PARENT_MISSes.
3020 To only send ICP queries to this neighbor infrequently.
3021 This is used to keep the neighbor round trip time updated
3022 and is usually used in conjunction with weighted-round-robin.
3025 ==== HTCP OPTIONS ====
3027 You MUST also set htcp_port and htcp_access explicitly when using these options.
3028 The defaults will prevent peer traffic using HTCP.
3031 htcp Send HTCP, instead of ICP, queries to the neighbor.
3032 You probably also want to set the "icp-port" to 4827
3033 instead of 3130. This directive accepts a comma separated
3034 list of options described below.
3036 htcp=oldsquid Send HTCP to old Squid versions (2.5 or earlier).
3038 htcp=no-clr Send HTCP to the neighbor but without
3039 sending any CLR requests. This cannot be used with
3042 htcp=only-clr Send HTCP to the neighbor but ONLY CLR requests.
3043 This cannot be used with no-clr.
3046 Send HTCP to the neighbor including CLRs but only when
3047 they do not result from PURGE requests.
3050 Forward any HTCP CLR requests this proxy receives to the peer.
3053 ==== PEER SELECTION METHODS ====
3055 The default peer selection method is ICP, with the first responding peer
3056 being used as source. These options can be used for better load balancing.
3059 default This is a parent cache which can be used as a "last-resort"
3060 if a peer cannot be located by any of the peer-selection methods.
3061 If specified more than once, only the first is used.
3063 round-robin Load-Balance parents which should be used in a round-robin
3064 fashion in the absence of any ICP queries.
3065 weight=N can be used to add bias.
3067 weighted-round-robin
3068 Load-Balance parents which should be used in a round-robin
3069 fashion with the frequency of each parent being based on the
3070 round trip time. Closer parents are used more often.
3071 Usually used for background-ping parents.
3072 weight=N can be used to add bias.
3074 carp Load-Balance parents which should be used as a CARP array.
3075 The requests will be distributed among the parents based on the
3076 CARP load balancing hash function based on their weight.
3078 userhash Load-balance parents based on the client proxy_auth or ident username.
3080 sourcehash Load-balance parents based on the client source IP.
3083 To be used only for cache peers of type "multicast".
3084 ALL members of this multicast group have "sibling"
3085 relationship with it, not "parent". This is to a multicast
3086 group when the requested object would be fetched only from
3087 a "parent" cache, anyway. It's useful, e.g., when
3088 configuring a pool of redundant Squid proxies, being
3089 members of the same multicast group.
3092 ==== PEER SELECTION OPTIONS ====
3094 weight=N use to affect the selection of a peer during any weighted
3095 peer-selection mechanisms.
3096 The weight must be an integer; default is 1,
3097 larger weights are favored more.
3098 This option does not affect parent selection if a peering
3099 protocol is not in use.
3101 basetime=N Specify a base amount to be subtracted from round trip
3103 It is subtracted before division by weight in calculating
3104 which parent to fectch from. If the rtt is less than the
3105 base time the rtt is set to a minimal value.
3107 ttl=N Specify a TTL to use when sending multicast ICP queries
3109 Only useful when sending to a multicast group.
3110 Because we don't accept ICP replies from random
3111 hosts, you must configure other group members as
3112 peers with the 'multicast-responder' option.
3114 no-delay To prevent access to this neighbor from influencing the
3117 digest-url=URL Tell Squid to fetch the cache digest (if digests are
3118 enabled) for this host from the specified URL rather
3119 than the Squid default location.
3122 ==== CARP OPTIONS ====
3124 carp-key=key-specification
3125 use a different key than the full URL to hash against the peer.
3126 the key-specification is a comma-separated list of the keywords
3127 scheme, host, port, path, params
3128 Order is not important.
3130 ==== ACCELERATOR / REVERSE-PROXY OPTIONS ====
3132 originserver Causes this parent to be contacted as an origin server.
3133 Meant to be used in accelerator setups when the peer
3137 Set the Host header of requests forwarded to this peer.
3138 Useful in accelerator setups where the server (peer)
3139 expects a certain domain name but clients may request
3140 others. ie example.com or www.example.com
3142 no-digest Disable request of cache digests.
3145 Disables requesting ICMP RTT database (NetDB).
3148 ==== AUTHENTICATION OPTIONS ====
3151 If this is a personal/workgroup proxy and your parent
3152 requires proxy authentication.
3154 Note: The string can include URL escapes (i.e. %20 for
3155 spaces). This also means % must be written as %%.
3158 Send login details received from client to this peer.
3159 Both Proxy- and WWW-Authorization headers are passed
3160 without alteration to the peer.
3161 Authentication is not required by Squid for this to work.
3163 Note: This will pass any form of authentication but
3164 only Basic auth will work through a proxy unless the
3165 connection-auth options are also used.
3167 login=PASS Send login details received from client to this peer.
3168 Authentication is not required by this option.
3170 If there are no client-provided authentication headers
3171 to pass on, but username and password are available
3172 from an external ACL user= and password= result tags
3173 they may be sent instead.
3175 Note: To combine this with proxy_auth both proxies must
3176 share the same user database as HTTP only allows for
3177 a single login (one for proxy, one for origin server).
3178 Also be warned this will expose your users proxy
3179 password to the peer. USE WITH CAUTION
3182 Send the username to the upstream cache, but with a
3183 fixed password. This is meant to be used when the peer
3184 is in another administrative domain, but it is still
3185 needed to identify each user.
3186 The star can optionally be followed by some extra
3187 information which is added to the username. This can
3188 be used to identify this proxy to the peer, similar to
3189 the login=username:password option above.
3192 If this is a personal/workgroup proxy and your parent
3193 requires a secure proxy authentication.
3194 The first principal from the default keytab or defined by
3195 the environment variable KRB5_KTNAME will be used.
3197 WARNING: The connection may transmit requests from multiple
3198 clients. Negotiate often assumes end-to-end authentication
3199 and a single-client. Which is not strictly true here.
3201 login=NEGOTIATE:principal_name
3202 If this is a personal/workgroup proxy and your parent
3203 requires a secure proxy authentication.
3204 The principal principal_name from the default keytab or
3205 defined by the environment variable KRB5_KTNAME will be
3208 WARNING: The connection may transmit requests from multiple
3209 clients. Negotiate often assumes end-to-end authentication
3210 and a single-client. Which is not strictly true here.
3212 connection-auth=on|off
3213 Tell Squid that this peer does or not support Microsoft
3214 connection oriented authentication, and any such
3215 challenges received from there should be ignored.
3216 Default is auto to automatically determine the status
3220 Do not use a keytab to authenticate to a peer when
3221 login=NEGOTIATE is specified. Let the GSSAPI
3222 implementation determine which already existing
3223 credentials cache to use instead.
3226 ==== SSL / HTTPS / TLS OPTIONS ====
3228 ssl Encrypt connections to this peer with SSL/TLS.
3230 sslcert=/path/to/ssl/certificate
3231 A client SSL certificate to use when connecting to
3234 sslkey=/path/to/ssl/key
3235 The private SSL key corresponding to sslcert above.
3236 If 'sslkey' is not specified 'sslcert' is assumed to
3237 reference a combined file containing both the
3238 certificate and the key.
3240 sslcipher=... The list of valid SSL ciphers to use when connecting
3244 The minimum TLS protocol version to permit. To control
3245 SSLv3 use the ssloptions= parameter.
3246 Supported Values: 1.0 (default), 1.1, 1.2
3248 ssloptions=... Specify various SSL implementation options:
3250 NO_SSLv3 Disallow the use of SSLv3
3252 NO_TLSv1 Disallow the use of TLSv1.0
3254 NO_TLSv1_1 Disallow the use of TLSv1.1
3256 NO_TLSv1_2 Disallow the use of TLSv1.2
3259 Always create a new key when using
3260 temporary/ephemeral DH key exchanges
3263 Disable use of RFC5077 session tickets.
3264 Some servers may have problems
3265 understanding the TLS extension due
3266 to ambiguous specification in RFC4507.
3268 ALL Enable various bug workarounds
3269 suggested as "harmless" by OpenSSL
3270 Be warned that this reduces SSL/TLS
3271 strength to some attacks.
3273 See the OpenSSL SSL_CTX_set_options documentation for a
3276 tls-cafile= PEM file containing CA certificates to use when verifying
3277 the peer certificate. May be repeated to load multiple files.
3279 sslcapath=... A directory containing additional CA certificates to
3280 use when verifying the peer certificate.
3281 Requires OpenSSL or LibreSSL.
3283 sslcrlfile=... A certificate revocation list file to use when
3284 verifying the peer certificate.
3286 sslflags=... Specify various flags modifying the SSL implementation:
3289 Accept certificates even if they fail to
3293 Don't verify the peer certificate
3294 matches the server name
3296 ssldomain= The peer name as advertised in it's certificate.
3297 Used for verifying the correctness of the received peer
3298 certificate. If not specified the peer hostname will be
3302 Enable the "Front-End-Https: On" header needed when
3303 using Squid as a SSL frontend in front of Microsoft OWA.
3304 See MS KB document Q307347 for details on this header.
3305 If set to auto the header will only be added if the
3306 request is forwarded as a https:// URL.
3309 Do not use the system default Trusted CA.
3311 tls-no-npn Do not use the TLS NPN extension to advertise HTTP/1.1.
3313 ==== GENERAL OPTIONS ====
3316 A peer-specific connect timeout.
3317 Also see the peer_connect_timeout directive.
3319 connect-fail-limit=N
3320 How many times connecting to a peer must fail before
3321 it is marked as down. Standby connection failures
3322 count towards this limit. Default is 10.
3324 allow-miss Disable Squid's use of only-if-cached when forwarding
3325 requests to siblings. This is primarily useful when
3326 icp_hit_stale is used by the sibling. Excessive use
3327 of this option may result in forwarding loops. One way
3328 to prevent peering loops when using this option, is to
3329 deny cache peer usage on requests from a peer:
3331 cache_peer_access peerName deny fromPeer
3333 max-conn=N Limit the number of concurrent connections the Squid
3334 may open to this peer, including already opened idle
3335 and standby connections. There is no peer-specific
3336 connection limit by default.
3338 A peer exceeding the limit is not used for new
3339 requests unless a standby connection is available.
3341 max-conn currently works poorly with idle persistent
3342 connections: When a peer reaches its max-conn limit,
3343 and there are idle persistent connections to the peer,
3344 the peer may not be selected because the limiting code
3345 does not know whether Squid can reuse those idle
3348 standby=N Maintain a pool of N "hot standby" connections to an
3349 UP peer, available for requests when no idle
3350 persistent connection is available (or safe) to use.
3351 By default and with zero N, no such pool is maintained.
3352 N must not exceed the max-conn limit (if any).
3354 At start or after reconfiguration, Squid opens new TCP
3355 standby connections until there are N connections
3356 available and then replenishes the standby pool as
3357 opened connections are used up for requests. A used
3358 connection never goes back to the standby pool, but
3359 may go to the regular idle persistent connection pool
3360 shared by all peers and origin servers.
3362 Squid never opens multiple new standby connections
3363 concurrently. This one-at-a-time approach minimizes
3364 flooding-like effect on peers. Furthermore, just a few
3365 standby connections should be sufficient in most cases
3366 to supply most new requests with a ready-to-use
3369 Standby connections obey server_idle_pconn_timeout.
3370 For the feature to work as intended, the peer must be
3371 configured to accept and keep them open longer than
3372 the idle timeout at the connecting Squid, to minimize
3373 race conditions typical to idle used persistent
3374 connections. Default request_timeout and
3375 server_idle_pconn_timeout values ensure such a
3378 name=xxx Unique name for the peer.
3379 Required if you have multiple peers on the same host
3380 but different ports.
3381 This name can be used in cache_peer_access and similar
3382 directives to identify the peer.
3383 Can be used by outgoing access controls through the
3386 no-tproxy Do not use the client-spoof TPROXY support when forwarding
3387 requests to this peer. Use normal address selection instead.
3388 This overrides the spoof_client_ip ACL.
3390 proxy-only objects fetched from the peer will not be stored locally.
3394 NAME: cache_peer_access
3397 DEFAULT_DOC: No peer usage restrictions.
3400 Restricts usage of cache_peer proxies.
3403 cache_peer_access peer-name allow|deny [!]aclname ...
3405 For the required peer-name parameter, use either the value of the
3406 cache_peer name=value parameter or, if name=value is missing, the
3407 cache_peer hostname parameter.
3409 This directive narrows down the selection of peering candidates, but
3410 does not determine the order in which the selected candidates are
3411 contacted. That order is determined by the peer selection algorithms
3412 (see PEER SELECTION sections in the cache_peer documentation).
3414 If a deny rule matches, the corresponding peer will not be contacted
3415 for the current transaction -- Squid will not send ICP queries and
3416 will not forward HTTP requests to that peer. An allow match leaves
3417 the corresponding peer in the selection. The first match for a given
3418 peer wins for that peer.
3420 The relative order of cache_peer_access directives for the same peer
3421 matters. The relative order of any two cache_peer_access directives
3422 for different peers does not matter. To ease interpretation, it is a
3423 good idea to group cache_peer_access directives for the same peer
3426 A single cache_peer_access directive may be evaluated multiple times
3427 for a given transaction because individual peer selection algorithms
3428 may check it independently from each other. These redundant checks
3429 may be optimized away in future Squid versions.
3431 This clause only supports fast acl types.
3432 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
3436 NAME: neighbor_type_domain
3437 TYPE: hostdomaintype
3439 DEFAULT_DOC: The peer type from cache_peer directive is used for all requests to that peer.
3442 Modify the cache_peer neighbor type when passing requests
3443 about specific domains to the peer.
3446 neighbor_type_domain neighbor parent|sibling domain domain ...
3449 cache_peer foo.example.com parent 3128 3130
3450 neighbor_type_domain foo.example.com sibling .au .de
3452 The above configuration treats all requests to foo.example.com as a
3453 parent proxy unless the request is for a .au or .de ccTLD domain name.
3456 NAME: dead_peer_timeout
3460 LOC: Config.Timeout.deadPeer
3462 This controls how long Squid waits to declare a peer cache
3463 as "dead." If there are no ICP replies received in this
3464 amount of time, Squid will declare the peer dead and not
3465 expect to receive any further ICP replies. However, it
3466 continues to send ICP queries, and will mark the peer as
3467 alive upon receipt of the first subsequent ICP reply.
3469 This timeout also affects when Squid expects to receive ICP
3470 replies from peers. If more than 'dead_peer' seconds have
3471 passed since the last ICP reply was received, Squid will not
3472 expect to receive an ICP reply on the next query. Thus, if
3473 your time between requests is greater than this timeout, you
3474 will see a lot of requests sent DIRECT to origin servers
3475 instead of to your parents.
3478 NAME: forward_max_tries
3481 LOC: Config.forward_max_tries
3483 Controls how many different forward paths Squid will try
3484 before giving up. See also forward_timeout.
3486 NOTE: connect_retries (default: none) can make each of these
3487 possible forwarding paths be tried multiple times.
3491 MEMORY CACHE OPTIONS
3492 -----------------------------------------------------------------------------
3499 LOC: Config.memMaxSize
3501 NOTE: THIS PARAMETER DOES NOT SPECIFY THE MAXIMUM PROCESS SIZE.
3502 IT ONLY PLACES A LIMIT ON HOW MUCH ADDITIONAL MEMORY SQUID WILL
3503 USE AS A MEMORY CACHE OF OBJECTS. SQUID USES MEMORY FOR OTHER
3504 THINGS AS WELL. SEE THE SQUID FAQ SECTION 8 FOR DETAILS.
3506 'cache_mem' specifies the ideal amount of memory to be used
3508 * In-Transit objects
3510 * Negative-Cached objects
3512 Data for these objects are stored in 4 KB blocks. This
3513 parameter specifies the ideal upper limit on the total size of
3514 4 KB blocks allocated. In-Transit objects take the highest
3517 In-transit objects have priority over the others. When
3518 additional space is needed for incoming data, negative-cached
3519 and hot objects will be released. In other words, the
3520 negative-cached and hot objects will fill up any unused space
3521 not needed for in-transit objects.
3523 If circumstances require, this limit will be exceeded.
3524 Specifically, if your incoming request rate requires more than
3525 'cache_mem' of memory to hold in-transit objects, Squid will
3526 exceed this limit to satisfy the new requests. When the load
3527 decreases, blocks will be freed until the high-water mark is
3528 reached. Thereafter, blocks will be used to store hot
3531 If shared memory caching is enabled, Squid does not use the shared
3532 cache space for in-transit objects, but they still consume as much
3533 local memory as they need. For more details about the shared memory
3534 cache, see memory_cache_shared.
3537 NAME: maximum_object_size_in_memory
3541 LOC: Config.Store.maxInMemObjSize
3543 Objects greater than this size will not be attempted to kept in
3544 the memory cache. This should be set high enough to keep objects
3545 accessed frequently in memory to improve performance whilst low
3546 enough to keep larger objects from hoarding cache_mem.
3549 NAME: memory_cache_shared
3552 LOC: Config.memShared
3554 DEFAULT_DOC: "on" where supported if doing memory caching with multiple SMP workers.
3556 Controls whether the memory cache is shared among SMP workers.
3558 The shared memory cache is meant to occupy cache_mem bytes and replace
3559 the non-shared memory cache, although some entities may still be
3560 cached locally by workers for now (e.g., internal and in-transit
3561 objects may be served from a local memory cache even if shared memory
3562 caching is enabled).
3564 By default, the memory cache is shared if and only if all of the
3565 following conditions are satisfied: Squid runs in SMP mode with
3566 multiple workers, cache_mem is positive, and Squid environment
3567 supports required IPC primitives (e.g., POSIX shared memory segments
3568 and GCC-style atomic operations).
3570 To avoid blocking locks, shared memory uses opportunistic algorithms
3571 that do not guarantee that every cachable entity that could have been
3572 shared among SMP workers will actually be shared.
3574 Currently, entities exceeding 32KB in size cannot be shared.
3577 NAME: memory_cache_mode
3581 DEFAULT_DOC: Keep the most recently fetched objects in memory
3583 Controls which objects to keep in the memory cache (cache_mem)
3585 always Keep most recently fetched objects in memory (default)
3587 disk Only disk cache hits are kept in memory, which means
3588 an object must first be cached on disk and then hit
3589 a second time before cached in memory.
3591 network Only objects fetched from network is kept in memory
3594 NAME: memory_replacement_policy
3596 LOC: Config.memPolicy
3599 The memory replacement policy parameter determines which
3600 objects are purged from memory when memory space is needed.
3602 See cache_replacement_policy for details on algorithms.
3607 -----------------------------------------------------------------------------
3610 NAME: cache_replacement_policy
3612 LOC: Config.replPolicy
3615 The cache replacement policy parameter determines which
3616 objects are evicted (replaced) when disk space is needed.
3618 lru : Squid's original list based LRU policy
3619 heap GDSF : Greedy-Dual Size Frequency
3620 heap LFUDA: Least Frequently Used with Dynamic Aging
3621 heap LRU : LRU policy implemented using a heap
3623 Applies to any cache_dir lines listed below this directive.
3625 The LRU policies keeps recently referenced objects.
3627 The heap GDSF policy optimizes object hit rate by keeping smaller
3628 popular objects in cache so it has a better chance of getting a
3629 hit. It achieves a lower byte hit rate than LFUDA though since
3630 it evicts larger (possibly popular) objects.
3632 The heap LFUDA policy keeps popular objects in cache regardless of
3633 their size and thus optimizes byte hit rate at the expense of
3634 hit rate since one large, popular object will prevent many
3635 smaller, slightly less popular objects from being cached.
3637 Both policies utilize a dynamic aging mechanism that prevents
3638 cache pollution that can otherwise occur with frequency-based
3639 replacement policies.
3641 NOTE: if using the LFUDA replacement policy you should increase
3642 the value of maximum_object_size above its default of 4 MB to
3643 to maximize the potential byte hit rate improvement of LFUDA.
3645 For more information about the GDSF and LFUDA cache replacement
3646 policies see http://www.hpl.hp.com/techreports/1999/HPL-1999-69.html
3647 and http://fog.hpl.external.hp.com/techreports/98/HPL-98-173.html.
3650 NAME: minimum_object_size
3654 DEFAULT_DOC: no limit
3655 LOC: Config.Store.minObjectSize
3657 Objects smaller than this size will NOT be saved on disk. The
3658 value is specified in bytes, and the default is 0 KB, which
3659 means all responses can be stored.
3662 NAME: maximum_object_size
3666 LOC: Config.Store.maxObjectSize
3668 Set the default value for max-size parameter on any cache_dir.
3669 The value is specified in bytes, and the default is 4 MB.
3671 If you wish to get a high BYTES hit ratio, you should probably
3672 increase this (one 32 MB object hit counts for 3200 10KB
3675 If you wish to increase hit ratio more than you want to
3676 save bandwidth you should leave this low.
3678 NOTE: if using the LFUDA replacement policy you should increase
3679 this value to maximize the byte hit rate improvement of LFUDA!
3680 See cache_replacement_policy for a discussion of this policy.
3686 DEFAULT_DOC: No disk cache. Store cache ojects only in memory.
3687 LOC: Config.cacheSwap
3690 cache_dir Type Directory-Name Fs-specific-data [options]
3692 You can specify multiple cache_dir lines to spread the
3693 cache among different disk partitions.
3695 Type specifies the kind of storage system to use. Only "ufs"
3696 is built by default. To enable any of the other storage systems
3697 see the --enable-storeio configure option.
3699 'Directory' is a top-level directory where cache swap
3700 files will be stored. If you want to use an entire disk
3701 for caching, this can be the mount-point directory.
3702 The directory must exist and be writable by the Squid
3703 process. Squid will NOT create this directory for you.
3705 In SMP configurations, cache_dir must not precede the workers option
3706 and should use configuration macros or conditionals to give each
3707 worker interested in disk caching a dedicated cache directory.
3710 ==== The ufs store type ====
3712 "ufs" is the old well-known Squid storage format that has always
3716 cache_dir ufs Directory-Name Mbytes L1 L2 [options]
3718 'Mbytes' is the amount of disk space (MB) to use under this
3719 directory. The default is 100 MB. Change this to suit your
3720 configuration. Do NOT put the size of your disk drive here.
3721 Instead, if you want Squid to use the entire disk drive,
3722 subtract 20% and use that value.
3724 'L1' is the number of first-level subdirectories which
3725 will be created under the 'Directory'. The default is 16.
3727 'L2' is the number of second-level subdirectories which
3728 will be created under each first-level directory. The default
3732 ==== The aufs store type ====
3734 "aufs" uses the same storage format as "ufs", utilizing
3735 POSIX-threads to avoid blocking the main Squid process on
3736 disk-I/O. This was formerly known in Squid as async-io.
3739 cache_dir aufs Directory-Name Mbytes L1 L2 [options]
3741 see argument descriptions under ufs above
3744 ==== The diskd store type ====
3746 "diskd" uses the same storage format as "ufs", utilizing a
3747 separate process to avoid blocking the main Squid process on
3751 cache_dir diskd Directory-Name Mbytes L1 L2 [options] [Q1=n] [Q2=n]
3753 see argument descriptions under ufs above
3755 Q1 specifies the number of unacknowledged I/O requests when Squid
3756 stops opening new files. If this many messages are in the queues,
3757 Squid won't open new files. Default is 64
3759 Q2 specifies the number of unacknowledged messages when Squid
3760 starts blocking. If this many messages are in the queues,
3761 Squid blocks until it receives some replies. Default is 72
3763 When Q1 < Q2 (the default), the cache directory is optimized
3764 for lower response time at the expense of a decrease in hit
3765 ratio. If Q1 > Q2, the cache directory is optimized for
3766 higher hit ratio at the expense of an increase in response
3770 ==== The rock store type ====
3773 cache_dir rock Directory-Name Mbytes [options]
3775 The Rock Store type is a database-style storage. All cached
3776 entries are stored in a "database" file, using fixed-size slots.
