· 10 years ago · Apr 13, 2016, 07:39 AM
1# WELCOME TO SQUID 2.7.STABLE8
2# ----------------------------
3#
4# This is the default Squid configuration file. You may wish
5# to look at the Squid home page (http://www.squid-cache.org/)
6# for the FAQ and other documentation.
7#
8# The default Squid config file shows what the defaults for
9# various options happen to be. If you don't need to change the
10# default, you shouldn't uncomment the line. Doing so may cause
11# run-time problems. In some cases "none" refers to no default
12# setting at all, while in other cases it refers to a valid
13# option - the comments for that keyword indicate if this is the
14# case.
15#
16
17
18# Configuration options can be included using the "include" directive.
19# Include takes a list of files to include. Quoting and wildcards is
20# supported.
21#
22# For example,
23#
24# include /path/to/included/file/squid.acl.config
25#
26# Includes can be nested up to a hard-coded depth of 16 levels.
27# This arbitrary restriction is to prevent recursive include references
28# from causing Squid entering an infinite loop whilst trying to load
29# configuration files.
30
31
32# OPTIONS FOR AUTHENTICATION
33# -----------------------------------------------------------------------------
34
35# TAG: auth_param
36# This is used to define parameters for the various authentication
37# schemes supported by Squid.
38#
39# format: auth_param scheme parameter [setting]
40#
41# The order in which authentication schemes are presented to the client is
42# dependent on the order the scheme first appears in config file. IE
43# has a bug (it's not RFC 2617 compliant) in that it will use the basic
44# scheme if basic is the first entry presented, even if more secure
45# schemes are presented. For now use the order in the recommended
46# settings section below. If other browsers have difficulties (don't
47# recognize the schemes offered even if you are using basic) either
48# put basic first, or disable the other schemes (by commenting out their
49# program entry).
50#
51# Once an authentication scheme is fully configured, it can only be
52# shutdown by shutting squid down and restarting. Changes can be made on
53# the fly and activated with a reconfigure. I.E. You can change to a
54# different helper, but not unconfigure the helper completely.
55#
56# Please note that while this directive defines how Squid processes
57# authentication it does not automatically activate authentication.
58# To use authentication you must in addition make use of ACLs based
59# on login name in http_access (proxy_auth, proxy_auth_regex or
60# external with %LOGIN used in the format tag). The browser will be
61# challenged for authentication on the first such acl encountered
62# in http_access processing and will also be re-challenged for new
63# login credentials if the request is being denied by a proxy_auth
64# type acl.
65#
66# WARNING: authentication can't be used in a transparently intercepting
67# proxy as the client then thinks it is talking to an origin server and
68# not the proxy. This is a limitation of bending the TCP/IP protocol to
69# transparently intercepting port 80, not a limitation in Squid.
70#
71# === Parameters for the basic scheme follow. ===
72#
73# "program" cmdline
74# Specify the command for the external authenticator. Such a program
75# reads a line containing "username password" and replies "OK" or
76# "ERR" in an endless loop. "ERR" responses may optionally be followed
77# by a error description available as %m in the returned error page.
78#
79# By default, the basic authentication scheme is not used unless a
80# program is specified.
81#
82# If you want to use the traditional proxy authentication, jump over to
83# the helpers/basic_auth/NCSA directory and type:
84# % make
85# % make install
86#
87# Then, set this line to something like
88#
89# auth_param basic program c:/squid/libexec/ncsa_auth c:/squid/etc/passwd
90#
91# "children" numberofchildren
92# The number of authenticator processes to spawn. If you start too few
93# squid will have to wait for them to process a backlog of credential
94# verifications, slowing it down. When credential verifications are
95# done via a (slow) network you are likely to need lots of
96# authenticator processes.
97# auth_param basic children 5
98#
99# "concurrency" numberofconcurrentrequests
100# The number of concurrent requests/channels the helper supports.