3777 A single entry occupies one or more slots.
3779 If possible, Squid using Rock Store creates a dedicated kid
3780 process called "disker" to avoid blocking Squid worker(s) on disk
3781 I/O. One disker kid is created for each rock cache_dir. Diskers
3782 are created only when Squid, running in daemon mode, has support
3783 for the IpcIo disk I/O module.
3785 swap-timeout=msec: Squid will not start writing a miss to or
3786 reading a hit from disk if it estimates that the swap operation
3787 will take more than the specified number of milliseconds. By
3788 default and when set to zero, disables the disk I/O time limit
3789 enforcement. Ignored when using blocking I/O module because
3790 blocking synchronous I/O does not allow Squid to estimate the
3791 expected swap wait time.
3793 max-swap-rate=swaps/sec: Artificially limits disk access using
3794 the specified I/O rate limit. Swap out requests that
3795 would cause the average I/O rate to exceed the limit are
3796 delayed. Individual swap in requests (i.e., hits or reads) are
3797 not delayed, but they do contribute to measured swap rate and
3798 since they are placed in the same FIFO queue as swap out
3799 requests, they may wait longer if max-swap-rate is smaller.
3800 This is necessary on file systems that buffer "too
3801 many" writes and then start blocking Squid and other processes
3802 while committing those writes to disk. Usually used together
3803 with swap-timeout to avoid excessive delays and queue overflows
3804 when disk demand exceeds available disk "bandwidth". By default
3805 and when set to zero, disables the disk I/O rate limit
3806 enforcement. Currently supported by IpcIo module only.
3808 slot-size=bytes: The size of a database "record" used for
3809 storing cached responses. A cached response occupies at least
3810 one slot and all database I/O is done using individual slots so
3811 increasing this parameter leads to more disk space waste while
3812 decreasing it leads to more disk I/O overheads. Should be a
3813 multiple of your operating system I/O page size. Defaults to
3814 16KBytes. A housekeeping header is stored with each slot and
3815 smaller slot-sizes will be rejected. The header is smaller than
3819 ==== COMMON OPTIONS ====
3821 no-store no new objects should be stored to this cache_dir.
3823 min-size=n the minimum object size in bytes this cache_dir
3824 will accept. It's used to restrict a cache_dir
3825 to only store large objects (e.g. AUFS) while
3826 other stores are optimized for smaller objects
3830 max-size=n the maximum object size in bytes this cache_dir
3832 The value in maximum_object_size directive sets
3833 the default unless more specific details are
3834 available (ie a small store capacity).
3836 Note: To make optimal use of the max-size limits you should order
3837 the cache_dir lines with the smallest max-size value first.
3841 # Uncomment and adjust the following to add a disk cache directory.
3842 #cache_dir ufs @DEFAULT_SWAP_DIR@ 100 16 256
3846 NAME: store_dir_select_algorithm
3848 LOC: Config.store_dir_select_algorithm
3851 How Squid selects which cache_dir to use when the response
3852 object will fit into more than one.
3854 Regardless of which algorithm is used the cache_dir min-size
3855 and max-size parameters are obeyed. As such they can affect
3856 the selection algorithm by limiting the set of considered
3863 This algorithm is suited to caches with similar cache_dir
3864 sizes and disk speeds.
3866 The disk with the least I/O pending is selected.
3867 When there are multiple disks with the same I/O load ranking
3868 the cache_dir with most available capacity is selected.
3870 When a mix of cache_dir sizes are configured the faster disks
3871 have a naturally lower I/O loading and larger disks have more
3872 capacity. So space used to store objects and data throughput
3873 may be very unbalanced towards larger disks.
3878 This algorithm is suited to caches with unequal cache_dir
3881 Each cache_dir is selected in a rotation. The next suitable
3884 Available cache_dir capacity is only considered in relation
3885 to whether the object will fit and meets the min-size and
3886 max-size parameters.
3888 Disk I/O loading is only considered to prevent overload on slow
3889 disks. This algorithm does not spread objects by size, so any
3890 I/O loading per-disk may appear very unbalanced and volatile.
3892 If several cache_dirs use similar min-size, max-size, or other
3893 limits to to reject certain responses, then do not group such
3894 cache_dir lines together, to avoid round-robin selection bias
3895 towards the first cache_dir after the group. Instead, interleave
3896 cache_dir lines from different groups. For example:
3898 store_dir_select_algorithm round-robin
3899 cache_dir rock /hdd1 ... min-size=100000
3900 cache_dir rock /ssd1 ... max-size=99999
3901 cache_dir rock /hdd2 ... min-size=100000
3902 cache_dir rock /ssd2 ... max-size=99999
3903 cache_dir rock /hdd3 ... min-size=100000
3904 cache_dir rock /ssd3 ... max-size=99999
3907 NAME: max_open_disk_fds
3909 LOC: Config.max_open_disk_fds
3911 DEFAULT_DOC: no limit
3913 To avoid having disk as the I/O bottleneck Squid can optionally
3914 bypass the on-disk cache if more than this amount of disk file
3915 descriptors are open.
3917 A value of 0 indicates no limit.
3920 NAME: cache_swap_low
3921 COMMENT: (percent, 0-100)
3924 LOC: Config.Swap.lowWaterMark
3926 The low-water mark for AUFS/UFS/diskd cache object eviction by
3927 the cache_replacement_policy algorithm.
3929 Removal begins when the swap (disk) usage of a cache_dir is
3930 above this low-water mark and attempts to maintain utilization
3931 near the low-water mark.
3933 As swap utilization increases towards the high-water mark set
3934 by cache_swap_high object eviction becomes more agressive.
3936 The value difference in percentages between low- and high-water
3937 marks represent an eviction rate of 300 objects per second and
3938 the rate continues to scale in agressiveness by multiples of
3939 this above the high-water mark.
3941 Defaults are 90% and 95%. If you have a large cache, 5% could be
3942 hundreds of MB. If this is the case you may wish to set these
3943 numbers closer together.
3945 See also cache_swap_high and cache_replacement_policy
3948 NAME: cache_swap_high
3949 COMMENT: (percent, 0-100)
3952 LOC: Config.Swap.highWaterMark
3954 The high-water mark for AUFS/UFS/diskd cache object eviction by
3955 the cache_replacement_policy algorithm.
3957 Removal begins when the swap (disk) usage of a cache_dir is
3958 above the low-water mark set by cache_swap_low and attempts to
3959 maintain utilization near the low-water mark.
3961 As swap utilization increases towards this high-water mark object
3962 eviction becomes more agressive.
3964 The value difference in percentages between low- and high-water
3965 marks represent an eviction rate of 300 objects per second and
3966 the rate continues to scale in agressiveness by multiples of
3967 this above the high-water mark.
3969 Defaults are 90% and 95%. If you have a large cache, 5% could be
3970 hundreds of MB. If this is the case you may wish to set these
3971 numbers closer together.
3973 See also cache_swap_low and cache_replacement_policy
3978 -----------------------------------------------------------------------------
3985 DEFAULT_DOC: The format definitions squid, common, combined, referrer, useragent are built in.
3989 logformat <name> <format specification>
3991 Defines an access log format.
3993 The <format specification> is a string with embedded % format codes
3995 % format codes all follow the same basic structure where all but
3996 the formatcode is optional. Output strings are automatically escaped
3997 as required according to their context and the output format
3998 modifiers are usually not needed, but can be specified if an explicit
3999 output format is desired.
4001 % ["|[|'|#|/] [-] [[0]width] [{arg}] formatcode [{arg}]
4003 " output in quoted string format
4004 [ output in squid text log format as used by log_mime_hdrs
4005 # output in URL quoted format
4006 / output in shell \-escaped format
4011 width minimum and/or maximum field width:
4012 [width_min][.width_max]
4013 When minimum starts with 0, the field is zero-padded.
4014 String values exceeding maximum width are truncated.
4016 {arg} argument such as header name etc. This field may be
4017 placed before or after the token, but not both at once.
4021 % a literal % character
4022 sn Unique sequence number per log line entry
4023 err_code The ID of an error response served by Squid or
4024 a similar internal error identifier.
4025 err_detail Additional err_code-dependent error information.
4026 note The annotation specified by the argument. Also
4027 logs the adaptation meta headers set by the
4028 adaptation_meta configuration parameter.
4029 If no argument given all annotations logged.
4030 The argument may include a separator to use with
4033 By default, multiple note values are separated with ","
4034 and multiple notes are separated with "\r\n".
4035 When logging named notes with %{name}note, the
4036 explicitly configured separator is used between note
4037 values. When logging all notes with %note, the
4038 explicitly configured separator is used between
4039 individual notes. There is currently no way to
4040 specify both value and notes separators when logging
4041 all notes with %note.
4043 Connection related format codes:
4045 >a Client source IP address
4047 >p Client source port
4048 >eui Client source EUI (MAC address, EUI-48 or EUI-64 identifier)
4049 >la Local IP address the client connected to
4050 >lp Local port number the client connected to
4051 >qos Client connection TOS/DSCP value set by Squid
4052 >nfmark Client connection netfilter mark set by Squid
4054 la Local listening IP address the client connection was connected to.
4055 lp Local listening port number the client connection was connected to.
4057 <a Server IP address of the last server or peer connection
4058 <A Server FQDN or peer name
4059 <p Server port number of the last server or peer connection
4060 <la Local IP address of the last server or peer connection
4061 <lp Local port number of the last server or peer connection
4062 <qos Server connection TOS/DSCP value set by Squid
4063 <nfmark Server connection netfilter mark set by Squid
4065 Time related format codes:
4067 ts Seconds since epoch
4068 tu subsecond time (milliseconds)
4069 tl Local time. Optional strftime format argument
4070 default %d/%b/%Y:%H:%M:%S %z
4071 tg GMT time. Optional strftime format argument
4072 default %d/%b/%Y:%H:%M:%S %z
4073 tr Response time (milliseconds)
4074 dt Total time spent making DNS lookups (milliseconds)
4075 tS Approximate master transaction start time in
4076 <full seconds since epoch>.<fractional seconds> format.
4077 Currently, Squid considers the master transaction
4078 started when a complete HTTP request header initiating
4079 the transaction is received from the client. This is
4080 the same value that Squid uses to calculate transaction
4081 response time when logging %tr to access.log. Currently,
4082 Squid uses millisecond resolution for %tS values,
4083 similar to the default access.log "current time" field
4086 Access Control related format codes:
4088 et Tag returned by external acl
4089 ea Log string returned by external acl
4090 un User name (any available)
4091 ul User name from authentication
4092 ue User name from external acl helper
4093 ui User name from ident
4094 un A user name. Expands to the first available name
4095 from the following list of information sources:
4096 - authenticated user name, like %ul
4097 - user name supplied by an external ACL, like %ue
4098 - SSL client name, like %us
4099 - ident user name, like %ui
4100 credentials Client credentials. The exact meaning depends on
4101 the authentication scheme: For Basic authentication,
4102 it is the password; for Digest, the realm sent by the
4103 client; for NTLM and Negotiate, the client challenge
4104 or client credentials prefixed with "YR " or "KK ".
4106 HTTP related format codes:
4110 [http::]rm Request method (GET/POST etc)
4111 [http::]>rm Request method from client
4112 [http::]<rm Request method sent to server or peer
4113 [http::]ru Request URL from client (historic, filtered for logging)
4114 [http::]>ru Request URL from client
4115 [http::]<ru Request URL sent to server or peer
4116 [http::]>rs Request URL scheme from client
4117 [http::]<rs Request URL scheme sent to server or peer
4118 [http::]>rd Request URL domain from client
4119 [http::]<rd Request URL domain sent to server or peer
4120 [http::]>rP Request URL port from client
4121 [http::]<rP Request URL port sent to server or peer
4122 [http::]rp Request URL path excluding hostname
4123 [http::]>rp Request URL path excluding hostname from client
4124 [http::]<rp Request URL path excluding hostname sent to server or peer
4125 [http::]rv Request protocol version
4126 [http::]>rv Request protocol version from client
4127 [http::]<rv Request protocol version sent to server or peer
4129 [http::]>h Original received request header.
4130 Usually differs from the request header sent by
4131 Squid, although most fields are often preserved.
4132 Accepts optional header field name/value filter
4133 argument using name[:[separator]element] format.
4134 [http::]>ha Received request header after adaptation and
4135 redirection (pre-cache REQMOD vectoring point).
4136 Usually differs from the request header sent by
4137 Squid, although most fields are often preserved.
4138 Optional header name argument as for >h
4142 [http::]<Hs HTTP status code received from the next hop
4143 [http::]>Hs HTTP status code sent to the client
4145 [http::]<h Reply header. Optional header name argument
4148 [http::]mt MIME content type
4153 [http::]st Total size of request + reply traffic with client
4154 [http::]>st Total size of request received from client.
4155 Excluding chunked encoding bytes.
4156 [http::]<st Total size of reply sent to client (after adaptation)
4158 [http::]>sh Size of request headers received from client
4159 [http::]<sh Size of reply headers sent to client (after adaptation)
4161 [http::]<sH Reply high offset sent
4162 [http::]<sS Upstream object size
4164 [http::]<bs Number of HTTP-equivalent message body bytes
4165 received from the next hop, excluding chunked
4166 transfer encoding and control messages.
4167 Generated FTP/Gopher listings are treated as
4172 [http::]<pt Peer response time in milliseconds. The timer starts
4173 when the last request byte is sent to the next hop
4174 and stops when the last response byte is received.
4175 [http::]<tt Total time in milliseconds. The timer
4176 starts with the first connect request (or write I/O)
4177 sent to the first selected peer. The timer stops
4178 with the last I/O with the last peer.
4180 Squid handling related format codes:
4182 Ss Squid request status (TCP_MISS etc)
4183 Sh Squid hierarchy status (DEFAULT_PARENT etc)
4185 SSL-related format codes:
4187 ssl::bump_mode SslBump decision for the transaction:
4189 For CONNECT requests that initiated bumping of
4190 a connection and for any request received on
4191 an already bumped connection, Squid logs the
4192 corresponding SslBump mode ("server-first" or
4193 "client-first"). See the ssl_bump option for
4194 more information about these modes.
4196 A "none" token is logged for requests that
4197 triggered "ssl_bump" ACL evaluation matching
4198 either a "none" rule or no rules at all.
4200 In all other cases, a single dash ("-") is
4203 ssl::>sni SSL client SNI sent to Squid. Available only
4204 after the peek, stare, or splice SSL bumping
4208 The Subject field of the received client
4209 SSL certificate or a dash ('-') if Squid has
4210 received an invalid/malformed certificate or
4211 no certificate at all. Consider encoding the
4212 logged value because Subject often has spaces.
4215 The Issuer field of the received client
4216 SSL certificate or a dash ('-') if Squid has
4217 received an invalid/malformed certificate or
4218 no certificate at all. Consider encoding the
4219 logged value because Issuer often has spaces.
4222 The list of certificate validation errors
4223 detected by Squid (including OpenSSL and
4224 certificate validation helper components). The
4225 errors are listed in the discovery order. By
4226 default, the error codes are separated by ':'.
4227 Accepts an optional separator argument.
4229 %ssl::>negotiated_version The negotiated TLS version of the
4232 %ssl::<negotiated_version The negotiated TLS version of the
4233 last server or peer connection.
4235 %ssl::>received_hello_version The TLS version of the Hello
4236 message received from TLS client.
4238 %ssl::<received_hello_version The TLS version of the Hello
4239 message received from TLS server.
4241 %ssl::>received_supported_version The maximum TLS version
4242 supported by the TLS client.
4244 %ssl::<received_supported_version The maximum TLS version
4245 supported by the TLS server.
4247 %ssl::>negotiated_cipher The negotiated cipher of the
4250 %ssl::<negotiated_cipher The negotiated cipher of the
4251 last server or peer connection.
4253 If ICAP is enabled, the following code becomes available (as
4254 well as ICAP log codes documented with the icap_log option):
4256 icap::tt Total ICAP processing time for the HTTP
4257 transaction. The timer ticks when ICAP
4258 ACLs are checked and when ICAP
4259 transaction is in progress.
4261 If adaptation is enabled the following codes become available:
4263 adapt::<last_h The header of the last ICAP response or
4264 meta-information from the last eCAP
4265 transaction related to the HTTP transaction.
4266 Like <h, accepts an optional header name
4269 adapt::sum_trs Summed adaptation transaction response
4270 times recorded as a comma-separated list in
4271 the order of transaction start time. Each time
4272 value is recorded as an integer number,
4273 representing response time of one or more
4274 adaptation (ICAP or eCAP) transaction in
4275 milliseconds. When a failed transaction is
4276 being retried or repeated, its time is not
4277 logged individually but added to the
4278 replacement (next) transaction. See also:
4281 adapt::all_trs All adaptation transaction response times.
4282 Same as adaptation_strs but response times of
4283 individual transactions are never added
4284 together. Instead, all transaction response
4285 times are recorded individually.
4287 You can prefix adapt::*_trs format codes with adaptation
4288 service name in curly braces to record response time(s) specific
4289 to that service. For example: %{my_service}adapt::sum_trs
4291 The default formats available (which do not need re-defining) are:
4293 logformat squid %ts.%03tu %6tr %>a %Ss/%03>Hs %<st %rm %ru %[un %Sh/%<a %mt
4294 logformat common %>a %[ui %[un [%tl] "%rm %ru HTTP/%rv" %>Hs %<st %Ss:%Sh
4295 logformat combined %>a %[ui %[un [%tl] "%rm %ru HTTP/%rv" %>Hs %<st "%{Referer}>h" "%{User-Agent}>h" %Ss:%Sh
4296 logformat referrer %ts.%03tu %>a %{Referer}>h %ru
4297 logformat useragent %>a [%tl] "%{User-Agent}>h"
4299 NOTE: When the log_mime_hdrs directive is set to ON.
4300 The squid, common and combined formats have a safely encoded copy
4301 of the mime headers appended to each line within a pair of brackets.
4303 NOTE: The common and combined formats are not quite true to the Apache definition.
4304 The logs from Squid contain an extra status and hierarchy code appended.
4308 NAME: access_log cache_access_log
4310 LOC: Config.Log.accesslogs
4311 DEFAULT_IF_NONE: daemon:@DEFAULT_ACCESS_LOG@ squid
4313 Configures whether and how Squid logs HTTP and ICP transactions.
4314 If access logging is enabled, a single line is logged for every
4315 matching HTTP or ICP request. The recommended directive formats are:
4317 access_log <module>:<place> [option ...] [acl acl ...]
4318 access_log none [acl acl ...]
4320 The following directive format is accepted but may be deprecated:
4321 access_log <module>:<place> [<logformat name> [acl acl ...]]
4323 In most cases, the first ACL name must not contain the '=' character
4324 and should not be equal to an existing logformat name. You can always
4325 start with an 'all' ACL to work around those restrictions.
4327 Will log to the specified module:place using the specified format (which
4328 must be defined in a logformat directive) those entries which match
4329 ALL the acl's specified (which must be defined in acl clauses).
4330 If no acl is specified, all requests will be logged to this destination.
4332 ===== Available options for the recommended directive format =====
4334 logformat=name Names log line format (either built-in or
4335 defined by a logformat directive). Defaults
4338 buffer-size=64KB Defines approximate buffering limit for log
4339 records (see buffered_logs). Squid should not
4340 keep more than the specified size and, hence,
4341 should flush records before the buffer becomes
4342 full to avoid overflows under normal
4343 conditions (the exact flushing algorithm is
4344 module-dependent though). The on-error option
4345 controls overflow handling.
4347 on-error=die|drop Defines action on unrecoverable errors. The
4348 'drop' action ignores (i.e., does not log)
4349 affected log records. The default 'die' action
4350 kills the affected worker. The drop action
4351 support has not been tested for modules other
4354 rotate=N Specifies the number of log file rotations to
4355 make when you run 'squid -k rotate'. The default
4356 is to obey the logfile_rotate directive. Setting
4357 rotate=0 will disable the file name rotation,
4358 but the log files are still closed and re-opened.
4359 This will enable you to rename the logfiles
4360 yourself just before sending the rotate signal.
4361 Only supported by the stdio module.
4363 ===== Modules Currently available =====
4365 none Do not log any requests matching these ACL.
4366 Do not specify Place or logformat name.
4368 stdio Write each log line to disk immediately at the completion of
4370 Place: the filename and path to be written.
4372 daemon Very similar to stdio. But instead of writing to disk the log
4373 line is passed to a daemon helper for asychronous handling instead.
4374 Place: varies depending on the daemon.
4376 log_file_daemon Place: the file name and path to be written.
4378 syslog To log each request via syslog facility.
4379 Place: The syslog facility and priority level for these entries.
4380 Place Format: facility.priority
4382 where facility could be any of:
4383 authpriv, daemon, local0 ... local7 or user.
4385 And priority could be any of:
4386 err, warning, notice, info, debug.
4388 udp To send each log line as text data to a UDP receiver.
4389 Place: The destination host name or IP and port.
4390 Place Format: //host:port
4392 tcp To send each log line as text data to a TCP receiver.
4393 Lines may be accumulated before sending (see buffered_logs).
4394 Place: The destination host name or IP and port.
4395 Place Format: //host:port
4398 access_log daemon:@DEFAULT_ACCESS_LOG@ squid
4404 LOC: Config.Log.icaplogs
4407 ICAP log files record ICAP transaction summaries, one line per
4410 The icap_log option format is:
4411 icap_log <filepath> [<logformat name> [acl acl ...]]
4412 icap_log none [acl acl ...]]
4414 Please see access_log option documentation for details. The two
4415 kinds of logs share the overall configuration approach and many
4418 ICAP processing of a single HTTP message or transaction may
4419 require multiple ICAP transactions. In such cases, multiple
4420 ICAP transaction log lines will correspond to a single access
4423 ICAP log uses logformat codes that make sense for an ICAP
4424 transaction. Header-related codes are applied to the HTTP header
4425 embedded in an ICAP server response, with the following caveats:
4426 For REQMOD, there is no HTTP response header unless the ICAP
4427 server performed request satisfaction. For RESPMOD, the HTTP
4428 request header is the header sent to the ICAP server. For
4429 OPTIONS, there are no HTTP headers.
4431 The following format codes are also available for ICAP logs:
4433 icap::<A ICAP server IP address. Similar to <A.
4435 icap::<service_name ICAP service name from the icap_service
4436 option in Squid configuration file.
4438 icap::ru ICAP Request-URI. Similar to ru.
4440 icap::rm ICAP request method (REQMOD, RESPMOD, or
4441 OPTIONS). Similar to existing rm.
4443 icap::>st Bytes sent to the ICAP server (TCP payload
4444 only; i.e., what Squid writes to the socket).
4446 icap::<st Bytes received from the ICAP server (TCP
4447 payload only; i.e., what Squid reads from
4450 icap::<bs Number of message body bytes received from the
4451 ICAP server. ICAP message body, if any, usually
4452 includes encapsulated HTTP message headers and
4453 possibly encapsulated HTTP message body. The
4454 HTTP body part is dechunked before its size is
4457 icap::tr Transaction response time (in
4458 milliseconds). The timer starts when
4459 the ICAP transaction is created and
4460 stops when the transaction is completed.
4463 icap::tio Transaction I/O time (in milliseconds). The
4464 timer starts when the first ICAP request
4465 byte is scheduled for sending. The timers
4466 stops when the last byte of the ICAP response
4469 icap::to Transaction outcome: ICAP_ERR* for all
4470 transaction errors, ICAP_OPT for OPTION
4471 transactions, ICAP_ECHO for 204
4472 responses, ICAP_MOD for message
4473 modification, and ICAP_SAT for request
4474 satisfaction. Similar to Ss.
4476 icap::Hs ICAP response status code. Similar to Hs.
4478 icap::>h ICAP request header(s). Similar to >h.
4480 icap::<h ICAP response header(s). Similar to <h.
4482 The default ICAP log format, which can be used without an explicit
4483 definition, is called icap_squid:
4485 logformat icap_squid %ts.%03tu %6icap::tr %>a %icap::to/%03icap::Hs %icap::<size %icap::rm %icap::ru% %un -/%icap::<A -
4487 See also: logformat, log_icap, and %adapt::<last_h
4490 NAME: logfile_daemon
4492 DEFAULT: @DEFAULT_LOGFILED@
4493 LOC: Log::TheConfig.logfile_daemon
4495 Specify the path to the logfile-writing daemon. This daemon is
4496 used to write the access and store logs, if configured.
4498 Squid sends a number of commands to the log daemon:
4499 L<data>\n - logfile data
4504 r<n>\n - set rotate count to <n>
4505 b<n>\n - 1 = buffer output, 0 = don't buffer output
4507 No responses is expected.
4510 NAME: stats_collection
4512 LOC: Config.accessList.stats_collection
4514 DEFAULT_DOC: Allow logging for all transactions.
4515 COMMENT: allow|deny acl acl...
4517 This options allows you to control which requests gets accounted
4518 in performance counters.
4520 This clause only supports fast acl types.
4521 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
4524 NAME: cache_store_log
4527 LOC: Config.Log.store
4529 Logs the activities of the storage manager. Shows which
4530 objects are ejected from the cache, and which objects are
4531 saved and for how long.
4532 There are not really utilities to analyze this data, so you can safely
4533 disable it (the default).
4535 Store log uses modular logging outputs. See access_log for the list
4536 of modules supported.
4539 cache_store_log stdio:@DEFAULT_STORE_LOG@
4540 cache_store_log daemon:@DEFAULT_STORE_LOG@
4543 NAME: cache_swap_state cache_swap_log
4545 LOC: Config.Log.swap
4547 DEFAULT_DOC: Store the journal inside its cache_dir
4549 Location for the cache "swap.state" file. This index file holds
4550 the metadata of objects saved on disk. It is used to rebuild
4551 the cache during startup. Normally this file resides in each
4552 'cache_dir' directory, but you may specify an alternate
4553 pathname here. Note you must give a full filename, not just
4554 a directory. Since this is the index for the whole object
4555 list you CANNOT periodically rotate it!
4557 If %s can be used in the file name it will be replaced with a
4558 a representation of the cache_dir name where each / is replaced
4559 with '.'. This is needed to allow adding/removing cache_dir
4560 lines when cache_swap_log is being used.
4562 If have more than one 'cache_dir', and %s is not used in the name
4563 these swap logs will have names such as:
4569 The numbered extension (which is added automatically)
4570 corresponds to the order of the 'cache_dir' lines in this
4571 configuration file. If you change the order of the 'cache_dir'
4572 lines in this file, these index files will NOT correspond to
4573 the correct 'cache_dir' entry (unless you manually rename
4574 them). We recommend you do NOT use this option. It is
4575 better to keep these index files in each 'cache_dir' directory.
4578 NAME: logfile_rotate
4581 LOC: Config.Log.rotateNumber
4583 Specifies the default number of logfile rotations to make when you
4584 type 'squid -k rotate'. The default is 10, which will rotate
4585 with extensions 0 through 9. Setting logfile_rotate to 0 will
4586 disable the file name rotation, but the logfiles are still closed
4587 and re-opened. This will enable you to rename the logfiles
4588 yourself just before sending the rotate signal.
4590 Note, from Squid-3.1 this option is only a default for cache.log,
4591 that log can be rotated separately by using debug_options.
4593 Note, from Squid-3.6 this option is only a default for access.log
4594 recorded by stdio: module. Those logs can be rotated separately by
4595 using the rotate=N option on their access_log directive.
4597 Note, the 'squid -k rotate' command normally sends a USR1
4598 signal to the running squid process. In certain situations
4599 (e.g. on Linux with Async I/O), USR1 is used for other
4600 purposes, so -k rotate uses another signal. It is best to get
4601 in the habit of using 'squid -k rotate' instead of 'kill -USR1
4608 DEFAULT: @DEFAULT_MIME_TABLE@
4609 LOC: Config.mimeTablePathname
4611 Path to Squid's icon configuration file.
4613 You shouldn't need to change this, but the default file contains
4614 examples and formatting information if you do.
4620 LOC: Config.onoff.log_mime_hdrs
4623 The Cache can record both the request and the response MIME
4624 headers for each HTTP transaction. The headers are encoded
4625 safely and will appear as two bracketed fields at the end of
4626 the access log (for either the native or httpd-emulated log
4627 formats). To enable this logging set log_mime_hdrs to 'on'.
4632 DEFAULT: @DEFAULT_PID_FILE@
4633 LOC: Config.pidFilename
4635 A filename to write the process-id to. To disable, enter "none".
4638 NAME: client_netmask
4640 LOC: Config.Addrs.client_netmask
4642 DEFAULT_DOC: Log full client IP address
4644 A netmask for client addresses in logfiles and cachemgr output.
4645 Change this to protect the privacy of your cache clients.
4646 A netmask of 255.255.255.0 will log all IP's in that range with
4647 the last digit set to '0'.
4650 NAME: strip_query_terms
4652 LOC: Config.onoff.strip_query_terms
4655 By default, Squid strips query terms from requested URLs before
4656 logging. This protects your user's privacy and reduces log size.
4658 When investigating HIT/MISS or other caching behaviour you
4659 will need to disable this to see the full URL used by Squid.
4666 LOC: Config.onoff.buffered_logs
4668 Whether to write/send access_log records ASAP or accumulate them and
4669 then write/send them in larger chunks. Buffering may improve
4670 performance because it decreases the number of I/Os. However,
4671 buffering increases the delay before log records become available to
4672 the final recipient (e.g., a disk file or logging daemon) and,
4673 hence, increases the risk of log records loss.
4675 Note that even when buffered_logs are off, Squid may have to buffer
4676 records if it cannot write/send them immediately due to pending I/Os
4677 (e.g., the I/O writing the previous log record) or connectivity loss.
4679 Currently honored by 'daemon' and 'tcp' access_log modules only.
4682 NAME: netdb_filename
4684 DEFAULT: stdio:@DEFAULT_NETDB_FILE@
4685 LOC: Config.netdbFilename
4688 Where Squid stores it's netdb journal.
4689 When enabled this journal preserves netdb state between restarts.
4691 To disable, enter "none".
4695 OPTIONS FOR TROUBLESHOOTING
4696 -----------------------------------------------------------------------------
4701 DEFAULT_IF_NONE: @DEFAULT_CACHE_LOG@
4702 LOC: Debug::cache_log
4704 Squid administrative logging file.
4706 This is where general information about Squid behavior goes. You can
4707 increase the amount of data logged to this file and how often it is
4708 rotated with "debug_options"
4714 DEFAULT_DOC: Log all critical and important messages.
4715 LOC: Debug::debugOptions
4717 Logging options are set as section,level where each source file
4718 is assigned a unique section. Lower levels result in less
4719 output, Full debugging (level 9) can result in a very large
4720 log file, so be careful.
4722 The magic word "ALL" sets debugging levels for all sections.
4723 The default is to run with "ALL,1" to record important warnings.
4725 The rotate=N option can be used to keep more or less of these logs
4726 than would otherwise be kept by logfile_rotate.
4727 For most uses a single log should be enough to monitor current
4728 events affecting Squid.
4733 LOC: Config.coredump_dir
4734 DEFAULT_IF_NONE: none
4735 DEFAULT_DOC: Use the directory from where Squid was started.
4737 By default Squid leaves core files in the directory from where
4738 it was started. If you set 'coredump_dir' to a directory
4739 that exists, Squid will chdir() to that directory at startup
4740 and coredump files will be left there.
4744 # Leave coredumps in the first cache dir
4745 coredump_dir @DEFAULT_SWAP_DIR@
4751 OPTIONS FOR FTP GATEWAYING
4752 -----------------------------------------------------------------------------
4758 LOC: Config.Ftp.anon_user
4760 If you want the anonymous login password to be more informative
4761 (and enable the use of picky FTP servers), set this to something
4762 reasonable for your domain, like wwwuser@somewhere.net
4764 The reason why this is domainless by default is the
4765 request can be made on the behalf of a user in any domain,
4766 depending on how the cache is used.
4767 Some FTP server also validate the email address is valid
4768 (for example perl.com).
4774 LOC: Config.Ftp.passive
4776 If your firewall does not allow Squid to use passive
4777 connections, turn off this option.
4779 Use of ftp_epsv_all option requires this to be ON.
4785 LOC: Config.Ftp.epsv_all
4787 FTP Protocol extensions permit the use of a special "EPSV ALL" command.
4789 NATs may be able to put the connection on a "fast path" through the
4790 translator, as the EPRT command will never be used and therefore,
4791 translation of the data portion of the segments will never be needed.
4793 When a client only expects to do two-way FTP transfers this may be
4795 If squid finds that it must do a three-way FTP transfer after issuing
4796 an EPSV ALL command, the FTP session will fail.
4798 If you have any doubts about this option do not use it.
4799 Squid will nicely attempt all other connection methods.
4801 Requires ftp_passive to be ON (default) for any effect.
4807 LOC: Config.accessList.ftp_epsv
4809 FTP Protocol extensions permit the use of a special "EPSV" command.
4811 NATs may be able to put the connection on a "fast path" through the
4812 translator using EPSV, as the EPRT command will never be used
4813 and therefore, translation of the data portion of the segments
4814 will never be needed.
4816 EPSV is often required to interoperate with FTP servers on IPv6
4817 networks. On the other hand, it may break some IPv4 servers.
4819 By default, EPSV may try EPSV with any FTP server. To fine tune
4820 that decision, you may restrict EPSV to certain clients or servers
4823 ftp_epsv allow|deny al1 acl2 ...
4825 WARNING: Disabling EPSV may cause problems with external NAT and IPv6.
4827 Only fast ACLs are supported.
4828 Requires ftp_passive to be ON (default) for any effect.
4834 LOC: Config.Ftp.eprt
4836 FTP Protocol extensions permit the use of a special "EPRT" command.
4838 This extension provides a protocol neutral alternative to the
4839 IPv4-only PORT command. When supported it enables active FTP data
4840 channels over IPv6 and efficient NAT handling.
4842 Turning this OFF will prevent EPRT being attempted and will skip
4843 straight to using PORT for IPv4 servers.
4845 Some devices are known to not handle this extension correctly and
4846 may result in crashes. Devices which suport EPRT enough to fail
4847 cleanly will result in Squid attempting PORT anyway. This directive
4848 should only be disabled when EPRT results in device failures.
4850 WARNING: Doing so will convert Squid back to the old behavior with all
4851 the related problems with external NAT devices/layers and IPv4-only FTP.
4854 NAME: ftp_sanitycheck
4857 LOC: Config.Ftp.sanitycheck
4859 For security and data integrity reasons Squid by default performs
4860 sanity checks of the addresses of FTP data connections ensure the
4861 data connection is to the requested server. If you need to allow
4862 FTP connections to servers using another IP address for the data
4863 connection turn this off.
4866 NAME: ftp_telnet_protocol
4869 LOC: Config.Ftp.telnet
4871 The FTP protocol is officially defined to use the telnet protocol
4872 as transport channel for the control connection. However, many
4873 implementations are broken and does not respect this aspect of
4876 If you have trouble accessing files with ASCII code 255 in the
4877 path or similar problems involving this ASCII code you can
4878 try setting this directive to off. If that helps, report to the
4879 operator of the FTP server in question that their FTP server
4880 is broken and does not follow the FTP standard.
4884 OPTIONS FOR EXTERNAL SUPPORT PROGRAMS
4885 -----------------------------------------------------------------------------
4890 DEFAULT: @DEFAULT_DISKD@
4891 LOC: Config.Program.diskd
4893 Specify the location of the diskd executable.
4894 Note this is only useful if you have compiled in
4895 diskd as one of the store io modules.
4898 NAME: unlinkd_program
4901 DEFAULT: @DEFAULT_UNLINKD@
4902 LOC: Config.Program.unlinkd
4904 Specify the location of the executable for file deletion process.
4907 NAME: pinger_program
4910 DEFAULT: @DEFAULT_PINGER@
4913 Specify the location of the executable for the pinger process.
4922 Control whether the pinger is active at run-time.
4923 Enables turning ICMP pinger on and off with a simple
4924 squid -k reconfigure.
4929 OPTIONS FOR URL REWRITING
4930 -----------------------------------------------------------------------------
4933 NAME: url_rewrite_program redirect_program
4935 LOC: Config.Program.redirect
4938 Specify the location of the executable URL rewriter to use.
4939 Since they can perform almost any function there isn't one included.
4941 For each requested URL, the rewriter will receive on line with the format
4943 [channel-ID <SP>] URL [<SP> extras]<NL>
4945 See url_rewrite_extras on how to send "extras" with optional values to
4947 After processing the request the helper must reply using the following format:
4949 [channel-ID <SP>] result [<SP> kv-pairs]
4951 The result code can be:
4953 OK status=30N url="..."
4954 Redirect the URL to the one supplied in 'url='.
4955 'status=' is optional and contains the status code to send
4956 the client in Squids HTTP response. It must be one of the
4957 HTTP redirect status codes: 301, 302, 303, 307, 308.
4958 When no status is given Squid will use 302.
4960 OK rewrite-url="..."
4961 Rewrite the URL to the one supplied in 'rewrite-url='.
4962 The new URL is fetched directly by Squid and returned to
4963 the client as the response to its request.
4966 When neither of url= and rewrite-url= are sent Squid does
4970 Do not change the URL.
4973 An internal error occurred in the helper, preventing
4974 a result being identified. The 'message=' key name is
4975 reserved for delivering a log message.
4978 In addition to the above kv-pairs Squid also understands the following
4979 optional kv-pairs received from URL rewriters:
4981 Associates a TAG with the client TCP connection.
4982 The TAG is treated as a regular annotation but persists across
4983 future requests on the client connection rather than just the
4984 current request. A helper may update the TAG during subsequent
4985 requests be returning a new kv-pair.
4987 When using the concurrency= option the protocol is changed by
4988 introducing a query channel tag in front of the request/response.
4989 The query channel tag is a number between 0 and concurrency-1.
4990 This value must be echoed back unchanged to Squid as the first part
4991 of the response relating to its request.
4993 WARNING: URL re-writing ability should be avoided whenever possible.
4994 Use the URL redirect form of response instead.
4996 Re-write creates a difference in the state held by the client
4997 and server. Possibly causing confusion when the server response
4998 contains snippets of its view state. Embeded URLs, response
4999 and content Location headers, etc. are not re-written by this
5002 By default, a URL rewriter is not used.
5005 NAME: url_rewrite_children redirect_children
5006 TYPE: HelperChildConfig
5007 DEFAULT: 20 startup=0 idle=1 concurrency=0
5008 LOC: Config.redirectChildren
5010 The maximum number of redirector processes to spawn. If you limit
5011 it too few Squid will have to wait for them to process a backlog of
5012 URLs, slowing it down. If you allow too many they will use RAM
5013 and other system resources noticably.
5015 The startup= and idle= options allow some measure of skew in your
5020 Sets a minimum of how many processes are to be spawned when Squid
5021 starts or reconfigures. When set to zero the first request will
5022 cause spawning of the first child process to handle it.
5024 Starting too few will cause an initial slowdown in traffic as Squid
5025 attempts to simultaneously spawn enough processes to cope.
5029 Sets a minimum of how many processes Squid is to try and keep available
5030 at all times. When traffic begins to rise above what the existing
5031 processes can handle this many more will be spawned up to the maximum
5032 configured. A minimum setting of 1 is required.
5036 The number of requests each redirector helper can handle in
5037 parallel. Defaults to 0 which indicates the redirector
5038 is a old-style single threaded redirector.
5040 When this directive is set to a value >= 1 then the protocol
5041 used to communicate with the helper is modified to include
5042 an ID in front of the request/response. The ID from the request
5043 must be echoed back with the response to that request.
5047 Sets the maximum number of queued requests.
5048 If the queued requests exceed queue size and redirector_bypass
5049 configuration option is set, then redirector is bypassed. Otherwise, if
5050 overloading persists squid may abort its operation.
5051 The default value is set to 2*numberofchildren.
5054 NAME: url_rewrite_host_header redirect_rewrites_host_header
5057 LOC: Config.onoff.redir_rewrites_host
5059 To preserve same-origin security policies in browsers and
5060 prevent Host: header forgery by redirectors Squid rewrites
5061 any Host: header in redirected requests.
5063 If you are running an accelerator this may not be a wanted
5064 effect of a redirector. This directive enables you disable
5065 Host: alteration in reverse-proxy traffic.
5067 WARNING: Entries are cached on the result of the URL rewriting
5068 process, so be careful if you have domain-virtual hosts.
5070 WARNING: Squid and other software verifies the URL and Host
5071 are matching, so be careful not to relay through other proxies
5072 or inspecting firewalls with this disabled.
5075 NAME: url_rewrite_access redirector_access
5078 DEFAULT_DOC: Allow, unless rules exist in squid.conf.
5079 LOC: Config.accessList.redirector
5081 If defined, this access list specifies which requests are
5082 sent to the redirector processes.
5084 This clause supports both fast and slow acl types.
5085 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
5088 NAME: url_rewrite_bypass redirector_bypass
5090 LOC: Config.onoff.redirector_bypass
5093 When this is 'on', a request will not go through the
5094 redirector if all the helpers are busy. If this is 'off'
5095 and the redirector queue grows too large, Squid will exit
5096 with a FATAL error and ask you to increase the number of
5097 redirectors. You should only enable this if the redirectors
5098 are not critical to your caching system. If you use
5099 redirectors for access control, and you enable this option,
5100 users may have access to pages they should not
5101 be allowed to request.
5102 This options sets default queue-size option of the url_rewrite_children
5106 NAME: url_rewrite_extras
5107 TYPE: TokenOrQuotedString
5108 LOC: Config.redirector_extras
5109 DEFAULT: "%>a/%>A %un %>rm myip=%la myport=%lp"
5111 Specifies a string to be append to request line format for the
5112 rewriter helper. "Quoted" format values may contain spaces and
5113 logformat %macros. In theory, any logformat %macro can be used.
5114 In practice, a %macro expands as a dash (-) if the helper request is
5115 sent before the required macro information is available to Squid.
5118 NAME: url_rewrite_timeout
5119 TYPE: UrlHelperTimeout
5120 LOC: Config.onUrlRewriteTimeout
5122 DEFAULT_DOC: Squid waits for the helper response forever
5124 Squid times active requests to redirector. The timeout value and Squid
5125 reaction to a timed out request are configurable using the following
5128 url_rewrite_timeout timeout time-units on_timeout=<action> [response=<quoted-response>]
5130 supported timeout actions:
5131 fail Squid return a ERR_GATEWAY_FAILURE error page
5133 bypass Do not re-write the URL
5135 retry Send the lookup to the helper again
5137 use_configured_response
5138 Use the <quoted-response> as helper response
5142 OPTIONS FOR STORE ID
5143 -----------------------------------------------------------------------------
5146 NAME: store_id_program storeurl_rewrite_program
5148 LOC: Config.Program.store_id
5151 Specify the location of the executable StoreID helper to use.
5152 Since they can perform almost any function there isn't one included.
5154 For each requested URL, the helper will receive one line with the format
5156 [channel-ID <SP>] URL [<SP> extras]<NL>
5159 After processing the request the helper must reply using the following format:
5161 [channel-ID <SP>] result [<SP> kv-pairs]
5163 The result code can be:
5166 Use the StoreID supplied in 'store-id='.
5169 The default is to use HTTP request URL as the store ID.
5172 An internal error occured in the helper, preventing
5173 a result being identified.
5175 In addition to the above kv-pairs Squid also understands the following
5176 optional kv-pairs received from URL rewriters:
5178 Associates a TAG with the client TCP connection.
5179 Please see url_rewrite_program related documentation for this
5182 Helper programs should be prepared to receive and possibly ignore
5183 additional whitespace-separated tokens on each input line.
5185 When using the concurrency= option the protocol is changed by
5186 introducing a query channel tag in front of the request/response.
5187 The query channel tag is a number between 0 and concurrency-1.
5188 This value must be echoed back unchanged to Squid as the first part
5189 of the response relating to its request.
5191 NOTE: when using StoreID refresh_pattern will apply to the StoreID
5192 returned from the helper and not the URL.
5194 WARNING: Wrong StoreID value returned by a careless helper may result
5195 in the wrong cached response returned to the user.
5197 By default, a StoreID helper is not used.
5200 NAME: store_id_extras
5201 TYPE: TokenOrQuotedString
5202 LOC: Config.storeId_extras
5203 DEFAULT: "%>a/%>A %un %>rm myip=%la myport=%lp"
5205 Specifies a string to be append to request line format for the
5206 StoreId helper. "Quoted" format values may contain spaces and
5207 logformat %macros. In theory, any logformat %macro can be used.
5208 In practice, a %macro expands as a dash (-) if the helper request is
5209 sent before the required macro information is available to Squid.
5212 NAME: store_id_children storeurl_rewrite_children
5213 TYPE: HelperChildConfig
5214 DEFAULT: 20 startup=0 idle=1 concurrency=0
5215 LOC: Config.storeIdChildren
5217 The maximum number of StoreID helper processes to spawn. If you limit
5218 it too few Squid will have to wait for them to process a backlog of
5219 requests, slowing it down. If you allow too many they will use RAM
5220 and other system resources noticably.
5222 The startup= and idle= options allow some measure of skew in your
5227 Sets a minimum of how many processes are to be spawned when Squid
5228 starts or reconfigures. When set to zero the first request will
5229 cause spawning of the first child process to handle it.
5231 Starting too few will cause an initial slowdown in traffic as Squid
5232 attempts to simultaneously spawn enough processes to cope.
5236 Sets a minimum of how many processes Squid is to try and keep available
5237 at all times. When traffic begins to rise above what the existing
5238 processes can handle this many more will be spawned up to the maximum
5239 configured. A minimum setting of 1 is required.
5243 The number of requests each storeID helper can handle in
5244 parallel. Defaults to 0 which indicates the helper
5245 is a old-style single threaded program.
5247 When this directive is set to a value >= 1 then the protocol
5248 used to communicate with the helper is modified to include
5249 an ID in front of the request/response. The ID from the request
5250 must be echoed back with the response to that request.
5254 Sets the maximum number of queued requests.
5255 If the queued requests exceed queue size and store_id_bypass
5256 configuration option is set, then storeID helper is bypassed. Otherwise,
5257 if overloading persists squid may abort its operation.
5258 The default value is set to 2*numberofchildren.
5261 NAME: store_id_access storeurl_rewrite_access
5264 DEFAULT_DOC: Allow, unless rules exist in squid.conf.
5265 LOC: Config.accessList.store_id
5267 If defined, this access list specifies which requests are
5268 sent to the StoreID processes. By default all requests
5271 This clause supports both fast and slow acl types.
5272 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
5275 NAME: store_id_bypass storeurl_rewrite_bypass
5277 LOC: Config.onoff.store_id_bypass
5280 When this is 'on', a request will not go through the
5281 helper if all helpers are busy. If this is 'off'
5282 and the helper queue grows too large, Squid will exit
5283 with a FATAL error and ask you to increase the number of
5284 helpers. You should only enable this if the helperss
5285 are not critical to your caching system. If you use
5286 helpers for critical caching components, and you enable this
5287 option, users may not get objects from cache.
5288 This options sets default queue-size option of the store_id_children
5293 OPTIONS FOR TUNING THE CACHE
5294 -----------------------------------------------------------------------------
5297 NAME: cache no_cache
5300 DEFAULT_DOC: By default, this directive is unused and has no effect.
5301 LOC: Config.accessList.noCache
5303 Requests denied by this directive will not be served from the cache
5304 and their responses will not be stored in the cache. This directive
5305 has no effect on other transactions and on already cached responses.