101# Changes the protocol used to include a channel number first on
102# the request/response line, allowing multiple requests to be sent
103# to the same helper in parallell without wating for the response.
104# Must not be set unless it's known the helper supports this.
105#
106# "realm" realmstring
107# Specifies the realm name which is to be reported to the client for
108# the basic proxy authentication scheme (part of the text the user
109# will see when prompted their username and password).
110# auth_param basic realm Squid proxy-caching web server
111#
112# "credentialsttl" timetolive
113# Specifies how long squid assumes an externally validated
114# username:password pair is valid for - in other words how often the
115# helper program is called for that user. Set this low to force
116# revalidation with short lived passwords. Note that setting this high
117# does not impact your susceptibility to replay attacks unless you are
118# using an one-time password system (such as SecureID). If you are using
119# such a system, you will be vulnerable to replay attacks unless you
120# also use the max_user_ip ACL in an http_access rule.
121# auth_param basic credentialsttl 2 hours
122#
123# "casesensitive" on|off
124# Specifies if usernames are case sensitive. Most user databases are
125# case insensitive allowing the same username to be spelled using both
126# lower and upper case letters, but some are case sensitive. This
127# makes a big difference for user_max_ip ACL processing and similar.
128# auth_param basic casesensitive off
129#
130# "blankpassword" on|off
131# Specifies if blank passwords should be supported. Defaults to off
132# as there is multiple authentication backends which handles blank
133# passwords as "guest" access.
134#
135# === Parameters for the digest scheme follow ===
136#
137# "program" cmdline
138# Specify the command for the external authenticator. Such a program
139# reads a line containing "username":"realm" and replies with the
140# appropriate H(A1) value hex encoded or ERR if the user (or his H(A1)
141# hash) does not exists. See RFC 2616 for the definition of H(A1).
142# "ERR" responses may optionally be followed by a error description
143# available as %m in the returned error page.
144#
145# By default, the digest authentication scheme is not used unless a
146# program is specified.
147#
148# If you want to use a digest authenticator, jump over to the
149# helpers/digest_auth/ directory and choose the authenticator to use.
150# It it's directory type
151# % make
152# % make install
153#
154# Then, set this line to something like
155#
156# auth_param digest program c:/squid/libexec/digest_auth_pw c:/squid/etc/digpass
157#
158# "children" numberofchildren
159# The number of authenticator processes to spawn. If you start too few
160# squid will have to wait for them to process a backlog of credential
161# verifications, slowing it down. When credential verifications are
162# done via a (slow) network you are likely to need lots of
163# authenticator processes.
164# auth_param digest children 5
165#
166# "concurrency" numberofconcurrentrequests
167# The number of concurrent requests/channels the helper supports.
168# Changes the protocol used to include a channel number first on
169# the request/response line, allowing multiple requests to be sent
170# to the same helper in parallell without wating for the response.
171# Must not be set unless it's known the helper supports this.
172#
173# "realm" realmstring
174# Specifies the realm name which is to be reported to the client for the
175# digest proxy authentication scheme (part of the text the user will see
176# when prompted their username and password).
177# auth_param digest realm Squid proxy-caching web server
178#
179# "nonce_garbage_interval" timeinterval
180# Specifies the interval that nonces that have been issued to clients are
181# checked for validity.
182# auth_param digest nonce_garbage_interval 5 minutes
183#
184# "nonce_max_duration" timeinterval
185# Specifies the maximum length of time a given nonce will be valid for.
186# auth_param digest nonce_max_duration 30 minutes
187#
188# "nonce_max_count" number
189# Specifies the maximum number of times a given nonce can be used.
190# auth_param digest nonce_max_count 50
191#
192# "nonce_strictness" on|off
193# Determines if squid requires strict increment-by-1 behavior for nonce
194# counts, or just incrementing (off - for use when useragents generate
195# nonce counts that occasionally miss 1 (ie, 1,2,4,6)).