5307 This clause supports both fast and slow acl types.
5308 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
5310 This and the two other similar caching directives listed below are
5311 checked at different transaction processing stages, have different
5312 access to response information, affect different cache operations,
5313 and differ in slow ACLs support:
5315 * cache: Checked before Squid makes a hit/miss determination.
5316 No access to reply information!
5317 Denies both serving a hit and storing a miss.
5318 Supports both fast and slow ACLs.
5319 * send_hit: Checked after a hit was detected.
5320 Has access to reply (hit) information.
5321 Denies serving a hit only.
5322 Supports fast ACLs only.
5323 * store_miss: Checked before storing a cachable miss.
5324 Has access to reply (miss) information.
5325 Denies storing a miss only.
5326 Supports fast ACLs only.
5328 If you are not sure which of the three directives to use, apply the
5329 following decision logic:
5331 * If your ACL(s) are of slow type _and_ need response info, redesign.
5332 Squid does not support that particular combination at this time.
5334 * If your directive ACL(s) are of slow type, use "cache"; and/or
5335 * if your directive ACL(s) need no response info, use "cache".
5337 * If you do not want the response cached, use store_miss; and/or
5338 * if you do not want a hit on a cached response, use send_hit.
5344 DEFAULT_DOC: By default, this directive is unused and has no effect.
5345 LOC: Config.accessList.sendHit
5347 Responses denied by this directive will not be served from the cache
5348 (but may still be cached, see store_miss). This directive has no
5349 effect on the responses it allows and on the cached objects.
5351 Please see the "cache" directive for a summary of differences among
5352 store_miss, send_hit, and cache directives.
5354 Unlike the "cache" directive, send_hit only supports fast acl
5355 types. See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
5359 # apply custom Store ID mapping to some URLs
5360 acl MapMe dstdomain .c.example.com
5361 store_id_program ...
5362 store_id_access allow MapMe
5364 # but prevent caching of special responses
5365 # such as 302 redirects that cause StoreID loops
5366 acl Ordinary http_status 200-299
5367 store_miss deny MapMe !Ordinary
5369 # and do not serve any previously stored special responses
5370 # from the cache (in case they were already cached before
5371 # the above store_miss rule was in effect).
5372 send_hit deny MapMe !Ordinary
5378 DEFAULT_DOC: By default, this directive is unused and has no effect.
5379 LOC: Config.accessList.storeMiss
5381 Responses denied by this directive will not be cached (but may still
5382 be served from the cache, see send_hit). This directive has no
5383 effect on the responses it allows and on the already cached responses.
5385 Please see the "cache" directive for a summary of differences among
5386 store_miss, send_hit, and cache directives. See the
5387 send_hit directive for a usage example.
5389 Unlike the "cache" directive, store_miss only supports fast acl
5390 types. See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
5396 LOC: Config.maxStale
5399 This option puts an upper limit on how stale content Squid
5400 will serve from the cache if cache validation fails.
5401 Can be overriden by the refresh_pattern max-stale option.
5404 NAME: refresh_pattern
5405 TYPE: refreshpattern
5409 usage: refresh_pattern [-i] regex min percent max [options]
5411 By default, regular expressions are CASE-SENSITIVE. To make
5412 them case-insensitive, use the -i option.
5414 'Min' is the time (in minutes) an object without an explicit
5415 expiry time should be considered fresh. The recommended
5416 value is 0, any higher values may cause dynamic applications
5417 to be erroneously cached unless the application designer
5418 has taken the appropriate actions.
5420 'Percent' is a percentage of the objects age (time since last
5421 modification age) an object without explicit expiry time
5422 will be considered fresh.
5424 'Max' is an upper limit on how long objects without an explicit
5425 expiry time will be considered fresh.
5427 options: override-expire
5437 override-expire enforces min age even if the server
5438 sent an explicit expiry time (e.g., with the
5439 Expires: header or Cache-Control: max-age). Doing this
5440 VIOLATES the HTTP standard. Enabling this feature
5441 could make you liable for problems which it causes.
5443 Note: override-expire does not enforce staleness - it only extends
5444 freshness / min. If the server returns a Expires time which
5445 is longer than your max time, Squid will still consider
5446 the object fresh for that period of time.
5448 override-lastmod enforces min age even on objects
5449 that were modified recently.
5451 reload-into-ims changes a client no-cache or ``reload''
5452 request for a cached entry into a conditional request using
5453 If-Modified-Since and/or If-None-Match headers, provided the
5454 cached entry has a Last-Modified and/or a strong ETag header.
5455 Doing this VIOLATES the HTTP standard. Enabling this feature
5456 could make you liable for problems which it causes.
5458 ignore-reload ignores a client no-cache or ``reload''
5459 header. Doing this VIOLATES the HTTP standard. Enabling
5460 this feature could make you liable for problems which
5463 ignore-no-store ignores any ``Cache-control: no-store''
5464 headers received from a server. Doing this VIOLATES
5465 the HTTP standard. Enabling this feature could make you
5466 liable for problems which it causes.
5468 ignore-private ignores any ``Cache-control: private''
5469 headers received from a server. Doing this VIOLATES
5470 the HTTP standard. Enabling this feature could make you
5471 liable for problems which it causes.
5473 refresh-ims causes squid to contact the origin server
5474 when a client issues an If-Modified-Since request. This
5475 ensures that the client will receive an updated version
5476 if one is available.
5478 store-stale stores responses even if they don't have explicit
5479 freshness or a validator (i.e., Last-Modified or an ETag)
5480 present, or if they're already stale. By default, Squid will
5481 not cache such responses because they usually can't be
5482 reused. Note that such responses will be stale by default.
5484 max-stale=NN provide a maximum staleness factor. Squid won't
5485 serve objects more stale than this even if it failed to
5486 validate the object. Default: use the max_stale global limit.
5488 Basically a cached object is:
5490 FRESH if expire > now, else STALE
5492 FRESH if lm-factor < percent, else STALE
5496 The refresh_pattern lines are checked in the order listed here.
5497 The first entry which matches is used. If none of the entries
5498 match the default will be used.
5500 Note, you must uncomment all the default lines if you want
5501 to change one. The default setting is only active if none is
5507 # Add any of your own refresh_pattern entries above these.
5509 refresh_pattern ^ftp: 1440 20% 10080
5510 refresh_pattern ^gopher: 1440 0% 1440
5511 refresh_pattern -i (/cgi-bin/|\?) 0 0% 0
5512 refresh_pattern . 0 20% 4320
5516 NAME: quick_abort_min
5520 LOC: Config.quickAbort.min
5523 NAME: quick_abort_max
5527 LOC: Config.quickAbort.max
5530 NAME: quick_abort_pct
5534 LOC: Config.quickAbort.pct
5536 The cache by default continues downloading aborted requests
5537 which are almost completed (less than 16 KB remaining). This
5538 may be undesirable on slow (e.g. SLIP) links and/or very busy
5539 caches. Impatient users may tie up file descriptors and
5540 bandwidth by repeatedly requesting and immediately aborting
5543 When the user aborts a request, Squid will check the
5544 quick_abort values to the amount of data transferred until
5547 If the transfer has less than 'quick_abort_min' KB remaining,
5548 it will finish the retrieval.
5550 If the transfer has more than 'quick_abort_max' KB remaining,
5551 it will abort the retrieval.
5553 If more than 'quick_abort_pct' of the transfer has completed,
5554 it will finish the retrieval.
5556 If you do not want any retrieval to continue after the client
5557 has aborted, set both 'quick_abort_min' and 'quick_abort_max'
5560 If you want retrievals to always continue if they are being
5561 cached set 'quick_abort_min' to '-1 KB'.
5564 NAME: read_ahead_gap
5565 COMMENT: buffer-size
5567 LOC: Config.readAheadGap
5570 The amount of data the cache will buffer ahead of what has been
5571 sent to the client when retrieving an object from another server.
5575 IFDEF: USE_HTTP_VIOLATIONS
5578 LOC: Config.negativeTtl
5581 Set the Default Time-to-Live (TTL) for failed requests.
5582 Certain types of failures (such as "connection refused" and
5583 "404 Not Found") are able to be negatively-cached for a short time.
5584 Modern web servers should provide Expires: header, however if they
5585 do not this can provide a minimum TTL.
5586 The default is not to cache errors with unknown expiry details.
5588 Note that this is different from negative caching of DNS lookups.
5590 WARNING: Doing this VIOLATES the HTTP standard. Enabling
5591 this feature could make you liable for problems which it
5595 NAME: positive_dns_ttl
5598 LOC: Config.positiveDnsTtl
5601 Upper limit on how long Squid will cache positive DNS responses.
5602 Default is 6 hours (360 minutes). This directive must be set
5603 larger than negative_dns_ttl.
5606 NAME: negative_dns_ttl
5609 LOC: Config.negativeDnsTtl
5612 Time-to-Live (TTL) for negative caching of failed DNS lookups.
5613 This also sets the lower cache limit on positive lookups.
5614 Minimum value is 1 second, and it is not recommendable to go
5615 much below 10 seconds.
5618 NAME: range_offset_limit
5619 COMMENT: size [acl acl...]
5621 LOC: Config.rangeOffsetLimit
5624 usage: (size) [units] [[!]aclname]
5626 Sets an upper limit on how far (number of bytes) into the file
5627 a Range request may be to cause Squid to prefetch the whole file.
5628 If beyond this limit, Squid forwards the Range request as it is and
5629 the result is NOT cached.
5631 This is to stop a far ahead range request (lets say start at 17MB)
5632 from making Squid fetch the whole object up to that point before
5633 sending anything to the client.
5635 Multiple range_offset_limit lines may be specified, and they will
5636 be searched from top to bottom on each request until a match is found.
5637 The first match found will be used. If no line matches a request, the
5638 default limit of 0 bytes will be used.
5640 'size' is the limit specified as a number of units.
5642 'units' specifies whether to use bytes, KB, MB, etc.
5643 If no units are specified bytes are assumed.
5645 A size of 0 causes Squid to never fetch more than the
5646 client requested. (default)
5648 A size of 'none' causes Squid to always fetch the object from the
5649 beginning so it may cache the result. (2.0 style)
5651 'aclname' is the name of a defined ACL.
5653 NP: Using 'none' as the byte value here will override any quick_abort settings
5654 that may otherwise apply to the range request. The range request will
5655 be fully fetched from start to finish regardless of the client
5656 actions. This affects bandwidth usage.
5659 NAME: minimum_expiry_time
5662 LOC: Config.minimum_expiry_time
5665 The minimum caching time according to (Expires - Date)
5666 headers Squid honors if the object can't be revalidated.
5667 The default is 60 seconds.
5669 In reverse proxy environments it might be desirable to honor
5670 shorter object lifetimes. It is most likely better to make
5671 your server return a meaningful Last-Modified header however.
5673 In ESI environments where page fragments often have short
5674 lifetimes, this will often be best set to 0.
5677 NAME: store_avg_object_size
5681 LOC: Config.Store.avgObjectSize
5683 Average object size, used to estimate number of objects your
5684 cache can hold. The default is 13 KB.
5686 This is used to pre-seed the cache index memory allocation to
5687 reduce expensive reallocate operations while handling clients
5688 traffic. Too-large values may result in memory allocation during
5689 peak traffic, too-small values will result in wasted memory.
5691 Check the cache manager 'info' report metrics for the real
5692 object sizes seen by your Squid before tuning this.
5695 NAME: store_objects_per_bucket
5698 LOC: Config.Store.objectsPerBucket
5700 Target number of objects per bucket in the store hash table.
5701 Lowering this value increases the total number of buckets and
5702 also the storage maintenance rate. The default is 20.
5707 -----------------------------------------------------------------------------
5710 NAME: request_header_max_size
5714 LOC: Config.maxRequestHeaderSize
5716 This specifies the maximum size for HTTP headers in a request.
5717 Request headers are usually relatively small (about 512 bytes).
5718 Placing a limit on the request header size will catch certain
5719 bugs (for example with persistent connections) and possibly
5720 buffer-overflow or denial-of-service attacks.
5723 NAME: reply_header_max_size
5727 LOC: Config.maxReplyHeaderSize
5729 This specifies the maximum size for HTTP headers in a reply.
5730 Reply headers are usually relatively small (about 512 bytes).
5731 Placing a limit on the reply header size will catch certain
5732 bugs (for example with persistent connections) and possibly
5733 buffer-overflow or denial-of-service attacks.
5736 NAME: request_body_max_size
5740 DEFAULT_DOC: No limit.
5741 LOC: Config.maxRequestBodySize
5743 This specifies the maximum size for an HTTP request body.
5744 In other words, the maximum size of a PUT/POST request.
5745 A user who attempts to send a request with a body larger
5746 than this limit receives an "Invalid Request" error message.
5747 If you set this parameter to a zero (the default), there will
5748 be no limit imposed.
5750 See also client_request_buffer_max_size for an alternative
5751 limitation on client uploads which can be configured.
5754 NAME: client_request_buffer_max_size
5758 LOC: Config.maxRequestBufferSize
5760 This specifies the maximum buffer size of a client request.
5761 It prevents squid eating too much memory when somebody uploads
5766 IFDEF: USE_HTTP_VIOLATIONS
5769 DEFAULT_DOC: Obey RFC 2616.
5770 LOC: Config.accessList.brokenPosts
5772 A list of ACL elements which, if matched, causes Squid to send
5773 an extra CRLF pair after the body of a PUT/POST request.
5775 Some HTTP servers has broken implementations of PUT/POST,
5776 and rely on an extra CRLF pair sent by some WWW clients.
5778 Quote from RFC2616 section 4.1 on this matter:
5780 Note: certain buggy HTTP/1.0 client implementations generate an
5781 extra CRLF's after a POST request. To restate what is explicitly
5782 forbidden by the BNF, an HTTP/1.1 client must not preface or follow
5783 a request with an extra CRLF.
5785 This clause only supports fast acl types.
5786 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
5789 acl buggy_server url_regex ^http://....
5790 broken_posts allow buggy_server
5793 NAME: adaptation_uses_indirect_client icap_uses_indirect_client
5796 IFDEF: FOLLOW_X_FORWARDED_FOR&&USE_ADAPTATION
5798 LOC: Adaptation::Config::use_indirect_client
5800 Controls whether the indirect client IP address (instead of the direct
5801 client IP address) is passed to adaptation services.
5803 See also: follow_x_forwarded_for adaptation_send_client_ip
5807 IFDEF: USE_HTTP_VIOLATIONS
5811 LOC: Config.onoff.via
5813 If set (default), Squid will include a Via header in requests and
5814 replies as required by RFC2616.
5820 LOC: Config.onoff.ie_refresh
5823 Microsoft Internet Explorer up until version 5.5 Service
5824 Pack 1 has an issue with transparent proxies, wherein it
5825 is impossible to force a refresh. Turning this on provides
5826 a partial fix to the problem, by causing all IMS-REFRESH
5827 requests from older IE versions to check the origin server
5828 for fresh content. This reduces hit ratio by some amount
5829 (~10% in my experience), but allows users to actually get
5830 fresh content when they want it. Note because Squid
5831 cannot tell if the user is using 5.5 or 5.5SP1, the behavior
5832 of 5.5 is unchanged from old versions of Squid (i.e. a
5833 forced refresh is impossible). Newer versions of IE will,
5834 hopefully, continue to have the new behavior and will be
5835 handled based on that assumption. This option defaults to
5836 the old Squid behavior, which is better for hit ratios but
5837 worse for clients using IE, if they need to be able to
5838 force fresh content.
5841 NAME: vary_ignore_expire
5844 LOC: Config.onoff.vary_ignore_expire
5847 Many HTTP servers supporting Vary gives such objects
5848 immediate expiry time with no cache-control header
5849 when requested by a HTTP/1.0 client. This option
5850 enables Squid to ignore such expiry times until
5851 HTTP/1.1 is fully implemented.
5853 WARNING: If turned on this may eventually cause some
5854 varying objects not intended for caching to get cached.
5857 NAME: request_entities
5859 LOC: Config.onoff.request_entities
5862 Squid defaults to deny GET and HEAD requests with request entities,
5863 as the meaning of such requests are undefined in the HTTP standard
5864 even if not explicitly forbidden.
5866 Set this directive to on if you have clients which insists
5867 on sending request entities in GET or HEAD requests. But be warned
5868 that there is server software (both proxies and web servers) which
5869 can fail to properly process this kind of request which may make you
5870 vulnerable to cache pollution attacks if enabled.
5873 NAME: request_header_access
5874 IFDEF: USE_HTTP_VIOLATIONS
5875 TYPE: http_header_access
5876 LOC: Config.request_header_access
5878 DEFAULT_DOC: No limits.
5880 Usage: request_header_access header_name allow|deny [!]aclname ...
5882 WARNING: Doing this VIOLATES the HTTP standard. Enabling
5883 this feature could make you liable for problems which it
5886 This option replaces the old 'anonymize_headers' and the
5887 older 'http_anonymizer' option with something that is much
5888 more configurable. A list of ACLs for each header name allows
5889 removal of specific header fields under specific conditions.
5891 This option only applies to outgoing HTTP request headers (i.e.,
5892 headers sent by Squid to the next HTTP hop such as a cache peer
5893 or an origin server). The option has no effect during cache hit
5894 detection. The equivalent adaptation vectoring point in ICAP
5895 terminology is post-cache REQMOD.
5897 The option is applied to individual outgoing request header
5898 fields. For each request header field F, Squid uses the first
5899 qualifying sets of request_header_access rules:
5901 1. Rules with header_name equal to F's name.
5902 2. Rules with header_name 'Other', provided F's name is not
5903 on the hard-coded list of commonly used HTTP header names.
5904 3. Rules with header_name 'All'.
5906 Within that qualifying rule set, rule ACLs are checked as usual.
5907 If ACLs of an "allow" rule match, the header field is allowed to
5908 go through as is. If ACLs of a "deny" rule match, the header is
5909 removed and request_header_replace is then checked to identify
5910 if the removed header has a replacement. If no rules within the
5911 set have matching ACLs, the header field is left as is.
5913 For example, to achieve the same behavior as the old
5914 'http_anonymizer standard' option, you should use:
5916 request_header_access From deny all
5917 request_header_access Referer deny all
5918 request_header_access User-Agent deny all
5920 Or, to reproduce the old 'http_anonymizer paranoid' feature
5923 request_header_access Authorization allow all
5924 request_header_access Proxy-Authorization allow all
5925 request_header_access Cache-Control allow all
5926 request_header_access Content-Length allow all
5927 request_header_access Content-Type allow all
5928 request_header_access Date allow all
5929 request_header_access Host allow all
5930 request_header_access If-Modified-Since allow all
5931 request_header_access Pragma allow all
5932 request_header_access Accept allow all
5933 request_header_access Accept-Charset allow all
5934 request_header_access Accept-Encoding allow all
5935 request_header_access Accept-Language allow all
5936 request_header_access Connection allow all
5937 request_header_access All deny all
5939 HTTP reply headers are controlled with the reply_header_access directive.
5941 By default, all headers are allowed (no anonymizing is performed).
5944 NAME: reply_header_access
5945 IFDEF: USE_HTTP_VIOLATIONS
5946 TYPE: http_header_access
5947 LOC: Config.reply_header_access
5949 DEFAULT_DOC: No limits.
5951 Usage: reply_header_access header_name allow|deny [!]aclname ...
5953 WARNING: Doing this VIOLATES the HTTP standard. Enabling
5954 this feature could make you liable for problems which it
5957 This option only applies to reply headers, i.e., from the
5958 server to the client.
5960 This is the same as request_header_access, but in the other
5961 direction. Please see request_header_access for detailed
5964 For example, to achieve the same behavior as the old
5965 'http_anonymizer standard' option, you should use:
5967 reply_header_access Server deny all
5968 reply_header_access WWW-Authenticate deny all
5969 reply_header_access Link deny all
5971 Or, to reproduce the old 'http_anonymizer paranoid' feature
5974 reply_header_access Allow allow all
5975 reply_header_access WWW-Authenticate allow all
5976 reply_header_access Proxy-Authenticate allow all
5977 reply_header_access Cache-Control allow all
5978 reply_header_access Content-Encoding allow all
5979 reply_header_access Content-Length allow all
5980 reply_header_access Content-Type allow all
5981 reply_header_access Date allow all
5982 reply_header_access Expires allow all
5983 reply_header_access Last-Modified allow all
5984 reply_header_access Location allow all
5985 reply_header_access Pragma allow all
5986 reply_header_access Content-Language allow all
5987 reply_header_access Retry-After allow all
5988 reply_header_access Title allow all
5989 reply_header_access Content-Disposition allow all
5990 reply_header_access Connection allow all
5991 reply_header_access All deny all
5993 HTTP request headers are controlled with the request_header_access directive.
5995 By default, all headers are allowed (no anonymizing is
5999 NAME: request_header_replace header_replace
6000 IFDEF: USE_HTTP_VIOLATIONS
6001 TYPE: http_header_replace
6002 LOC: Config.request_header_access
6005 Usage: request_header_replace header_name message
6006 Example: request_header_replace User-Agent Nutscrape/1.0 (CP/M; 8-bit)
6008 This option allows you to change the contents of headers
6009 denied with request_header_access above, by replacing them
6010 with some fixed string.
6012 This only applies to request headers, not reply headers.
6014 By default, headers are removed if denied.
6017 NAME: reply_header_replace
6018 IFDEF: USE_HTTP_VIOLATIONS
6019 TYPE: http_header_replace
6020 LOC: Config.reply_header_access
6023 Usage: reply_header_replace header_name message
6024 Example: reply_header_replace Server Foo/1.0
6026 This option allows you to change the contents of headers
6027 denied with reply_header_access above, by replacing them
6028 with some fixed string.
6030 This only applies to reply headers, not request headers.
6032 By default, headers are removed if denied.
6035 NAME: request_header_add
6036 TYPE: HeaderWithAclList
6037 LOC: Config.request_header_add
6040 Usage: request_header_add field-name field-value acl1 [acl2] ...
6041 Example: request_header_add X-Client-CA "CA=%ssl::>cert_issuer" all
6043 This option adds header fields to outgoing HTTP requests (i.e.,
6044 request headers sent by Squid to the next HTTP hop such as a
6045 cache peer or an origin server). The option has no effect during
6046 cache hit detection. The equivalent adaptation vectoring point
6047 in ICAP terminology is post-cache REQMOD.
6049 Field-name is a token specifying an HTTP header name. If a
6050 standard HTTP header name is used, Squid does not check whether
6051 the new header conflicts with any existing headers or violates
6052 HTTP rules. If the request to be modified already contains a
6053 field with the same name, the old field is preserved but the
6054 header field values are not merged.
6056 Field-value is either a token or a quoted string. If quoted
6057 string format is used, then the surrounding quotes are removed
6058 while escape sequences and %macros are processed.
6060 In theory, all of the logformat codes can be used as %macros.
6061 However, unlike logging (which happens at the very end of
6062 transaction lifetime), the transaction may not yet have enough
6063 information to expand a macro when the new header value is needed.
6064 And some information may already be available to Squid but not yet
6065 committed where the macro expansion code can access it (report
6066 such instances!). The macro will be expanded into a single dash
6067 ('-') in such cases. Not all macros have been tested.
6069 One or more Squid ACLs may be specified to restrict header
6070 injection to matching requests. As always in squid.conf, all
6071 ACLs in an option ACL list must be satisfied for the insertion
6072 to happen. The request_header_add option supports fast ACLs
6081 This option used to log custom information about the master
6082 transaction. For example, an admin may configure Squid to log
6083 which "user group" the transaction belongs to, where "user group"
6084 will be determined based on a set of ACLs and not [just]
6085 authentication information.
6086 Values of key/value pairs can be logged using %{key}note macros:
6088 note key value acl ...
6089 logformat myFormat ... %{key}note ...
6092 NAME: relaxed_header_parser
6093 COMMENT: on|off|warn
6095 LOC: Config.onoff.relaxed_header_parser
6098 In the default "on" setting Squid accepts certain forms
6099 of non-compliant HTTP messages where it is unambiguous
6100 what the sending application intended even if the message
6101 is not correctly formatted. The messages is then normalized
6102 to the correct form when forwarded by Squid.
6104 If set to "warn" then a warning will be emitted in cache.log
6105 each time such HTTP error is encountered.
6107 If set to "off" then such HTTP errors will cause the request
6108 or response to be rejected.
6111 NAME: collapsed_forwarding
6114 LOC: Config.onoff.collapsed_forwarding
6117 This option controls whether Squid is allowed to merge multiple
6118 potentially cachable requests for the same URI before Squid knows
6119 whether the response is going to be cachable.
6121 This feature is disabled by default: Enabling collapsed forwarding
6122 needlessly delays forwarding requests that look cachable (when they are
6123 collapsed) but then need to be forwarded individually anyway because
6124 they end up being for uncachable content. However, in some cases, such
6125 as accelleration of highly cachable content with periodic or groupped
6126 expiration times, the gains from collapsing [large volumes of
6127 simultenous refresh requests] outweigh losses from such delays.
6132 -----------------------------------------------------------------------------
6135 NAME: forward_timeout
6138 LOC: Config.Timeout.forward
6141 This parameter specifies how long Squid should at most attempt in
6142 finding a forwarding path for the request before giving up.
6145 NAME: connect_timeout
6148 LOC: Config.Timeout.connect
6151 This parameter specifies how long to wait for the TCP connect to
6152 the requested server or peer to complete before Squid should
6153 attempt to find another path where to forward the request.
6156 NAME: peer_connect_timeout
6159 LOC: Config.Timeout.peer_connect
6162 This parameter specifies how long to wait for a pending TCP
6163 connection to a peer cache. The default is 30 seconds. You
6164 may also set different timeout values for individual neighbors
6165 with the 'connect-timeout' option on a 'cache_peer' line.
6171 LOC: Config.Timeout.read
6174 Applied on peer server connections.
6176 After each successful read(), the timeout will be extended by this
6177 amount. If no data is read again after this amount of time,
6178 the request is aborted and logged with ERR_READ_TIMEOUT.
6180 The default is 15 minutes.
6186 LOC: Config.Timeout.write
6189 This timeout is tracked for all connections that have data
6190 available for writing and are waiting for the socket to become
6191 ready. After each successful write, the timeout is extended by
6192 the configured amount. If Squid has data to write but the
6193 connection is not ready for the configured duration, the
6194 transaction associated with the connection is terminated. The
6195 default is 15 minutes.
6198 NAME: request_timeout
6200 LOC: Config.Timeout.request
6203 How long to wait for complete HTTP request headers after initial
6204 connection establishment.
6207 NAME: request_start_timeout
6209 LOC: Config.Timeout.request_start_timeout
6212 How long to wait for the first request byte after initial
6213 connection establishment.
6216 NAME: client_idle_pconn_timeout persistent_request_timeout
6218 LOC: Config.Timeout.clientIdlePconn
6221 How long to wait for the next HTTP request on a persistent
6222 client connection after the previous request completes.
6225 NAME: ftp_client_idle_timeout
6227 LOC: Config.Timeout.ftpClientIdle
6230 How long to wait for an FTP request on a connection to Squid ftp_port.
6231 Many FTP clients do not deal with idle connection closures well,
6232 necessitating a longer default timeout than client_idle_pconn_timeout
6233 used for incoming HTTP requests.
6236 NAME: client_lifetime
6239 LOC: Config.Timeout.lifetime
6242 The maximum amount of time a client (browser) is allowed to
6243 remain connected to the cache process. This protects the Cache
6244 from having a lot of sockets (and hence file descriptors) tied up
6245 in a CLOSE_WAIT state from remote clients that go away without
6246 properly shutting down (either because of a network failure or
6247 because of a poor client implementation). The default is one
6250 NOTE: The default value is intended to be much larger than any
6251 client would ever need to be connected to your cache. You
6252 should probably change client_lifetime only as a last resort.
6253 If you seem to have many client connections tying up
6254 filedescriptors, we recommend first tuning the read_timeout,
6255 request_timeout, persistent_request_timeout and quick_abort values.
6258 NAME: pconn_lifetime
6261 LOC: Config.Timeout.pconnLifetime
6264 Desired maximum lifetime of a persistent connection.
6265 When set, Squid will close a now-idle persistent connection that
6266 exceeded configured lifetime instead of moving the connection into
6267 the idle connection pool (or equivalent). No effect on ongoing/active
6268 transactions. Connection lifetime is the time period from the
6269 connection acceptance or opening time until "now".
6271 This limit is useful in environments with long-lived connections
6272 where Squid configuration or environmental factors change during a
6273 single connection lifetime. If unrestricted, some connections may
6274 last for hours and even days, ignoring those changes that should
6275 have affected their behavior or their existence.
6277 Currently, a new lifetime value supplied via Squid reconfiguration
6278 has no effect on already idle connections unless they become busy.
6280 When set to '0' this limit is not used.
6283 NAME: half_closed_clients
6285 LOC: Config.onoff.half_closed_clients
6288 Some clients may shutdown the sending side of their TCP
6289 connections, while leaving their receiving sides open. Sometimes,
6290 Squid can not tell the difference between a half-closed and a
6291 fully-closed TCP connection.
6293 By default, Squid will immediately close client connections when
6294 read(2) returns "no more data to read."
6296 Change this option to 'on' and Squid will keep open connections
6297 until a read(2) or write(2) on the socket returns an error.
6298 This may show some benefits for reverse proxies. But if not
6299 it is recommended to leave OFF.
6302 NAME: server_idle_pconn_timeout pconn_timeout
6304 LOC: Config.Timeout.serverIdlePconn
6307 Timeout for idle persistent connections to servers and other
6314 LOC: Ident::TheConfig.timeout
6317 Maximum time to wait for IDENT lookups to complete.
6319 If this is too high, and you enabled IDENT lookups from untrusted
6320 users, you might be susceptible to denial-of-service by having
6321 many ident requests going at once.
6324 NAME: shutdown_lifetime
6327 LOC: Config.shutdownLifetime
6330 When SIGTERM or SIGHUP is received, the cache is put into
6331 "shutdown pending" mode until all active sockets are closed.
6332 This value is the lifetime to set for all open descriptors
6333 during shutdown mode. Any active clients after this many
6334 seconds will receive a 'timeout' message.
6338 ADMINISTRATIVE PARAMETERS
6339 -----------------------------------------------------------------------------
6345 LOC: Config.adminEmail
6347 Email-address of local cache manager who will receive
6348 mail if the cache dies. The default is "webmaster".
6354 LOC: Config.EmailFrom
6356 From: email-address for mail sent when the cache dies.
6357 The default is to use 'squid@unique_hostname'.
6359 See also: unique_hostname directive.
6365 LOC: Config.EmailProgram
6367 Email program used to send mail if the cache dies.
6368 The default is "mail". The specified program must comply
6369 with the standard Unix mail syntax:
6370 mail-program recipient < mailfile
6372 Optional command line options can be specified.
6375 NAME: cache_effective_user
6377 DEFAULT: @DEFAULT_CACHE_EFFECTIVE_USER@
6378 LOC: Config.effectiveUser
6380 If you start Squid as root, it will change its effective/real
6381 UID/GID to the user specified below. The default is to change
6382 to UID of @DEFAULT_CACHE_EFFECTIVE_USER@.
6383 see also; cache_effective_group
6386 NAME: cache_effective_group
6389 DEFAULT_DOC: Use system group memberships of the cache_effective_user account
6390 LOC: Config.effectiveGroup
6392 Squid sets the GID to the effective user's default group ID
6393 (taken from the password file) and supplementary group list
6394 from the groups membership.
6396 If you want Squid to run with a specific GID regardless of
6397 the group memberships of the effective user then set this
6398 to the group (or GID) you want Squid to run as. When set
6399 all other group privileges of the effective user are ignored
6400 and only this GID is effective. If Squid is not started as
6401 root the user starting Squid MUST be member of the specified
6404 This option is not recommended by the Squid Team.
6405 Our preference is for administrators to configure a secure
6406 user account for squid with UID/GID matching system policies.
6409 NAME: httpd_suppress_version_string
6413 LOC: Config.onoff.httpd_suppress_version_string
6415 Suppress Squid version string info in HTTP headers and HTML error pages.
6418 NAME: visible_hostname
6420 LOC: Config.visibleHostname
6422 DEFAULT_DOC: Automatically detect the system host name
6424 If you want to present a special hostname in error messages, etc,
6425 define this. Otherwise, the return value of gethostname()
6426 will be used. If you have multiple caches in a cluster and
6427 get errors about IP-forwarding you must set them to have individual
6428 names with this setting.
6431 NAME: unique_hostname
6433 LOC: Config.uniqueHostname
6435 DEFAULT_DOC: Copy the value from visible_hostname
6437 If you want to have multiple machines with the same
6438 'visible_hostname' you must give each machine a different
6439 'unique_hostname' so forwarding loops can be detected.
6442 NAME: hostname_aliases
6444 LOC: Config.hostnameAliases
6447 A list of other DNS names your cache has.
6455 Minimum umask which should be enforced while the proxy
6456 is running, in addition to the umask set at startup.
6458 For a traditional octal representation of umasks, start
6463 OPTIONS FOR THE CACHE REGISTRATION SERVICE
6464 -----------------------------------------------------------------------------
6466 This section contains parameters for the (optional) cache
6467 announcement service. This service is provided to help
6468 cache administrators locate one another in order to join or
6469 create cache hierarchies.
6471 An 'announcement' message is sent (via UDP) to the registration
6472 service by Squid. By default, the announcement message is NOT
6473 SENT unless you enable it with 'announce_period' below.
6475 The announcement message includes your hostname, plus the
6476 following information from this configuration file:
6482 All current information is processed regularly and made
6483 available on the Web at http://www.ircache.net/Cache/Tracker/.
6486 NAME: announce_period
6488 LOC: Config.Announce.period
6490 DEFAULT_DOC: Announcement messages disabled.
6492 This is how frequently to send cache announcements.
6494 To enable announcing your cache, just set an announce period.
6497 announce_period 1 day
6502 DEFAULT: tracker.ircache.net
6503 LOC: Config.Announce.host
6505 Set the hostname where announce registration messages will be sent.
6507 See also announce_port and announce_file
6513 LOC: Config.Announce.file
6515 The contents of this file will be included in the announce
6516 registration messages.
6522 LOC: Config.Announce.port
6524 Set the port where announce registration messages will be sent.
6526 See also announce_host and announce_file
6530 HTTPD-ACCELERATOR OPTIONS
6531 -----------------------------------------------------------------------------
6534 NAME: httpd_accel_surrogate_id
6537 DEFAULT_DOC: visible_hostname is used if no specific ID is set.
6538 LOC: Config.Accel.surrogate_id
6540 Surrogates (http://www.esi.org/architecture_spec_1.0.html)
6541 need an identification token to allow control targeting. Because
6542 a farm of surrogates may all perform the same tasks, they may share
6543 an identification token.
6546 NAME: http_accel_surrogate_remote
6550 LOC: Config.onoff.surrogate_is_remote
6552 Remote surrogates (such as those in a CDN) honour the header
6553 "Surrogate-Control: no-store-remote".
6555 Set this to on to have squid behave as a remote surrogate.
6559 IFDEF: USE_SQUID_ESI
6560 COMMENT: libxml2|expat|custom
6562 LOC: ESIParser::Type
6565 ESI markup is not strictly XML compatible. The custom ESI parser
6566 will give higher performance, but cannot handle non ASCII character
6571 DELAY POOL PARAMETERS
6572 -----------------------------------------------------------------------------
6576 TYPE: delay_pool_count
6578 IFDEF: USE_DELAY_POOLS
6581 This represents the number of delay pools to be used. For example,
6582 if you have one class 2 delay pool and one class 3 delays pool, you
6583 have a total of 2 delay pools.
6585 See also delay_parameters, delay_class, delay_access for pool
6586 configuration details.
6590 TYPE: delay_pool_class
6592 IFDEF: USE_DELAY_POOLS
6595 This defines the class of each delay pool. There must be exactly one
6596 delay_class line for each delay pool. For example, to define two
6597 delay pools, one of class 2 and one of class 3, the settings above
6601 delay_pools 4 # 4 delay pools
6602 delay_class 1 2 # pool 1 is a class 2 pool
6603 delay_class 2 3 # pool 2 is a class 3 pool
6604 delay_class 3 4 # pool 3 is a class 4 pool
6605 delay_class 4 5 # pool 4 is a class 5 pool
6607 The delay pool classes are:
6609 class 1 Everything is limited by a single aggregate
6612 class 2 Everything is limited by a single aggregate
6613 bucket as well as an "individual" bucket chosen
6614 from bits 25 through 32 of the IPv4 address.
6616 class 3 Everything is limited by a single aggregate
6617 bucket as well as a "network" bucket chosen
6618 from bits 17 through 24 of the IP address and a
6619 "individual" bucket chosen from bits 17 through
6620 32 of the IPv4 address.
6622 class 4 Everything in a class 3 delay pool, with an
6623 additional limit on a per user basis. This
6624 only takes effect if the username is established
6625 in advance - by forcing authentication in your
6628 class 5 Requests are grouped according their tag (see
6629 external_acl's tag= reply).
6632 Each pool also requires a delay_parameters directive to configure the pool size
6633 and speed limits used whenever the pool is applied to a request. Along with
6634 a set of delay_access directives to determine when it is used.
6636 NOTE: If an IP address is a.b.c.d
6637 -> bits 25 through 32 are "d"
6638 -> bits 17 through 24 are "c"
6639 -> bits 17 through 32 are "c * 256 + d"
6641 NOTE-2: Due to the use of bitmasks in class 2,3,4 pools they only apply to
6642 IPv4 traffic. Class 1 and 5 pools may be used with IPv6 traffic.
6644 This clause only supports fast acl types.
6645 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
6647 See also delay_parameters and delay_access.
6651 TYPE: delay_pool_access
6653 DEFAULT_DOC: Deny using the pool, unless allow rules exist in squid.conf for the pool.
6654 IFDEF: USE_DELAY_POOLS
6657 This is used to determine which delay pool a request falls into.
6659 delay_access is sorted per pool and the matching starts with pool 1,
6660 then pool 2, ..., and finally pool N. The first delay pool where the
6661 request is allowed is selected for the request. If it does not allow
6662 the request to any pool then the request is not delayed (default).
6664 For example, if you want some_big_clients in delay
6665 pool 1 and lotsa_little_clients in delay pool 2:
6667 delay_access 1 allow some_big_clients
6668 delay_access 1 deny all
6669 delay_access 2 allow lotsa_little_clients
6670 delay_access 2 deny all
6671 delay_access 3 allow authenticated_clients
6673 See also delay_parameters and delay_class.
6677 NAME: delay_parameters
6678 TYPE: delay_pool_rates
6680 IFDEF: USE_DELAY_POOLS
6683 This defines the parameters for a delay pool. Each delay pool has
6684 a number of "buckets" associated with it, as explained in the
6685 description of delay_class.
6687 For a class 1 delay pool, the syntax is:
6689 delay_parameters pool aggregate
6691 For a class 2 delay pool:
6693 delay_parameters pool aggregate individual
6695 For a class 3 delay pool:
6697 delay_parameters pool aggregate network individual
6699 For a class 4 delay pool:
6701 delay_parameters pool aggregate network individual user
6703 For a class 5 delay pool:
6705 delay_parameters pool tagrate
6707 The option variables are:
6709 pool a pool number - ie, a number between 1 and the
6710 number specified in delay_pools as used in
6713 aggregate the speed limit parameters for the aggregate bucket
6716 individual the speed limit parameters for the individual
6717 buckets (class 2, 3).
6719 network the speed limit parameters for the network buckets
6722 user the speed limit parameters for the user buckets
6725 tagrate the speed limit parameters for the tag buckets
6728 A pair of delay parameters is written restore/maximum, where restore is
6729 the number of bytes (not bits - modem and network speeds are usually
6730 quoted in bits) per second placed into the bucket, and maximum is the
6731 maximum number of bytes which can be in the bucket at any time.
6733 There must be one delay_parameters line for each delay pool.
6736 For example, if delay pool number 1 is a class 2 delay pool as in the
6737 above example, and is being used to strictly limit each host to 64Kbit/sec
6738 (plus overheads), with no overall limit, the line is:
6740 delay_parameters 1 none 8000/8000
6742 Note that 8 x 8K Byte/sec -> 64K bit/sec.
6744 Note that the word 'none' is used to represent no limit.
6747 And, if delay pool number 2 is a class 3 delay pool as in the above
6748 example, and you want to limit it to a total of 256Kbit/sec (strict limit)
6749 with each 8-bit network permitted 64Kbit/sec (strict limit) and each
6750 individual host permitted 4800bit/sec with a bucket maximum size of 64Kbits
6751 to permit a decent web page to be downloaded at a decent speed
6752 (if the network is not being limited due to overuse) but slow down
6753 large downloads more significantly:
6755 delay_parameters 2 32000/32000 8000/8000 600/8000
6757 Note that 8 x 32K Byte/sec -> 256K bit/sec.
6758 8 x 8K Byte/sec -> 64K bit/sec.
6759 8 x 600 Byte/sec -> 4800 bit/sec.
6762 Finally, for a class 4 delay pool as in the example - each user will
6763 be limited to 128Kbits/sec no matter how many workstations they are logged into.:
6765 delay_parameters 4 32000/32000 8000/8000 600/64000 16000/16000
6768 See also delay_class and delay_access.
6772 NAME: delay_initial_bucket_level
6773 COMMENT: (percent, 0-100)
6776 IFDEF: USE_DELAY_POOLS
6777 LOC: Config.Delay.initial
6779 The initial bucket percentage is used to determine how much is put
6780 in each bucket when squid starts, is reconfigured, or first notices
6781 a host accessing it (in class 2 and class 3, individual hosts and
6782 networks only have buckets associated with them once they have been
6787 CLIENT DELAY POOL PARAMETERS
6788 -----------------------------------------------------------------------------
6791 NAME: client_delay_pools
6792 TYPE: client_delay_pool_count
6794 IFDEF: USE_DELAY_POOLS
6795 LOC: Config.ClientDelay
6797 This option specifies the number of client delay pools used. It must
6798 preceed other client_delay_* options.
6801 client_delay_pools 2
6803 See also client_delay_parameters and client_delay_access.
6806 NAME: client_delay_initial_bucket_level
6807 COMMENT: (percent, 0-no_limit)
6810 IFDEF: USE_DELAY_POOLS
6811 LOC: Config.ClientDelay.initial
6813 This option determines the initial bucket size as a percentage of
6814 max_bucket_size from client_delay_parameters. Buckets are created
6815 at the time of the "first" connection from the matching IP. Idle
6816 buckets are periodically deleted up.
6818 You can specify more than 100 percent but note that such "oversized"
6819 buckets are not refilled until their size goes down to max_bucket_size
6820 from client_delay_parameters.
6823 client_delay_initial_bucket_level 50
6826 NAME: client_delay_parameters
6827 TYPE: client_delay_pool_rates
6829 IFDEF: USE_DELAY_POOLS
6830 LOC: Config.ClientDelay
6833 This option configures client-side bandwidth limits using the
6836 client_delay_parameters pool speed_limit max_bucket_size
6838 pool is an integer ID used for client_delay_access matching.
6840 speed_limit is bytes added to the bucket per second.
6842 max_bucket_size is the maximum size of a bucket, enforced after any
6843 speed_limit additions.
6845 Please see the delay_parameters option for more information and
6849 client_delay_parameters 1 1024 2048
6850 client_delay_parameters 2 51200 16384
6852 See also client_delay_access.
6856 NAME: client_delay_access
6857 TYPE: client_delay_pool_access
6859 DEFAULT_DOC: Deny use of the pool, unless allow rules exist in squid.conf for the pool.
6860 IFDEF: USE_DELAY_POOLS
6861 LOC: Config.ClientDelay
6863 This option determines the client-side delay pool for the
6866 client_delay_access pool_ID allow|deny acl_name
6868 All client_delay_access options are checked in their pool ID
6869 order, starting with pool 1. The first checked pool with allowed
6870 request is selected for the request. If no ACL matches or there
6871 are no client_delay_access options, the request bandwidth is not
6874 The ACL-selected pool is then used to find the
6875 client_delay_parameters for the request. Client-side pools are
6876 not used to aggregate clients. Clients are always aggregated
6877 based on their source IP addresses (one bucket per source IP).
6879 This clause only supports fast acl types.
6880 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
6881 Additionally, only the client TCP connection details are available.
6882 ACLs testing HTTP properties will not work.
6884 Please see delay_access for more examples.
6887 client_delay_access 1 allow low_rate_network
6888 client_delay_access 2 allow vips_network
6891 See also client_delay_parameters and client_delay_pools.
6895 WCCPv1 AND WCCPv2 CONFIGURATION OPTIONS
6896 -----------------------------------------------------------------------------
6901 LOC: Config.Wccp.router
6903 DEFAULT_DOC: WCCP disabled.
6906 Use this option to define your WCCP ``home'' router for
6909 wccp_router supports a single WCCP(v1) router
6911 wccp2_router supports multiple WCCPv2 routers
6913 only one of the two may be used at the same time and defines
6914 which version of WCCP to use.
6918 TYPE: IpAddress_list
6919 LOC: Config.Wccp2.router
6921 DEFAULT_DOC: WCCPv2 disabled.
6924 Use this option to define your WCCP ``home'' router for
6927 wccp_router supports a single WCCP(v1) router
6929 wccp2_router supports multiple WCCPv2 routers
6931 only one of the two may be used at the same time and defines
6932 which version of WCCP to use.
6937 LOC: Config.Wccp.version
6941 This directive is only relevant if you need to set up WCCP(v1)
6942 to some very old and end-of-life Cisco routers. In all other
6943 setups it must be left unset or at the default setting.
6944 It defines an internal version in the WCCP(v1) protocol,
6945 with version 4 being the officially documented protocol.
6947 According to some users, Cisco IOS 11.2 and earlier only
6948 support WCCP version 3. If you're using that or an earlier
6949 version of IOS, you may need to change this value to 3, otherwise
6950 do not specify this parameter.
6953 NAME: wccp2_rebuild_wait
6955 LOC: Config.Wccp2.rebuildwait
6959 If this is enabled Squid will wait for the cache dir rebuild to finish
6960 before sending the first wccp2 HereIAm packet
6963 NAME: wccp2_forwarding_method
6965 LOC: Config.Wccp2.forwarding_method
6969 WCCP2 allows the setting of forwarding methods between the
6970 router/switch and the cache. Valid values are as follows:
6972 gre - GRE encapsulation (forward the packet in a GRE/WCCP tunnel)
6973 l2 - L2 redirect (forward the packet using Layer 2/MAC rewriting)
6975 Currently (as of IOS 12.4) cisco routers only support GRE.
6976 Cisco switches only support the L2 redirect assignment method.
6979 NAME: wccp2_return_method
6981 LOC: Config.Wccp2.return_method
6985 WCCP2 allows the setting of return methods between the
6986 router/switch and the cache for packets that the cache
6987 decides not to handle. Valid values are as follows:
6989 gre - GRE encapsulation (forward the packet in a GRE/WCCP tunnel)
6990 l2 - L2 redirect (forward the packet using Layer 2/MAC rewriting)
6992 Currently (as of IOS 12.4) cisco routers only support GRE.
6993 Cisco switches only support the L2 redirect assignment.
6995 If the "ip wccp redirect exclude in" command has been
6996 enabled on the cache interface, then it is still safe for
6997 the proxy server to use a l2 redirect method even if this
6998 option is set to GRE.