196# auth_param digest nonce_strictness off
197#
198# "check_nonce_count" on|off
199# This directive if set to off can disable the nonce count check
200# completely to work around buggy digest qop implementations in certain
201# mainstream browser versions. Default on to check the nonce count to
202# protect from authentication replay attacks.
203# auth_param digest check_nonce_count on
204#
205# "post_workaround" on|off
206# This is a workaround to certain buggy browsers who sends an incorrect
207# request digest in POST requests when reusing the same nonce as acquired
208# earlier in response to a GET request.
209# auth_param digest post_workaround off
210#
211# === NTLM scheme options follow ===
212#
213# "program" cmdline
214# Specify the command for the external NTLM authenticator. Such a
215# program participates in the NTLMSSP exchanges between Squid and the
216# client and reads commands according to the Squid NTLMSSP helper
217# protocol. See helpers/ntlm_auth/ for details. Recommended ntlm
218# authenticator is ntlm_auth from Samba-3.X, but a number of other
219# ntlm authenticators is available.
220#
221# By default, the ntlm authentication scheme is not used unless a
222# program is specified.
223#
224# auth_param ntlm program /path/to/samba/bin/ntlm_auth --helper-protocol=squid-2.5-ntlmssp
225#
226# "children" numberofchildren
227# The number of authenticator processes to spawn. If you start too few
228# squid will have to wait for them to process a backlog of credential
229# verifications, slowing it down. When credential verifications are
230# done via a (slow) network you are likely to need lots of
231# authenticator processes.
232# auth_param ntlm children 5
233#
234# "keep_alive" on|off
235# This option enables the use of keep-alive on the initial
236# authentication request. It has been reported some versions of MSIE
237# have problems if this is enabled, but performance will be increased
238# if enabled.
239#
240# auth_param ntlm keep_alive on
241#
242# === Negotiate scheme options follow ===
243#
244# "program" cmdline
245# Specify the command for the external Negotiate authenticator. Such a
246# program participates in the SPNEGO exchanges between Squid and the
247# client and reads commands according to the Squid ntlmssp helper
248# protocol. See helpers/ntlm_auth/ for details. Recommended SPNEGO
249# authenticator is ntlm_auth from Samba-4.X.
250#
251# By default, the Negotiate authentication scheme is not used unless a
252# program is specified.
253#
254# auth_param negotiate program /path/to/samba/bin/ntlm_auth --helper-protocol=gss-spnego
255#
256# "children" numberofchildren
257# The number of authenticator processes to spawn. If you start too few
258# squid will have to wait for them to process a backlog of credential
259# verifications, slowing it down. When credential verifications are
260# done via a (slow) network you are likely to need lots of
261# authenticator processes.
262# auth_param negotiate children 5
263#
264# "keep_alive" on|off
265# If you experience problems with PUT/POST requests when using the
266# Negotiate authentication scheme then you can try setting this to
267# off. This will cause Squid to forcibly close the connection on
268# the initial requests where the browser asks which schemes are
269# supported by the proxy.
270#
271# auth_param negotiate keep_alive on
272#
273#Recommended minimum configuration per scheme:
274#auth_param negotiate program <uncomment and complete this line to activate>
275#auth_param negotiate children 5
276#auth_param negotiate keep_alive on
277#auth_param ntlm program <uncomment and complete this line to activate>
278#auth_param ntlm children 5
279#auth_param ntlm keep_alive on
280#auth_param digest program <uncomment and complete this line>
281#auth_param digest children 5
282#auth_param digest realm Squid proxy-caching web server
283#auth_param digest nonce_garbage_interval 5 minutes
284#auth_param digest nonce_max_duration 30 minutes
285#auth_param digest nonce_max_count 50
286#auth_param basic program <uncomment and complete this line>
287#auth_param basic children 5
288#auth_param basic realm Squid proxy-caching web server
289#auth_param basic credentialsttl 2 hours
290#auth_param basic casesensitive off
291
292# TAG: authenticate_cache_garbage_interval
293# The time period between garbage collection across the username cache.