7001 NAME: wccp2_assignment_method
7003 LOC: Config.Wccp2.assignment_method
7007 WCCP2 allows the setting of methods to assign the WCCP hash
7008 Valid values are as follows:
7010 hash - Hash assignment
7011 mask - Mask assignment
7013 As a general rule, cisco routers support the hash assignment method
7014 and cisco switches support the mask assignment method.
7019 LOC: Config.Wccp2.info
7020 DEFAULT_IF_NONE: standard 0
7021 DEFAULT_DOC: Use the 'web-cache' standard service.
7024 WCCP2 allows for multiple traffic services. There are two
7025 types: "standard" and "dynamic". The standard type defines
7026 one service id - http (id 0). The dynamic service ids can be from
7027 51 to 255 inclusive. In order to use a dynamic service id
7028 one must define the type of traffic to be redirected; this is done
7029 using the wccp2_service_info option.
7031 The "standard" type does not require a wccp2_service_info option,
7032 just specifying the service id will suffice.
7034 MD5 service authentication can be enabled by adding
7035 "password=<password>" to the end of this service declaration.
7039 wccp2_service standard 0 # for the 'web-cache' standard service
7040 wccp2_service dynamic 80 # a dynamic service type which will be
7041 # fleshed out with subsequent options.
7042 wccp2_service standard 0 password=foo
7045 NAME: wccp2_service_info
7046 TYPE: wccp2_service_info
7047 LOC: Config.Wccp2.info
7051 Dynamic WCCPv2 services require further information to define the
7052 traffic you wish to have diverted.
7056 wccp2_service_info <id> protocol=<protocol> flags=<flag>,<flag>..
7057 priority=<priority> ports=<port>,<port>..
7059 The relevant WCCPv2 flags:
7060 + src_ip_hash, dst_ip_hash
7061 + source_port_hash, dst_port_hash
7062 + src_ip_alt_hash, dst_ip_alt_hash
7063 + src_port_alt_hash, dst_port_alt_hash
7066 The port list can be one to eight entries.
7070 wccp2_service_info 80 protocol=tcp flags=src_ip_hash,ports_source
7071 priority=240 ports=80
7073 Note: the service id must have been defined by a previous
7074 'wccp2_service dynamic <id>' entry.
7079 LOC: Config.Wccp2.weight
7083 Each cache server gets assigned a set of the destination
7084 hash proportional to their weight.
7089 LOC: Config.Wccp.address
7091 DEFAULT_DOC: Address selected by the operating system.
7094 Use this option if you require WCCPv2 to use a specific
7097 The default behavior is to not bind to any specific address.
7102 LOC: Config.Wccp2.address
7104 DEFAULT_DOC: Address selected by the operating system.
7107 Use this option if you require WCCP to use a specific
7110 The default behavior is to not bind to any specific address.
7114 PERSISTENT CONNECTION HANDLING
7115 -----------------------------------------------------------------------------
7117 Also see "pconn_timeout" in the TIMEOUTS section
7120 NAME: client_persistent_connections
7122 LOC: Config.onoff.client_pconns
7125 Persistent connection support for clients.
7126 Squid uses persistent connections (when allowed). You can use
7127 this option to disable persistent connections with clients.
7130 NAME: server_persistent_connections
7132 LOC: Config.onoff.server_pconns
7135 Persistent connection support for servers.
7136 Squid uses persistent connections (when allowed). You can use
7137 this option to disable persistent connections with servers.
7140 NAME: persistent_connection_after_error
7142 LOC: Config.onoff.error_pconns
7145 With this directive the use of persistent connections after
7146 HTTP errors can be disabled. Useful if you have clients
7147 who fail to handle errors on persistent connections proper.
7150 NAME: detect_broken_pconn
7152 LOC: Config.onoff.detect_broken_server_pconns
7155 Some servers have been found to incorrectly signal the use
7156 of HTTP/1.0 persistent connections even on replies not
7157 compatible, causing significant delays. This server problem
7158 has mostly been seen on redirects.
7160 By enabling this directive Squid attempts to detect such
7161 broken replies and automatically assume the reply is finished
7162 after 10 seconds timeout.
7166 CACHE DIGEST OPTIONS
7167 -----------------------------------------------------------------------------
7170 NAME: digest_generation
7171 IFDEF: USE_CACHE_DIGESTS
7173 LOC: Config.onoff.digest_generation
7176 This controls whether the server will generate a Cache Digest
7177 of its contents. By default, Cache Digest generation is
7178 enabled if Squid is compiled with --enable-cache-digests defined.
7181 NAME: digest_bits_per_entry
7182 IFDEF: USE_CACHE_DIGESTS
7184 LOC: Config.digest.bits_per_entry
7187 This is the number of bits of the server's Cache Digest which
7188 will be associated with the Digest entry for a given HTTP
7189 Method and URL (public key) combination. The default is 5.
7192 NAME: digest_rebuild_period
7193 IFDEF: USE_CACHE_DIGESTS
7196 LOC: Config.digest.rebuild_period
7199 This is the wait time between Cache Digest rebuilds.
7202 NAME: digest_rewrite_period
7204 IFDEF: USE_CACHE_DIGESTS
7206 LOC: Config.digest.rewrite_period
7209 This is the wait time between Cache Digest writes to
7213 NAME: digest_swapout_chunk_size
7216 IFDEF: USE_CACHE_DIGESTS
7217 LOC: Config.digest.swapout_chunk_size
7220 This is the number of bytes of the Cache Digest to write to
7221 disk at a time. It defaults to 4096 bytes (4KB), the Squid
7225 NAME: digest_rebuild_chunk_percentage
7226 COMMENT: (percent, 0-100)
7227 IFDEF: USE_CACHE_DIGESTS
7229 LOC: Config.digest.rebuild_chunk_percentage
7232 This is the percentage of the Cache Digest to be scanned at a
7233 time. By default it is set to 10% of the Cache Digest.
7238 -----------------------------------------------------------------------------
7243 LOC: Config.Port.snmp
7245 DEFAULT_DOC: SNMP disabled.
7248 The port number where Squid listens for SNMP requests. To enable
7249 SNMP support set this to a suitable port number. Port number
7250 3401 is often used for the Squid SNMP agent. By default it's
7251 set to "0" (disabled)
7259 LOC: Config.accessList.snmp
7261 DEFAULT_DOC: Deny, unless rules exist in squid.conf.
7264 Allowing or denying access to the SNMP port.
7266 All access to the agent is denied by default.
7269 snmp_access allow|deny [!]aclname ...
7271 This clause only supports fast acl types.
7272 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
7275 snmp_access allow snmppublic localhost
7276 snmp_access deny all
7279 NAME: snmp_incoming_address
7281 LOC: Config.Addrs.snmp_incoming
7283 DEFAULT_DOC: Accept SNMP packets from all machine interfaces.
7286 Just like 'udp_incoming_address', but for the SNMP port.
7288 snmp_incoming_address is used for the SNMP socket receiving
7289 messages from SNMP agents.
7291 The default snmp_incoming_address is to listen on all
7292 available network interfaces.
7295 NAME: snmp_outgoing_address
7297 LOC: Config.Addrs.snmp_outgoing
7299 DEFAULT_DOC: Use snmp_incoming_address or an address selected by the operating system.
7302 Just like 'udp_outgoing_address', but for the SNMP port.
7304 snmp_outgoing_address is used for SNMP packets returned to SNMP
7307 If snmp_outgoing_address is not set it will use the same socket
7308 as snmp_incoming_address. Only change this if you want to have
7309 SNMP replies sent using another address than where this Squid
7310 listens for SNMP queries.
7312 NOTE, snmp_incoming_address and snmp_outgoing_address can not have
7313 the same value since they both use the same port.
7318 -----------------------------------------------------------------------------
7321 NAME: icp_port udp_port
7324 DEFAULT_DOC: ICP disabled.
7325 LOC: Config.Port.icp
7327 The port number where Squid sends and receives ICP queries to
7328 and from neighbor caches. The standard UDP port for ICP is 3130.
7331 icp_port @DEFAULT_ICP_PORT@
7338 DEFAULT_DOC: HTCP disabled.
7339 LOC: Config.Port.htcp
7341 The port number where Squid sends and receives HTCP queries to
7342 and from neighbor caches. To turn it on you want to set it to
7349 NAME: log_icp_queries
7353 LOC: Config.onoff.log_udp
7355 If set, ICP queries are logged to access.log. You may wish
7356 do disable this if your ICP load is VERY high to speed things
7357 up or to simplify log analysis.
7360 NAME: udp_incoming_address
7362 LOC:Config.Addrs.udp_incoming
7364 DEFAULT_DOC: Accept packets from all machine interfaces.
7366 udp_incoming_address is used for UDP packets received from other
7369 The default behavior is to not bind to any specific address.
7371 Only change this if you want to have all UDP queries received on
7372 a specific interface/address.
7374 NOTE: udp_incoming_address is used by the ICP, HTCP, and DNS
7375 modules. Altering it will affect all of them in the same manner.
7377 see also; udp_outgoing_address
7379 NOTE, udp_incoming_address and udp_outgoing_address can not
7380 have the same value since they both use the same port.
7383 NAME: udp_outgoing_address
7385 LOC: Config.Addrs.udp_outgoing
7387 DEFAULT_DOC: Use udp_incoming_address or an address selected by the operating system.
7389 udp_outgoing_address is used for UDP packets sent out to other
7392 The default behavior is to not bind to any specific address.
7394 Instead it will use the same socket as udp_incoming_address.
7395 Only change this if you want to have UDP queries sent using another
7396 address than where this Squid listens for UDP queries from other
7399 NOTE: udp_outgoing_address is used by the ICP, HTCP, and DNS
7400 modules. Altering it will affect all of them in the same manner.
7402 see also; udp_incoming_address
7404 NOTE, udp_incoming_address and udp_outgoing_address can not
7405 have the same value since they both use the same port.
7412 LOC: Config.onoff.icp_hit_stale
7414 If you want to return ICP_HIT for stale cache objects, set this
7415 option to 'on'. If you have sibling relationships with caches
7416 in other administrative domains, this should be 'off'. If you only
7417 have sibling relationships with caches under your control,
7418 it is probably okay to set this to 'on'.
7419 If set to 'on', your siblings should use the option "allow-miss"
7420 on their cache_peer lines for connecting to you.
7423 NAME: minimum_direct_hops
7426 LOC: Config.minDirectHops
7428 If using the ICMP pinging stuff, do direct fetches for sites
7429 which are no more than this many hops away.
7432 NAME: minimum_direct_rtt
7436 LOC: Config.minDirectRtt
7438 If using the ICMP pinging stuff, do direct fetches for sites
7439 which are no more than this many rtt milliseconds away.
7445 LOC: Config.Netdb.low
7447 The low water mark for the ICMP measurement database.
7449 Note: high watermark controlled by netdb_high directive.
7451 These watermarks are counts, not percents. The defaults are
7452 (low) 900 and (high) 1000. When the high water mark is
7453 reached, database entries will be deleted until the low
7460 LOC: Config.Netdb.high
7462 The high water mark for the ICMP measurement database.
7464 Note: low watermark controlled by netdb_low directive.
7466 These watermarks are counts, not percents. The defaults are
7467 (low) 900 and (high) 1000. When the high water mark is
7468 reached, database entries will be deleted until the low
7472 NAME: netdb_ping_period
7474 LOC: Config.Netdb.period
7477 The minimum period for measuring a site. There will be at
7478 least this much delay between successive pings to the same
7479 network. The default is five minutes.
7486 LOC: Config.onoff.query_icmp
7488 If you want to ask your peers to include ICMP data in their ICP
7489 replies, enable this option.
7491 If your peer has configured Squid (during compilation) with
7492 '--enable-icmp' that peer will send ICMP pings to origin server
7493 sites of the URLs it receives. If you enable this option the
7494 ICP replies from that peer will include the ICMP data (if available).
7495 Then, when choosing a parent cache, Squid will choose the parent with
7496 the minimal RTT to the origin server. When this happens, the
7497 hierarchy field of the access.log will be
7498 "CLOSEST_PARENT_MISS". This option is off by default.
7501 NAME: test_reachability
7505 LOC: Config.onoff.test_reachability
7507 When this is 'on', ICP MISS replies will be ICP_MISS_NOFETCH
7508 instead of ICP_MISS if the target host is NOT in the ICMP
7509 database, or has a zero RTT.
7512 NAME: icp_query_timeout
7515 DEFAULT_DOC: Dynamic detection.
7517 LOC: Config.Timeout.icp_query
7519 Normally Squid will automatically determine an optimal ICP
7520 query timeout value based on the round-trip-time of recent ICP
7521 queries. If you want to override the value determined by
7522 Squid, set this 'icp_query_timeout' to a non-zero value. This
7523 value is specified in MILLISECONDS, so, to use a 2-second
7524 timeout (the old default), you would write:
7526 icp_query_timeout 2000
7529 NAME: maximum_icp_query_timeout
7533 LOC: Config.Timeout.icp_query_max
7535 Normally the ICP query timeout is determined dynamically. But
7536 sometimes it can lead to very large values (say 5 seconds).
7537 Use this option to put an upper limit on the dynamic timeout
7538 value. Do NOT use this option to always use a fixed (instead
7539 of a dynamic) timeout value. To set a fixed timeout see the
7540 'icp_query_timeout' directive.
7543 NAME: minimum_icp_query_timeout
7547 LOC: Config.Timeout.icp_query_min
7549 Normally the ICP query timeout is determined dynamically. But
7550 sometimes it can lead to very small timeouts, even lower than
7551 the normal latency variance on your link due to traffic.
7552 Use this option to put an lower limit on the dynamic timeout
7553 value. Do NOT use this option to always use a fixed (instead
7554 of a dynamic) timeout value. To set a fixed timeout see the
7555 'icp_query_timeout' directive.
7558 NAME: background_ping_rate
7562 LOC: Config.backgroundPingRate
7564 Controls how often the ICP pings are sent to siblings that
7565 have background-ping set.
7569 MULTICAST ICP OPTIONS
7570 -----------------------------------------------------------------------------
7575 LOC: Config.mcast_group_list
7578 This tag specifies a list of multicast groups which your server
7579 should join to receive multicasted ICP queries.
7581 NOTE! Be very careful what you put here! Be sure you
7582 understand the difference between an ICP _query_ and an ICP
7583 _reply_. This option is to be set only if you want to RECEIVE
7584 multicast queries. Do NOT set this option to SEND multicast
7585 ICP (use cache_peer for that). ICP replies are always sent via
7586 unicast, so this option does not affect whether or not you will
7587 receive replies from multicast group members.
7589 You must be very careful to NOT use a multicast address which
7590 is already in use by another group of caches.
7592 If you are unsure about multicast, please read the Multicast
7593 chapter in the Squid FAQ (http://www.squid-cache.org/FAQ/).
7595 Usage: mcast_groups 239.128.16.128 224.0.1.20
7597 By default, Squid doesn't listen on any multicast groups.
7600 NAME: mcast_miss_addr
7601 IFDEF: MULTICAST_MISS_STREAM
7603 LOC: Config.mcast_miss.addr
7605 DEFAULT_DOC: disabled.
7607 If you enable this option, every "cache miss" URL will
7608 be sent out on the specified multicast address.
7610 Do not enable this option unless you are are absolutely
7611 certain you understand what you are doing.
7614 NAME: mcast_miss_ttl
7615 IFDEF: MULTICAST_MISS_STREAM
7617 LOC: Config.mcast_miss.ttl
7620 This is the time-to-live value for packets multicasted
7621 when multicasting off cache miss URLs is enabled. By
7622 default this is set to 'site scope', i.e. 16.
7625 NAME: mcast_miss_port
7626 IFDEF: MULTICAST_MISS_STREAM
7628 LOC: Config.mcast_miss.port
7631 This is the port number to be used in conjunction with
7635 NAME: mcast_miss_encode_key
7636 IFDEF: MULTICAST_MISS_STREAM
7638 LOC: Config.mcast_miss.encode_key
7639 DEFAULT: XXXXXXXXXXXXXXXX
7641 The URLs that are sent in the multicast miss stream are
7642 encrypted. This is the encryption key.
7645 NAME: mcast_icp_query_timeout
7649 LOC: Config.Timeout.mcast_icp_query
7651 For multicast peers, Squid regularly sends out ICP "probes" to
7652 count how many other peers are listening on the given multicast
7653 address. This value specifies how long Squid should wait to
7654 count all the replies. The default is 2000 msec, or 2
7659 INTERNAL ICON OPTIONS
7660 -----------------------------------------------------------------------------
7663 NAME: icon_directory
7665 LOC: Config.icons.directory
7666 DEFAULT: @DEFAULT_ICON_DIR@
7668 Where the icons are stored. These are normally kept in
7672 NAME: global_internal_static
7674 LOC: Config.onoff.global_internal_static
7677 This directive controls is Squid should intercept all requests for
7678 /squid-internal-static/ no matter which host the URL is requesting
7679 (default on setting), or if nothing special should be done for
7680 such URLs (off setting). The purpose of this directive is to make
7681 icons etc work better in complex cache hierarchies where it may
7682 not always be possible for all corners in the cache mesh to reach
7683 the server generating a directory listing.
7686 NAME: short_icon_urls
7688 LOC: Config.icons.use_short_names
7691 If this is enabled Squid will use short URLs for icons.
7692 If disabled it will revert to the old behavior of including
7693 it's own name and port in the URL.
7695 If you run a complex cache hierarchy with a mix of Squid and
7696 other proxies you may need to disable this directive.
7701 -----------------------------------------------------------------------------
7704 NAME: error_directory
7706 LOC: Config.errorDirectory
7708 DEFAULT_DOC: Send error pages in the clients preferred language
7710 If you wish to create your own versions of the default
7711 error files to customize them to suit your company copy
7712 the error/template files to another directory and point
7715 WARNING: This option will disable multi-language support
7716 on error pages if used.
7718 The squid developers are interested in making squid available in
7719 a wide variety of languages. If you are making translations for a
7720 language that Squid does not currently provide please consider
7721 contributing your translation back to the project.
7722 http://wiki.squid-cache.org/Translations
7724 The squid developers working on translations are happy to supply drop-in
7725 translated error files in exchange for any new language contributions.
7728 NAME: error_default_language
7729 IFDEF: USE_ERR_LOCALES
7731 LOC: Config.errorDefaultLanguage
7733 DEFAULT_DOC: Generate English language pages.
7735 Set the default language which squid will send error pages in
7736 if no existing translation matches the clients language
7739 If unset (default) generic English will be used.
7741 The squid developers are interested in making squid available in
7742 a wide variety of languages. If you are interested in making
7743 translations for any language see the squid wiki for details.
7744 http://wiki.squid-cache.org/Translations
7747 NAME: error_log_languages
7748 IFDEF: USE_ERR_LOCALES
7750 LOC: Config.errorLogMissingLanguages
7753 Log to cache.log what languages users are attempting to
7754 auto-negotiate for translations.
7756 Successful negotiations are not logged. Only failures
7757 have meaning to indicate that Squid may need an upgrade
7758 of its error page translations.
7761 NAME: err_page_stylesheet
7763 LOC: Config.errorStylesheet
7764 DEFAULT: @DEFAULT_CONFIG_DIR@/errorpage.css
7766 CSS Stylesheet to pattern the display of Squid default error pages.
7768 For information on CSS see http://www.w3.org/Style/CSS/
7773 LOC: Config.errHtmlText
7776 HTML text to include in error messages. Make this a "mailto"
7777 URL to your admin address, or maybe just a link to your
7778 organizations Web page.
7780 To include this in your error messages, you must rewrite
7781 the error template files (found in the "errors" directory).
7782 Wherever you want the 'err_html_text' line to appear,
7783 insert a %L tag in the error template file.
7786 NAME: email_err_data
7789 LOC: Config.onoff.emailErrData
7792 If enabled, information about the occurred error will be
7793 included in the mailto links of the ERR pages (if %W is set)
7794 so that the email body contains the data.
7795 Syntax is <A HREF="mailto:%w%W">%w</A>
7800 LOC: Config.denyInfoList
7803 Usage: deny_info err_page_name acl
7804 or deny_info http://... acl
7805 or deny_info TCP_RESET acl
7807 This can be used to return a ERR_ page for requests which
7808 do not pass the 'http_access' rules. Squid remembers the last
7809 acl it evaluated in http_access, and if a 'deny_info' line exists
7810 for that ACL Squid returns a corresponding error page.
7812 The acl is typically the last acl on the http_access deny line which
7813 denied access. The exceptions to this rule are:
7814 - When Squid needs to request authentication credentials. It's then
7815 the first authentication related acl encountered
7816 - When none of the http_access lines matches. It's then the last
7817 acl processed on the last http_access line.
7818 - When the decision to deny access was made by an adaptation service,
7819 the acl name is the corresponding eCAP or ICAP service_name.
7821 NP: If providing your own custom error pages with error_directory
7822 you may also specify them by your custom file name:
7823 Example: deny_info ERR_CUSTOM_ACCESS_DENIED bad_guys
7825 By defaut Squid will send "403 Forbidden". A different 4xx or 5xx
7826 may be specified by prefixing the file name with the code and a colon.
7827 e.g. 404:ERR_CUSTOM_ACCESS_DENIED
7829 Alternatively you can tell Squid to reset the TCP connection
7830 by specifying TCP_RESET.
7832 Or you can specify an error URL or URL pattern. The browsers will
7833 get redirected to the specified URL after formatting tags have
7834 been replaced. Redirect will be done with 302 or 307 according to
7835 HTTP/1.1 specs. A different 3xx code may be specified by prefixing
7836 the URL. e.g. 303:http://example.com/
7839 %a - username (if available. Password NOT included)
7842 %E - Error description
7844 %H - Request domain name
7845 %i - Client IP Address
7847 %o - Message result from external ACL helper
7848 %p - Request Port number
7849 %P - Request Protocol name
7850 %R - Request URL path
7851 %T - Timestamp in RFC 1123 format
7852 %U - Full canonical URL from client
7853 (HTTPS URLs terminate with *)
7854 %u - Full canonical URL from client
7855 %w - Admin email from squid.conf
7857 %% - Literal percent (%) code
7862 OPTIONS INFLUENCING REQUEST FORWARDING
7863 -----------------------------------------------------------------------------
7866 NAME: nonhierarchical_direct
7868 LOC: Config.onoff.nonhierarchical_direct
7871 By default, Squid will send any non-hierarchical requests
7872 (not cacheable request type) direct to origin servers.
7874 When this is set to "off", Squid will prefer to send these
7875 requests to parents.
7877 Note that in most configurations, by turning this off you will only
7878 add latency to these request without any improvement in global hit
7881 This option only sets a preference. If the parent is unavailable a
7882 direct connection to the origin server may still be attempted. To
7883 completely prevent direct connections use never_direct.
7888 LOC: Config.onoff.prefer_direct
7891 Normally Squid tries to use parents for most requests. If you for some
7892 reason like it to first try going direct and only use a parent if
7893 going direct fails set this to on.
7895 By combining nonhierarchical_direct off and prefer_direct on you
7896 can set up Squid to use a parent as a backup path if going direct
7899 Note: If you want Squid to use parents for all requests see
7900 the never_direct directive. prefer_direct only modifies how Squid
7901 acts on cacheable requests.
7904 NAME: cache_miss_revalidate
7908 LOC: Config.onoff.cache_miss_revalidate
7910 RFC 7232 defines a conditional request mechanism to prevent
7911 response objects being unnecessarily transferred over the network.
7912 If that mechanism is used by the client and a cache MISS occurs
7913 it can prevent new cache entries being created.