294# This is a tradeoff between memory utilization (long intervals - say
295# 2 days) and CPU (short intervals - say 1 minute). Only change if you
296# have good reason to.
297#
298#Default:
299# authenticate_cache_garbage_interval 1 hour
300
301# TAG: authenticate_ttl
302# The time a user & their credentials stay in the logged in user cache
303# since their last request. When the garbage interval passes, all user
304# credentials that have passed their TTL are removed from memory.
305#
306#Default:
307# authenticate_ttl 1 hour
308
309# TAG: authenticate_ip_ttl
310# If you use proxy authentication and the 'max_user_ip' ACL, this
311# directive controls how long Squid remembers the IP addresses
312# associated with each user. Use a small value (e.g., 60 seconds) if
313# your users might change addresses quickly, as is the case with
314# dialups. You might be safe using a larger value (e.g., 2 hours) in a
315# corporate LAN environment with relatively static address assignments.
316#
317#Default:
318# authenticate_ip_ttl 0 seconds
319
320# TAG: authenticate_ip_shortcircuit_ttl
321# Cache authentication credentials per client IP address for this
322# long. Default is 0 seconds (disabled).
323#
324# See also authenticate_ip_shortcircuit_access directive.
325#
326#Default:
327# authenticate_ip_shortcircuit_ttl 0 seconds
328
329
330# ACCESS CONTROLS
331# -----------------------------------------------------------------------------
332
333# TAG: external_acl_type
334# This option defines external acl classes using a helper program to
335# look up the status
336#
337# external_acl_type name [options] FORMAT.. /path/to/helper [helper arguments..]
338#
339# Options:
340#
341# ttl=n TTL in seconds for cached results (defaults to 3600
342# for 1 hour)
343# negative_ttl=n
344# TTL for cached negative lookups (default same
345# as ttl)
346# children=n number of processes spawn to service external acl
347# lookups of this type. (default 5).
348# concurrency=n concurrency level per process. Only used with helpers
349# capable of processing more than one query at a time.
350# Note: see compatibility note below
351# cache=n result cache size, 0 is unbounded (default)
352# grace= Percentage remaining of TTL where a refresh of a
353# cached entry should be initiated without needing to
354# wait for a new reply. (default 0 for no grace period)
355# protocol=2.5 Compatibility mode for Squid-2.5 external acl helpers
356#
357# FORMAT specifications
358#
359# %LOGIN Authenticated user login name
360# %EXT_USER Username from external acl
361# %IDENT Ident user name
362# %SRC Client IP
363# %SRCPORT Client source port
364# %URI Requested URI
365# %DST Requested host
366# %PROTO Requested protocol
367# %PORT Requested port
368# %METHOD Request method
369# %MYADDR Squid interface address
370# %MYPORT Squid http_port number
371# %PATH Requested URL-path (including query-string if any)
372# %USER_CERT SSL User certificate in PEM format
373# %USER_CERTCHAIN SSL User certificate chain in PEM format
374# %USER_CERT_xx SSL User certificate subject attribute xx
375# %USER_CA_xx SSL User certificate issuer attribute xx
376# %{Header} HTTP request header "Header"
377# %{Hdr:member} HTTP request header "Hdr" list member "member"
378# %{Hdr:;member}
379# HTTP request header list member using ; as
380# list separator. ; can be any non-alphanumeric
381# character.
382# %ACL The ACL name
383# %DATA The ACL arguments. If not used then any arguments
384# is automatically added at the end
385#
386# In addition to the above, any string specified in the referencing
387# acl will also be included in the helper request line, after the
388# specified formats (see the "acl external" directive)
389#
390# The helper receives lines per the above format specification,
391# and returns lines starting with OK or ERR indicating the validity
392# of the request and optionally followed by additional keywords with
393# more details.
394#
395# General result syntax:
396#
397# OK/ERR keyword=value ...