7915 This option determines whether Squid on cache MISS will pass the
7916 client revalidation request to the server or tries to fetch new
7917 content for caching. It can be useful while the cache is mostly
7918 empty to more quickly have the cache populated by generating
7919 non-conditional GETs.
7921 When set to 'on' (default), Squid will pass all client If-* headers
7922 to the server. This permits server responses without a cacheable
7923 payload to be delivered and on MISS no new cache entry is created.
7925 When set to 'off' and if the request is cacheable, Squid will
7926 remove the clients If-Modified-Since and If-None-Match headers from
7927 the request sent to the server. This requests a 200 status response
7928 from the server to create a new cache entry with.
7933 LOC: Config.accessList.AlwaysDirect
7935 DEFAULT_DOC: Prevent any cache_peer being used for this request.
7937 Usage: always_direct allow|deny [!]aclname ...
7939 Here you can use ACL elements to specify requests which should
7940 ALWAYS be forwarded by Squid to the origin servers without using
7941 any peers. For example, to always directly forward requests for
7942 local servers ignoring any parents or siblings you may have use
7945 acl local-servers dstdomain my.domain.net
7946 always_direct allow local-servers
7948 To always forward FTP requests directly, use
7951 always_direct allow FTP
7953 NOTE: There is a similar, but opposite option named
7954 'never_direct'. You need to be aware that "always_direct deny
7955 foo" is NOT the same thing as "never_direct allow foo". You
7956 may need to use a deny rule to exclude a more-specific case of
7957 some other rule. Example:
7959 acl local-external dstdomain external.foo.net
7960 acl local-servers dstdomain .foo.net
7961 always_direct deny local-external
7962 always_direct allow local-servers
7964 NOTE: If your goal is to make the client forward the request
7965 directly to the origin server bypassing Squid then this needs
7966 to be done in the client configuration. Squid configuration
7967 can only tell Squid how Squid should fetch the object.
7969 NOTE: This directive is not related to caching. The replies
7970 is cached as usual even if you use always_direct. To not cache
7971 the replies see the 'cache' directive.
7973 This clause supports both fast and slow acl types.
7974 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
7979 LOC: Config.accessList.NeverDirect
7981 DEFAULT_DOC: Allow DNS results to be used for this request.
7983 Usage: never_direct allow|deny [!]aclname ...
7985 never_direct is the opposite of always_direct. Please read
7986 the description for always_direct if you have not already.
7988 With 'never_direct' you can use ACL elements to specify
7989 requests which should NEVER be forwarded directly to origin
7990 servers. For example, to force the use of a proxy for all
7991 requests, except those in your local domain use something like:
7993 acl local-servers dstdomain .foo.net
7994 never_direct deny local-servers
7995 never_direct allow all
7997 or if Squid is inside a firewall and there are local intranet
7998 servers inside the firewall use something like:
8000 acl local-intranet dstdomain .foo.net
8001 acl local-external dstdomain external.foo.net
8002 always_direct deny local-external
8003 always_direct allow local-intranet
8004 never_direct allow all
8006 This clause supports both fast and slow acl types.
8007 See http://wiki.squid-cache.org/SquidFaq/SquidAcl for details.
8011 ADVANCED NETWORKING OPTIONS
8012 -----------------------------------------------------------------------------
8015 NAME: incoming_udp_average incoming_icp_average
8018 LOC: Config.comm_incoming.udp.average
8020 Heavy voodoo here. I can't even believe you are reading this.
8021 Are you crazy? Don't even think about adjusting these unless
8022 you understand the algorithms in comm_select.c first!
8025 NAME: incoming_tcp_average incoming_http_average
8028 LOC: Config.comm_incoming.tcp.average
8030 Heavy voodoo here. I can't even believe you are reading this.
8031 Are you crazy? Don't even think about adjusting these unless
8032 you understand the algorithms in comm_select.c first!
8035 NAME: incoming_dns_average
8038 LOC: Config.comm_incoming.dns.average
8040 Heavy voodoo here. I can't even believe you are reading this.
8041 Are you crazy? Don't even think about adjusting these unless
8042 you understand the algorithms in comm_select.c first!
8045 NAME: min_udp_poll_cnt min_icp_poll_cnt
8048 LOC: Config.comm_incoming.udp.min_poll
8050 Heavy voodoo here. I can't even believe you are reading this.
8051 Are you crazy? Don't even think about adjusting these unless
8052 you understand the algorithms in comm_select.c first!
8055 NAME: min_dns_poll_cnt
8058 LOC: Config.comm_incoming.dns.min_poll
8060 Heavy voodoo here. I can't even believe you are reading this.
8061 Are you crazy? Don't even think about adjusting these unless
8062 you understand the algorithms in comm_select.c first!
8065 NAME: min_tcp_poll_cnt min_http_poll_cnt
8068 LOC: Config.comm_incoming.tcp.min_poll
8070 Heavy voodoo here. I can't even believe you are reading this.
8071 Are you crazy? Don't even think about adjusting these unless
8072 you understand the algorithms in comm_select.c first!
8078 LOC: Config.accept_filter
8082 The name of an accept(2) filter to install on Squid's
8083 listen socket(s). This feature is perhaps specific to
8084 FreeBSD and requires support in the kernel.
8086 The 'httpready' filter delays delivering new connections
8087 to Squid until a full HTTP request has been received.
8088 See the accf_http(9) man page for details.
8090 The 'dataready' filter delays delivering new connections
8091 to Squid until there is some data to process.
8092 See the accf_dataready(9) man page for details.
8096 The 'data' filter delays delivering of new connections
8097 to Squid until there is some data to process by TCP_ACCEPT_DEFER.
8098 You may optionally specify a number of seconds to wait by
8099 'data=N' where N is the number of seconds. Defaults to 30
8100 if not specified. See the tcp(7) man page for details.
8103 accept_filter httpready
8108 NAME: client_ip_max_connections
8110 LOC: Config.client_ip_max_connections
8112 DEFAULT_DOC: No limit.
8114 Set an absolute limit on the number of connections a single
8115 client IP can use. Any more than this and Squid will begin to drop
8116 new connections from the client until it closes some links.
8118 Note that this is a global limit. It affects all HTTP, HTCP, Gopher and FTP
8119 connections from the client. For finer control use the ACL access controls.
8121 Requires client_db to be enabled (the default).
8123 WARNING: This may noticably slow down traffic received via external proxies
8124 or NAT devices and cause them to rebound error messages back to their clients.
8127 NAME: tcp_recv_bufsize
8131 DEFAULT_DOC: Use operating system TCP defaults.
8132 LOC: Config.tcpRcvBufsz
8134 Size of receive buffer to set for TCP sockets. Probably just
8135 as easy to change your kernel's default.
8136 Omit from squid.conf to use the default buffer size.
8141 -----------------------------------------------------------------------------
8148 LOC: Adaptation::Icap::TheConfig.onoff
8151 If you want to enable the ICAP module support, set this to on.
8154 NAME: icap_connect_timeout
8157 LOC: Adaptation::Icap::TheConfig.connect_timeout_raw
8160 This parameter specifies how long to wait for the TCP connect to
8161 the requested ICAP server to complete before giving up and either
8162 terminating the HTTP transaction or bypassing the failure.
8164 The default for optional services is peer_connect_timeout.
8165 The default for essential services is connect_timeout.
8166 If this option is explicitly set, its value applies to all services.
8169 NAME: icap_io_timeout
8173 DEFAULT_DOC: Use read_timeout.
8174 LOC: Adaptation::Icap::TheConfig.io_timeout_raw
8177 This parameter specifies how long to wait for an I/O activity on
8178 an established, active ICAP connection before giving up and
8179 either terminating the HTTP transaction or bypassing the
8183 NAME: icap_service_failure_limit
8184 COMMENT: limit [in memory-depth time-units]
8185 TYPE: icap_service_failure_limit
8187 LOC: Adaptation::Icap::TheConfig
8190 The limit specifies the number of failures that Squid tolerates
8191 when establishing a new TCP connection with an ICAP service. If
8192 the number of failures exceeds the limit, the ICAP service is
8193 not used for new ICAP requests until it is time to refresh its
8196 A negative value disables the limit. Without the limit, an ICAP
8197 service will not be considered down due to connectivity failures
8198 between ICAP OPTIONS requests.
8200 Squid forgets ICAP service failures older than the specified
8201 value of memory-depth. The memory fading algorithm
8202 is approximate because Squid does not remember individual
8203 errors but groups them instead, splitting the option
8204 value into ten time slots of equal length.
8206 When memory-depth is 0 and by default this option has no
8207 effect on service failure expiration.
8209 Squid always forgets failures when updating service settings
8210 using an ICAP OPTIONS transaction, regardless of this option
8214 # suspend service usage after 10 failures in 5 seconds:
8215 icap_service_failure_limit 10 in 5 seconds
8218 NAME: icap_service_revival_delay
8221 LOC: Adaptation::Icap::TheConfig.service_revival_delay
8224 The delay specifies the number of seconds to wait after an ICAP
8225 OPTIONS request failure before requesting the options again. The
8226 failed ICAP service is considered "down" until fresh OPTIONS are
8229 The actual delay cannot be smaller than the hardcoded minimum
8230 delay of 30 seconds.
8233 NAME: icap_preview_enable
8237 LOC: Adaptation::Icap::TheConfig.preview_enable
8240 The ICAP Preview feature allows the ICAP server to handle the
8241 HTTP message by looking only at the beginning of the message body
8242 or even without receiving the body at all. In some environments,
8243 previews greatly speedup ICAP processing.
8245 During an ICAP OPTIONS transaction, the server may tell Squid what
8246 HTTP messages should be previewed and how big the preview should be.
8247 Squid will not use Preview if the server did not request one.
8249 To disable ICAP Preview for all ICAP services, regardless of
8250 individual ICAP server OPTIONS responses, set this option to "off".
8252 icap_preview_enable off
8255 NAME: icap_preview_size
8258 LOC: Adaptation::Icap::TheConfig.preview_size
8260 DEFAULT_DOC: No preview sent.
8262 The default size of preview data to be sent to the ICAP server.
8263 This value might be overwritten on a per server basis by OPTIONS requests.
8266 NAME: icap_206_enable
8270 LOC: Adaptation::Icap::TheConfig.allow206_enable
8273 206 (Partial Content) responses is an ICAP extension that allows the
8274 ICAP agents to optionally combine adapted and original HTTP message
8275 content. The decision to combine is postponed until the end of the
8276 ICAP response. Squid supports Partial Content extension by default.
8278 Activation of the Partial Content extension is negotiated with each
8279 ICAP service during OPTIONS exchange. Most ICAP servers should handle
8280 negotation correctly even if they do not support the extension, but
8281 some might fail. To disable Partial Content support for all ICAP
8282 services and to avoid any negotiation, set this option to "off".
8288 NAME: icap_default_options_ttl
8291 LOC: Adaptation::Icap::TheConfig.default_options_ttl
8294 The default TTL value for ICAP OPTIONS responses that don't have
8295 an Options-TTL header.
8298 NAME: icap_persistent_connections
8302 LOC: Adaptation::Icap::TheConfig.reuse_connections
8305 Whether or not Squid should use persistent connections to
8309 NAME: adaptation_send_client_ip icap_send_client_ip
8311 IFDEF: USE_ADAPTATION
8313 LOC: Adaptation::Config::send_client_ip
8316 If enabled, Squid shares HTTP client IP information with adaptation
8317 services. For ICAP, Squid adds the X-Client-IP header to ICAP requests.
8318 For eCAP, Squid sets the libecap::metaClientIp transaction option.
8320 See also: adaptation_uses_indirect_client
8323 NAME: adaptation_send_username icap_send_client_username
8325 IFDEF: USE_ADAPTATION
8327 LOC: Adaptation::Config::send_username
8330 This sends authenticated HTTP client username (if available) to
8331 the adaptation service.
8333 For ICAP, the username value is encoded based on the
8334 icap_client_username_encode option and is sent using the header
8335 specified by the icap_client_username_header option.
8338 NAME: icap_client_username_header
8341 LOC: Adaptation::Icap::TheConfig.client_username_header
8342 DEFAULT: X-Client-Username
8344 ICAP request header name to use for adaptation_send_username.
8347 NAME: icap_client_username_encode
8351 LOC: Adaptation::Icap::TheConfig.client_username_encode
8354 Whether to base64 encode the authenticated client username.
8358 TYPE: icap_service_type
8360 LOC: Adaptation::Icap::TheConfig
8363 Defines a single ICAP service using the following format:
8365 icap_service id vectoring_point uri [option ...]
8368 an opaque identifier or name which is used to direct traffic to
8369 this specific service. Must be unique among all adaptation
8370 services in squid.conf.
8372 vectoring_point: reqmod_precache|reqmod_postcache|respmod_precache|respmod_postcache
8373 This specifies at which point of transaction processing the
8374 ICAP service should be activated. *_postcache vectoring points
8375 are not yet supported.
8377 uri: icap://servername:port/servicepath
8378 ICAP server and service location.
8379 icaps://servername:port/servicepath
8380 The "icap:" URI scheme is used for traditional ICAP server and
8381 service location (default port is 1344, connections are not
8382 encrypted). The "icaps:" URI scheme is for Secure ICAP
8383 services that use SSL/TLS-encrypted ICAP connections (by
8384 default, on port 11344).
8386 ICAP does not allow a single service to handle both REQMOD and RESPMOD
8387 transactions. Squid does not enforce that requirement. You can specify
8388 services with the same service_url and different vectoring_points. You
8389 can even specify multiple identical services as long as their
8390 service_names differ.
8392 To activate a service, use the adaptation_access directive. To group
8393 services, use adaptation_service_chain and adaptation_service_set.
8395 Service options are separated by white space. ICAP services support
8396 the following name=value options:
8399 If set to 'on' or '1', the ICAP service is treated as
8400 optional. If the service cannot be reached or malfunctions,
8401 Squid will try to ignore any errors and process the message as
8402 if the service was not enabled. No all ICAP errors can be
8403 bypassed. If set to 0, the ICAP service is treated as
8404 essential and all ICAP errors will result in an error page
8405 returned to the HTTP client.
8407 Bypass is off by default: services are treated as essential.
8410 If set to 'on' or '1', the ICAP service is allowed to
8411 dynamically change the current message adaptation plan by
8412 returning a chain of services to be used next. The services
8413 are specified using the X-Next-Services ICAP response header
8414 value, formatted as a comma-separated list of service names.
8415 Each named service should be configured in squid.conf. Other
8416 services are ignored. An empty X-Next-Services value results
8417 in an empty plan which ends the current adaptation.
8419 Dynamic adaptation plan may cross or cover multiple supported
8420 vectoring points in their natural processing order.
8422 Routing is not allowed by default: the ICAP X-Next-Services
8423 response header is ignored.
8426 Only has effect on split-stack systems. The default on those systems
8427 is to use IPv4-only connections. When set to 'on' this option will
8428 make Squid use IPv6-only connections to contact this ICAP service.
8430 on-overload=block|bypass|wait|force
8431 If the service Max-Connections limit has been reached, do
8432 one of the following for each new ICAP transaction:
8433 * block: send an HTTP error response to the client
8434 * bypass: ignore the "over-connected" ICAP service
8435 * wait: wait (in a FIFO queue) for an ICAP connection slot
8436 * force: proceed, ignoring the Max-Connections limit
8438 In SMP mode with N workers, each worker assumes the service
8439 connection limit is Max-Connections/N, even though not all
8440 workers may use a given service.
8442 The default value is "bypass" if service is bypassable,
8443 otherwise it is set to "wait".
8447 Use the given number as the Max-Connections limit, regardless
8448 of the Max-Connections value given by the service, if any.
8450 ==== ICAPS / TLS OPTIONS ====
8452 These options are used for Secure ICAP (icaps://....) services only.
8454 tls-cert=/path/to/ssl/certificate
8455 A client SSL certificate to use when connecting to
8458 tls-key=/path/to/ssl/key
8459 The private TLS/SSL key corresponding to sslcert above.
8460 If 'tls-key' is not specified 'tls-cert' is assumed to
8461 reference a combined PEM format file containing both the
8462 certificate and the key.
8464 tls-cipher=... The list of valid TLS/SSL ciphers to use when connecting
8465 to this icap server.
8468 The minimum TLS protocol version to permit. To control
8469 SSLv3 use the ssloptions= parameter.
8470 Supported Values: 1.0 (default), 1.1, 1.2
8472 tls-options=... Specify various OpenSSL library options:
8474 NO_SSLv3 Disallow the use of SSLv3
8476 NO_TLSv1 Disallow the use of TLSv1.0
8477 NO_TLSv1_1 Disallow the use of TLSv1.1
8478 NO_TLSv1_2 Disallow the use of TLSv1.2
8481 Always create a new key when using
8482 temporary/ephemeral DH key exchanges
8484 ALL Enable various bug workarounds
8485 suggested as "harmless" by OpenSSL
8486 Be warned that this reduces SSL/TLS
8487 strength to some attacks.
8489 See the OpenSSL SSL_CTX_set_options documentation for a
8490 more complete list. Options relevant only to SSLv2 are
8493 tls-cafile= PEM file containing CA certificates to use when verifying
8494 the icap server certificate.
8495 Use to specify intermediate CA certificate(s) if not sent
8496 by the server. Or the full CA chain for the server when
8497 using the tls-no-default-ca flag.
8498 May be repeated to load multiple files.
8500 tls-capath=... A directory containing additional CA certificates to
8501 use when verifying the icap server certificate.
8502 Requires OpenSSL or LibreSSL.
8504 tls-crlfile=... A certificate revocation list file to use when
8505 verifying the icap server certificate.
8507 tls-flags=... Specify various flags modifying the Squid TLS implementation:
8510 Accept certificates even if they fail to
8513 Don't verify the icap server certificate
8514 matches the server name
8517 Do no use the system default Trusted CA.
8519 tls-domain= The icap server name as advertised in it's certificate.
8520 Used for verifying the correctness of the received icap
8521 server certificate. If not specified the icap server
8522 hostname extracted from ICAP URI will be used.
8524 Older icap_service format without optional named parameters is
8525 deprecated but supported for backward compatibility.
8528 icap_service svcBlocker reqmod_precache icap://icap1.mydomain.net:1344/reqmod bypass=0
8529 icap_service svcLogger reqmod_precache icaps://icap2.mydomain.net:11344/reqmod routing=on
8533 TYPE: icap_class_type
8538 This deprecated option was documented to define an ICAP service
8539 chain, even though it actually defined a set of similar, redundant
8540 services, and the chains were not supported.
8542 To define a set of redundant services, please use the
8543 adaptation_service_set directive. For service chains, use
8544 adaptation_service_chain.
8548 TYPE: icap_access_type
8553 This option is deprecated. Please use adaptation_access, which
8554 has the same ICAP functionality, but comes with better
8555 documentation, and eCAP support.
8560 -----------------------------------------------------------------------------
8567 LOC: Adaptation::Ecap::TheConfig.onoff
8570 Controls whether eCAP support is enabled.
8574 TYPE: ecap_service_type
8576 LOC: Adaptation::Ecap::TheConfig
8579 Defines a single eCAP service
8581 ecap_service id vectoring_point uri [option ...]
8584 an opaque identifier or name which is used to direct traffic to
8585 this specific service. Must be unique among all adaptation
8586 services in squid.conf.
8588 vectoring_point: reqmod_precache|reqmod_postcache|respmod_precache|respmod_postcache
8589 This specifies at which point of transaction processing the
8590 eCAP service should be activated. *_postcache vectoring points
8591 are not yet supported.
8593 uri: ecap://vendor/service_name?custom&cgi=style¶meters=optional
8594 Squid uses the eCAP service URI to match this configuration
8595 line with one of the dynamically loaded services. Each loaded
8596 eCAP service must have a unique URI. Obtain the right URI from
8597 the service provider.
8599 To activate a service, use the adaptation_access directive. To group
8600 services, use adaptation_service_chain and adaptation_service_set.
8602 Service options are separated by white space. eCAP services support
8603 the following name=value options:
8606 If set to 'on' or '1', the eCAP service is treated as optional.
8607 If the service cannot be reached or malfunctions, Squid will try
8608 to ignore any errors and process the message as if the service
8609 was not enabled. No all eCAP errors can be bypassed.
8610 If set to 'off' or '0', the eCAP service is treated as essential
8611 and all eCAP errors will result in an error page returned to the
8614 Bypass is off by default: services are treated as essential.
8617 If set to 'on' or '1', the eCAP service is allowed to
8618 dynamically change the current message adaptation plan by
8619 returning a chain of services to be used next.
8621 Dynamic adaptation plan may cross or cover multiple supported
8622 vectoring points in their natural processing order.
8624 Routing is not allowed by default.
8626 Older ecap_service format without optional named parameters is
8627 deprecated but supported for backward compatibility.
8631 ecap_service s1 reqmod_precache ecap://filters.R.us/leakDetector?on_error=block bypass=off
8632 ecap_service s2 respmod_precache ecap://filters.R.us/virusFilter config=/etc/vf.cfg bypass=on
8635 NAME: loadable_modules
8637 IFDEF: USE_LOADABLE_MODULES
8638 LOC: Config.loadable_module_names
8641 Instructs Squid to load the specified dynamic module(s) or activate
8642 preloaded module(s).
8644 loadable_modules @DEFAULT_PREFIX@/lib/MinimalAdapter.so
8648 MESSAGE ADAPTATION OPTIONS
8649 -----------------------------------------------------------------------------
8652 NAME: adaptation_service_set
8653 TYPE: adaptation_service_set_type
8654 IFDEF: USE_ADAPTATION
8659 Configures an ordered set of similar, redundant services. This is
8660 useful when hot standby or backup adaptation servers are available.
8662 adaptation_service_set set_name service_name1 service_name2 ...
8664 The named services are used in the set declaration order. The first
8665 applicable adaptation service from the set is used first. The next
8666 applicable service is tried if and only if the transaction with the
8667 previous service fails and the message waiting to be adapted is still
8670 When adaptation starts, broken services are ignored as if they were
8671 not a part of the set. A broken service is a down optional service.
8673 The services in a set must be attached to the same vectoring point
8674 (e.g., pre-cache) and use the same adaptation method (e.g., REQMOD).
8676 If all services in a set are optional then adaptation failures are
8677 bypassable. If all services in the set are essential, then a
8678 transaction failure with one service may still be retried using
8679 another service from the set, but when all services fail, the master
8680 transaction fails as well.
8682 A set may contain a mix of optional and essential services, but that
8683 is likely to lead to surprising results because broken services become
8684 ignored (see above), making previously bypassable failures fatal.
8685 Technically, it is the bypassability of the last failed service that
8688 See also: adaptation_access adaptation_service_chain
8691 adaptation_service_set svcBlocker urlFilterPrimary urlFilterBackup
8692 adaptation service_set svcLogger loggerLocal loggerRemote
8695 NAME: adaptation_service_chain
8696 TYPE: adaptation_service_chain_type
8697 IFDEF: USE_ADAPTATION
8702 Configures a list of complementary services that will be applied
8703 one-by-one, forming an adaptation chain or pipeline. This is useful
8704 when Squid must perform different adaptations on the same message.
8706 adaptation_service_chain chain_name service_name1 svc_name2 ...
8708 The named services are used in the chain declaration order. The first
8709 applicable adaptation service from the chain is used first. The next
8710 applicable service is applied to the successful adaptation results of
8711 the previous service in the chain.
8713 When adaptation starts, broken services are ignored as if they were
8714 not a part of the chain. A broken service is a down optional service.
8716 Request satisfaction terminates the adaptation chain because Squid
8717 does not currently allow declaration of RESPMOD services at the
8718 "reqmod_precache" vectoring point (see icap_service or ecap_service).
8720 The services in a chain must be attached to the same vectoring point
8721 (e.g., pre-cache) and use the same adaptation method (e.g., REQMOD).
8723 A chain may contain a mix of optional and essential services. If an
8724 essential adaptation fails (or the failure cannot be bypassed for
8725 other reasons), the master transaction fails. Otherwise, the failure
8726 is bypassed as if the failed adaptation service was not in the chain.
8728 See also: adaptation_access adaptation_service_set
8731 adaptation_service_chain svcRequest requestLogger urlFilter leakDetector
8734 NAME: adaptation_access
8735 TYPE: adaptation_access_type
8736 IFDEF: USE_ADAPTATION
8739 DEFAULT_DOC: Allow, unless rules exist in squid.conf.
8741 Sends an HTTP transaction to an ICAP or eCAP adaptation service.
8743 adaptation_access service_name allow|deny [!]aclname...
8744 adaptation_access set_name allow|deny [!]aclname...
8746 At each supported vectoring point, the adaptation_access
8747 statements are processed in the order they appear in this
8748 configuration file. Statements pointing to the following services
8749 are ignored (i.e., skipped without checking their ACL):
8751 - services serving different vectoring points
8752 - "broken-but-bypassable" services
8753 - "up" services configured to ignore such transactions
8754 (e.g., based on the ICAP Transfer-Ignore header).
8756 When a set_name is used, all services in the set are checked
8757 using the same rules, to find the first applicable one. See
8758 adaptation_service_set for details.
8760 If an access list is checked and there is a match, the
8761 processing stops: For an "allow" rule, the corresponding
8762 adaptation service is used for the transaction. For a "deny"
8763 rule, no adaptation service is activated.
8765 It is currently not possible to apply more than one adaptation
8766 service at the same vectoring point to the same HTTP transaction.
8768 See also: icap_service and ecap_service
8771 adaptation_access service_1 allow all
8774 NAME: adaptation_service_iteration_limit
8776 IFDEF: USE_ADAPTATION
8777 LOC: Adaptation::Config::service_iteration_limit
8780 Limits the number of iterations allowed when applying adaptation
8781 services to a message. If your longest adaptation set or chain
8782 may have more than 16 services, increase the limit beyond its
8783 default value of 16. If detecting infinite iteration loops sooner
8784 is critical, make the iteration limit match the actual number
8785 of services in your longest adaptation set or chain.
8787 Infinite adaptation loops are most likely with routing services.
8789 See also: icap_service routing=1
8792 NAME: adaptation_masterx_shared_names
8794 IFDEF: USE_ADAPTATION
8795 LOC: Adaptation::Config::masterx_shared_name
8798 For each master transaction (i.e., the HTTP request and response
8799 sequence, including all related ICAP and eCAP exchanges), Squid
8800 maintains a table of metadata. The table entries are (name, value)
8801 pairs shared among eCAP and ICAP exchanges. The table is destroyed
8802 with the master transaction.
8804 This option specifies the table entry names that Squid must accept
8805 from and forward to the adaptation transactions.
8807 An ICAP REQMOD or RESPMOD transaction may set an entry in the
8808 shared table by returning an ICAP header field with a name
8809 specified in adaptation_masterx_shared_names.
8811 An eCAP REQMOD or RESPMOD transaction may set an entry in the
8812 shared table by implementing the libecap::visitEachOption() API
8813 to provide an option with a name specified in
8814 adaptation_masterx_shared_names.
8816 Squid will store and forward the set entry to subsequent adaptation
8817 transactions within the same master transaction scope.
8819 Only one shared entry name is supported at this time.
8822 # share authentication information among ICAP services
8823 adaptation_masterx_shared_names X-Subscriber-ID
8826 NAME: adaptation_meta
8828 IFDEF: USE_ADAPTATION
8829 LOC: Adaptation::Config::metaHeaders
8832 This option allows Squid administrator to add custom ICAP request
8833 headers or eCAP options to Squid ICAP requests or eCAP transactions.
8834 Use it to pass custom authentication tokens and other
8835 transaction-state related meta information to an ICAP/eCAP service.
8837 The addition of a meta header is ACL-driven:
8838 adaptation_meta name value [!]aclname ...
8840 Processing for a given header name stops after the first ACL list match.
8841 Thus, it is impossible to add two headers with the same name. If no ACL
8842 lists match for a given header name, no such header is added. For
8845 # do not debug transactions except for those that need debugging
8846 adaptation_meta X-Debug 1 needs_debugging
8848 # log all transactions except for those that must remain secret
8849 adaptation_meta X-Log 1 !keep_secret
8851 # mark transactions from users in the "G 1" group
8852 adaptation_meta X-Authenticated-Groups "G 1" authed_as_G1
8854 The "value" parameter may be a regular squid.conf token or a "double
8855 quoted string". Within the quoted string, use backslash (\) to escape
8856 any character, which is currently only useful for escaping backslashes
8857 and double quotes. For example,
8858 "this string has one backslash (\\) and two \"quotes\""
8860 Used adaptation_meta header values may be logged via %note
8861 logformat code. If multiple adaptation_meta headers with the same name
8862 are used during master transaction lifetime, the header values are
8863 logged in the order they were used and duplicate values are ignored
8864 (only the first repeated value will be logged).
8870 LOC: Adaptation::Icap::TheConfig.repeat
8871 DEFAULT_IF_NONE: deny all
8873 This ACL determines which retriable ICAP transactions are
8874 retried. Transactions that received a complete ICAP response
8875 and did not have to consume or produce HTTP bodies to receive
8876 that response are usually retriable.
8878 icap_retry allow|deny [!]aclname ...
8880 Squid automatically retries some ICAP I/O timeouts and errors
8881 due to persistent connection race conditions.
8883 See also: icap_retry_limit
8886 NAME: icap_retry_limit
8889 LOC: Adaptation::Icap::TheConfig.repeat_limit
8891 DEFAULT_DOC: No retries are allowed.
8893 Limits the number of retries allowed.
8895 Communication errors due to persistent connection race
8896 conditions are unavoidable, automatically retried, and do not
8897 count against this limit.
8899 See also: icap_retry
8905 -----------------------------------------------------------------------------
8908 NAME: check_hostnames
8911 LOC: Config.onoff.check_hostnames
8913 For security and stability reasons Squid can check
8914 hostnames for Internet standard RFC compliance. If you want
8915 Squid to perform these checks turn this directive on.
8918 NAME: allow_underscore
8921 LOC: Config.onoff.allow_underscore
8923 Underscore characters is not strictly allowed in Internet hostnames
8924 but nevertheless used by many sites. Set this to off if you want
8925 Squid to be strict about the standard.
8926 This check is performed only when check_hostnames is set to on.
8929 NAME: dns_retransmit_interval
8932 LOC: Config.Timeout.idns_retransmit
8934 Initial retransmit interval for DNS queries. The interval is
8935 doubled each time all configured DNS servers have been tried.
8941 LOC: Config.Timeout.idns_query
8943 DNS Query timeout. If no response is received to a DNS query
8944 within this time all DNS servers for the queried domain
8945 are assumed to be unavailable.
8948 NAME: dns_packet_max
8950 DEFAULT_DOC: EDNS disabled
8952 LOC: Config.dns.packet_max
8954 Maximum number of bytes packet size to advertise via EDNS.
8955 Set to "none" to disable EDNS large packet support.
8957 For legacy reasons DNS UDP replies will default to 512 bytes which
8958 is too small for many responses. EDNS provides a means for Squid to
8959 negotiate receiving larger responses back immediately without having
8960 to failover with repeat requests. Responses larger than this limit
8961 will retain the old behaviour of failover to TCP DNS.
8963 Squid has no real fixed limit internally, but allowing packet sizes
8964 over 1500 bytes requires network jumbogram support and is usually not
8967 WARNING: The RFC also indicates that some older resolvers will reply
8968 with failure of the whole request if the extension is added. Some
8969 resolvers have already been identified which will reply with mangled
8970 EDNS response on occasion. Usually in response to many-KB jumbogram
8971 sizes being advertised by Squid.
8972 Squid will currently treat these both as an unable-to-resolve domain
8973 even if it would be resolvable without EDNS.
8980 DEFAULT_DOC: Search for single-label domain names is disabled.
8981 LOC: Config.onoff.res_defnames
8983 Normally the RES_DEFNAMES resolver option is disabled
8984 (see res_init(3)). This prevents caches in a hierarchy
8985 from interpreting single-component hostnames locally. To allow
8986 Squid to handle single-component names, enable this option.
8989 NAME: dns_multicast_local
8993 DEFAULT_DOC: Search for .local and .arpa names is disabled.
8994 LOC: Config.onoff.dns_mdns
8996 When set to on, Squid sends multicast DNS lookups on the local
8997 network for domains ending in .local and .arpa.
8998 This enables local servers and devices to be contacted in an
8999 ad-hoc or zero-configuration network environment.
9002 NAME: dns_nameservers
9005 DEFAULT_DOC: Use operating system definitions
9006 LOC: Config.dns_nameservers
9008 Use this if you want to specify a list of DNS name servers
9009 (IP addresses) to use instead of those given in your
9010 /etc/resolv.conf file.
9012 On Windows platforms, if no value is specified here or in
9013 the /etc/resolv.conf file, the list of DNS name servers are
9014 taken from the Windows registry, both static and dynamic DHCP
9015 configurations are supported.
9017 Example: dns_nameservers 10.0.0.1 192.172.0.4
9022 DEFAULT: @DEFAULT_HOSTS@
9023 LOC: Config.etcHostsPath
9025 Location of the host-local IP name-address associations
9026 database. Most Operating Systems have such a file on different
9028 - Un*X & Linux: /etc/hosts
9029 - Windows NT/2000: %SystemRoot%\system32\drivers\etc\hosts
9030 (%SystemRoot% value install default is c:\winnt)
9031 - Windows XP/2003: %SystemRoot%\system32\drivers\etc\hosts
9032 (%SystemRoot% value install default is c:\windows)
9033 - Windows 9x/Me: %windir%\hosts
9034 (%windir% value is usually c:\windows)
9035 - Cygwin: /etc/hosts
9037 The file contains newline-separated definitions, in the
9038 form ip_address_in_dotted_form name [name ...] names are
9039 whitespace-separated. Lines beginning with an hash (#)
9040 character are comments.
9042 The file is checked at startup and upon configuration.
9043 If set to 'none', it won't be checked.
9044 If append_domain is used, that domain will be added to
9045 domain-local (i.e. not containing any dot character) host
9051 LOC: Config.appendDomain
9053 DEFAULT_DOC: Use operating system definitions
9055 Appends local domain name to hostnames without any dots in
9056 them. append_domain must begin with a period.
9058 Be warned there are now Internet names with no dots in
9059 them using only top-domain names, so setting this may
9060 cause some Internet sites to become unavailable.
9063 append_domain .yourdomain.com
9066 NAME: ignore_unknown_nameservers
9068 LOC: Config.onoff.ignore_unknown_nameservers
9071 By default Squid checks that DNS responses are received
9072 from the same IP addresses they are sent to. If they
9073 don't match, Squid ignores the response and writes a warning
9074 message to cache.log. You can allow responses from unknown
9075 nameservers by setting this option to 'off'.
9081 LOC: Config.dns.v4_first
9083 With the IPv6 Internet being as fast or faster than IPv4 Internet
9084 for most networks Squid prefers to contact websites over IPv6.
9086 This option reverses the order of preference to make Squid contact
9087 dual-stack websites over IPv4 first. Squid will still perform both
9088 IPv6 and IPv4 DNS lookups before connecting.
9091 This option will restrict the situations under which IPv6
9092 connectivity is used (and tested). Hiding network problems
9093 which would otherwise be detected and warned about.
9097 COMMENT: (number of entries)
9100 LOC: Config.ipcache.size
9102 Maximum number of DNS IP cache entries.
9109 LOC: Config.ipcache.low
9116 LOC: Config.ipcache.high
9118 The size, low-, and high-water marks for the IP cache.
9121 NAME: fqdncache_size
9122 COMMENT: (number of entries)
9125 LOC: Config.fqdncache.size
9127 Maximum number of FQDN cache entries.
9132 -----------------------------------------------------------------------------
9135 NAME: configuration_includes_quoted_values
9137 TYPE: configuration_includes_quoted_values
9139 LOC: ConfigParser::RecognizeQuotedValues
9141 If set, Squid will recognize each "quoted string" after a configuration
9142 directive as a single parameter. The quotes are stripped before the
9143 parameter value is interpreted or used.
9144 See "Values with spaces, quotes, and other special characters"
9145 section for more details.
9152 LOC: Config.onoff.mem_pools
9154 If set, Squid will keep pools of allocated (but unused) memory
9155 available for future use. If memory is a premium on your
9156 system and you believe your malloc library outperforms Squid
9157 routines, disable this.
9160 NAME: memory_pools_limit
9164 LOC: Config.MemPools.limit
9166 Used only with memory_pools on:
9167 memory_pools_limit 50 MB
9169 If set to a non-zero value, Squid will keep at most the specified
9170 limit of allocated (but unused) memory in memory pools. All free()
9171 requests that exceed this limit will be handled by your malloc
9172 library. Squid does not pre-allocate any memory, just safe-keeps
9173 objects that otherwise would be free()d. Thus, it is safe to set
9174 memory_pools_limit to a reasonably high value even if your
9175 configuration will use less memory.
9177 If set to none, Squid will keep all memory it can. That is, there
9178 will be no limit on the total amount of memory used for safe-keeping.
9180 To disable memory allocation optimization, do not set
9181 memory_pools_limit to 0 or none. Set memory_pools to "off" instead.
9183 An overhead for maintaining memory pools is not taken into account
9184 when the limit is checked. This overhead is close to four bytes per
9185 object kept. However, pools may actually _save_ memory because of
9186 reduced memory thrashing in your malloc library.
9190 COMMENT: on|off|transparent|truncate|delete
9193 LOC: opt_forwarded_for
9195 If set to "on", Squid will append your client's IP address
9196 in the HTTP requests it forwards. By default it looks like:
9198 X-Forwarded-For: 192.1.2.3
9200 If set to "off", it will appear as
9202 X-Forwarded-For: unknown
9204 If set to "transparent", Squid will not alter the
9205 X-Forwarded-For header in any way.
9207 If set to "delete", Squid will delete the entire
9208 X-Forwarded-For header.
9210 If set to "truncate", Squid will remove all existing
9211 X-Forwarded-For entries, and place the client IP as the sole entry.
9214 NAME: cachemgr_passwd
9215 TYPE: cachemgrpasswd
9217 DEFAULT_DOC: No password. Actions which require password are denied.
9218 LOC: Config.passwd_list
9220 Specify passwords for cachemgr operations.
9222 Usage: cachemgr_passwd password action action ...
9224 Some valid actions are (see cache manager menu for a full list):
9264 * Indicates actions which will not be performed without a
9265 valid password, others can be performed if not listed here.
9267 To disable an action, set the password to "disable".
9268 To allow performing an action without a password, set the
9271 Use the keyword "all" to set the same password for all actions.
9274 cachemgr_passwd secret shutdown
9275 cachemgr_passwd lesssssssecret info stats/objects
9276 cachemgr_passwd disable all
9283 LOC: Config.onoff.client_db
9285 If you want to disable collecting per-client statistics,
9286 turn off client_db here.
9289 NAME: refresh_all_ims
9293 LOC: Config.onoff.refresh_all_ims
9295 When you enable this option, squid will always check
9296 the origin server for an update when a client sends an
9297 If-Modified-Since request. Many browsers use IMS
9298 requests when the user requests a reload, and this
9299 ensures those clients receive the latest version.
9301 By default (off), squid may return a Not Modified response
9302 based on the age of the cached version.
9305 NAME: reload_into_ims
9306 IFDEF: USE_HTTP_VIOLATIONS
9310 LOC: Config.onoff.reload_into_ims
9312 When you enable this option, client no-cache or ``reload''
9313 requests will be changed to If-Modified-Since requests.
9314 Doing this VIOLATES the HTTP standard. Enabling this
9315 feature could make you liable for problems which it
9318 see also refresh_pattern for a more selective approach.
9321 NAME: connect_retries
9323 LOC: Config.connect_retries
9325 DEFAULT_DOC: Do not retry failed connections.
9327 This sets the maximum number of connection attempts made for each
9328 TCP connection. The connect_retries attempts must all still
9329 complete within the connection timeout period.
9331 The default is not to re-try if the first connection attempt fails.
9332 The (not recommended) maximum is 10 tries.
9334 A warning message will be generated if it is set to a too-high
9335 value and the configured value will be over-ridden.
9337 Note: These re-tries are in addition to forward_max_tries
9338 which limit how many different addresses may be tried to find
9342 NAME: retry_on_error
9344 LOC: Config.retry.onerror
9347 If set to ON Squid will automatically retry requests when
9348 receiving an error response with status 403 (Forbidden),
9349 500 (Internal Error), 501 or 503 (Service not available).
9350 Status 502 and 504 (Gateway errors) are always retried.
9352 This is mainly useful if you are in a complex cache hierarchy to
9353 work around access control errors.
9355 NOTE: This retry will attempt to find another working destination.
9356 Which is different from the server which just failed.
9359 NAME: as_whois_server
9361 LOC: Config.as_whois_server
9362 DEFAULT: whois.ra.net
9364 WHOIS server to query for AS numbers. NOTE: AS numbers are
9365 queried only when Squid starts up, not for every request.
9370 LOC: Config.onoff.offline
9373 Enable this option and Squid will never try to validate cached
9377 NAME: uri_whitespace
9378 TYPE: uri_whitespace
9379 LOC: Config.uri_whitespace
9382 What to do with requests that have whitespace characters in the
9385 strip: The whitespace characters are stripped out of the URL.
9386 This is the behavior recommended by RFC2396 and RFC3986
9387 for tolerant handling of generic URI.
9388 NOTE: This is one difference between generic URI and HTTP URLs.
9390 deny: The request is denied. The user receives an "Invalid
9392 This is the behaviour recommended by RFC2616 for safe
9393 handling of HTTP request URL.
9395 allow: The request is allowed and the URI is not changed. The
9396 whitespace characters remain in the URI. Note the
9397 whitespace is passed to redirector processes if they
9399 Note this may be considered a violation of RFC2616
9400 request parsing where whitespace is prohibited in the
9403 encode: The request is allowed and the whitespace characters are
9404 encoded according to RFC1738.
9406 chop: The request is allowed and the URI is chopped at the
9410 NOTE the current Squid implementation of encode and chop violates
9411 RFC2616 by not using a 301 redirect after altering the URL.
9416 LOC: Config.chroot_dir
9419 Specifies a directory where Squid should do a chroot() while
9420 initializing. This also causes Squid to fully drop root
9421 privileges after initializing. This means, for example, if you
9422 use a HTTP port less than 1024 and try to reconfigure, you may
9423 get an error saying that Squid can not open the port.
9426 NAME: balance_on_multiple_ip
9428 LOC: Config.onoff.balance_on_multiple_ip
9431 Modern IP resolvers in squid sort lookup results by preferred access.
9432 By default squid will use these IP in order and only rotates to
9433 the next listed when the most preffered fails.
9435 Some load balancing servers based on round robin DNS have been
9436 found not to preserve user session state across requests
9437 to different IP addresses.
9439 Enabling this directive Squid rotates IP's per request.
9442 NAME: pipeline_prefetch
9443 TYPE: pipelinePrefetch
9444 LOC: Config.pipeline_max_prefetch
9446 DEFAULT_DOC: Do not pre-parse pipelined requests.
9448 HTTP clients may send a pipeline of 1+N requests to Squid using a
9449 single connection, without waiting for Squid to respond to the first
9450 of those requests. This option limits the number of concurrent
9451 requests Squid will try to handle in parallel. If set to N, Squid
9452 will try to receive and process up to 1+N requests on the same
9453 connection concurrently.
9455 Defaults to 0 (off) for bandwidth management and access logging
9458 NOTE: pipelining requires persistent connections to clients.
9460 WARNING: pipelining breaks NTLM and Negotiate/Kerberos authentication.
9463 NAME: high_response_time_warning
9466 LOC: Config.warnings.high_rptm
9468 DEFAULT_DOC: disabled.
9470 If the one-minute median response time exceeds this value,
9471 Squid prints a WARNING with debug level 0 to get the
9472 administrators attention. The value is in milliseconds.
9475 NAME: high_page_fault_warning
9477 LOC: Config.warnings.high_pf
9479 DEFAULT_DOC: disabled.
9481 If the one-minute average page fault rate exceeds this
9482 value, Squid prints a WARNING with debug level 0 to get
9483 the administrators attention. The value is in page faults
9487 NAME: high_memory_warning
9489 LOC: Config.warnings.high_memory
9490 IFDEF: HAVE_MSTATS&&HAVE_GNUMALLOC_H
9492 DEFAULT_DOC: disabled.
9494 If the memory usage (as determined by gnumalloc, if available and used)
9495 exceeds this amount, Squid prints a WARNING with debug level 0 to get
9496 the administrators attention.
9498 # TODO: link high_memory_warning to mempools?
9500 NAME: sleep_after_fork
9501 COMMENT: (microseconds)
9503 LOC: Config.sleep_after_fork
9506 When this is set to a non-zero value, the main Squid process
9507 sleeps the specified number of microseconds after a fork()
9508 system call. This sleep may help the situation where your
9509 system reports fork() failures due to lack of (virtual)
9510 memory. Note, however, if you have a lot of child
9511 processes, these sleep delays will add up and your
9512 Squid will not service requests for some amount of time
9513 until all the child processes have been started.
9514 On Windows value less then 1000 (1 milliseconds) are
9518 NAME: windows_ipaddrchangemonitor
9519 IFDEF: _SQUID_WINDOWS_
9523 LOC: Config.onoff.WIN32_IpAddrChangeMonitor
9525 On Windows Squid by default will monitor IP address changes and will
9526 reconfigure itself after any detected event. This is very useful for
9527 proxies connected to internet with dial-up interfaces.
9528 In some cases (a Proxy server acting as VPN gateway is one) it could be
9529 desiderable to disable this behaviour setting this to 'off'.
9530 Note: after changing this, Squid service must be restarted.
9535 IFDEF: USE_SQUID_EUI
9537 LOC: Eui::TheConfig.euiLookup
9539 Whether to lookup the EUI or MAC address of a connected client.
9542 NAME: max_filedescriptors max_filedesc
9545 DEFAULT_DOC: Use operating system limits set by ulimit.
9546 LOC: Config.max_filedescriptors
9548 Reduce the maximum number of filedescriptors supported below
9549 the usual operating system defaults.
9551 Remove from squid.conf to inherit the current ulimit setting.
9553 Note: Changing this requires a restart of Squid. Also
9554 not all I/O types supports large values (eg on Windows).
9557 NAME: force_request_body_continuation
9559 LOC: Config.accessList.forceRequestBodyContinuation
9561 DEFAULT_DOC: Deny, unless rules exist in squid.conf.
9563 This option controls how Squid handles data upload requests from HTTP
9564 and FTP agents that require a "Please Continue" control message response
9565 to actually send the request body to Squid. It is mostly useful in
9566 adaptation environments.
9568 When Squid receives an HTTP request with an "Expect: 100-continue"
9569 header or an FTP upload command (e.g., STOR), Squid normally sends the
9570 request headers or FTP command information to an adaptation service (or
9571 peer) and waits for a response. Most adaptation services (and some
9572 broken peers) may not respond to Squid at that stage because they may
9573 decide to wait for the HTTP request body or FTP data transfer. However,
9574 that request body or data transfer may never come because Squid has not
9575 responded with the HTTP 100 or FTP 150 (Please Continue) control message
9576 to the request sender yet!
9578 An allow match tells Squid to respond with the HTTP 100 or FTP 150
9579 (Please Continue) control message on its own, before forwarding the
9580 request to an adaptation service or peer. Such a response usually forces
9581 the request sender to proceed with sending the body. A deny match tells
9582 Squid to delay that control response until the origin server confirms
9583 that the request body is needed. Delaying is the default behavior.
projects / thirdparty / squid.git / blob