398#
399# Defined keywords:
400#
401# user= The users name (login also understood)
402# password= The users password (for PROXYPASS login= cache_peer)
403# message= Error message or similar used as %o in error messages
404# (error also understood)
405# log= String to be logged in access.log. Available as
406# %ea in logformat specifications
407#
408# If protocol=3.0 (the default) then URL escaping is used to protect
409# each value in both requests and responses.
410#
411# If using protocol=2.5 then all values need to be enclosed in quotes
412# if they may contain whitespace, or the whitespace escaped using \.
413# And quotes or \ characters within the keyword value must be \ escaped.
414#
415# When using the concurrency= option the protocol is changed by
416# introducing a query channel tag infront of the request/response.
417# The query channel tag is a number between 0 and concurrency-1.
418#
419# Compatibility Note: The children= option was named concurrency= in
420# Squid-2.5.STABLE3 and earlier, and was accepted as an alias for the
421# duration of the Squid-2.5 releases to keep compatibility. However,
422# the meaning of concurrency= option has changed in Squid-2.6 to match
423# that of Squid-3 and the old syntax no longer works.
424#
425#Default:
426# none
427
428# TAG: acl
429# Defining an Access List
430#
431# Every access list definition must begin with an aclname and acltype,
432# followed by either type-specific arguments or a quoted filename that
433# they are read from.
434#
435# acl aclname acltype argument ...
436# acl aclname acltype "file" ...
437#
438# when using "file", the file should contain one item per line.
439#
440# By default, regular expressions are CASE-SENSITIVE. To make
441# them case-insensitive, use the -i option.
442#
443# acl aclname src ip-address/netmask ... (clients IP address)
444# acl aclname src addr1-addr2/netmask ... (range of addresses)
445# acl aclname dst ip-address/netmask ... (URL host's IP address)
446# acl aclname myip ip-address/netmask ... (local socket IP address)
447#
448# acl aclname arp mac-address ... (xx:xx:xx:xx:xx:xx notation)
449# # The arp ACL requires the special configure option --enable-arp-acl.
450# # Furthermore, the arp ACL code is not portable to all operating systems.
451# # It works on Linux, Solaris, FreeBSD and some other *BSD variants.
452# #
453# # NOTE: Squid can only determine the MAC address for clients that are on
454# # the same subnet. If the client is on a different subnet, then Squid cannot
455# # find out its MAC address.
456#
457# acl aclname srcdomain .foo.com ... # reverse lookup, client IP
458# acl aclname dstdomain .foo.com ... # Destination server from URL
459# acl aclname srcdom_regex [-i] xxx ... # regex matching client name
460# acl aclname dstdom_regex [-i] xxx ... # regex matching server
461# # For dstdomain and dstdom_regex a reverse lookup is tried if a IP
462# # based URL is used and no match is found. The name "none" is used
463# # if the reverse lookup fails.
464#
465# acl aclname time [day-abbrevs] [h1:m1-h2:m2]
466# # day-abbrevs:
467# # S - Sunday
468# # M - Monday
469# # T - Tuesday
470# # W - Wednesday
471# # H - Thursday
472# # F - Friday
473# # A - Saturday
474# # h1:m1 must be less than h2:m2
475# acl aclname url_regex [-i] ^http:// ... # regex matching on whole URL
476# acl aclname urlpath_regex [-i] \.gif$ ... # regex matching on URL path
477# acl aclname urllogin [-i] [^a-zA-Z0-9] ... # regex matching on URL login field
478# acl aclname port 80 70 21 ...
479# acl aclname port 0-1024 ... # ranges allowed
480# acl aclname myport 3128 ... # (local socket TCP port)
481# acl aclname myportname 3128 ... # http(s)_port name
482# acl aclname proto HTTP FTP ...
483# acl aclname method GET POST ...
484# acl aclname browser [-i] regexp ...
485# # pattern match on User-Agent header (see also req_header below)
486# acl aclname referer_regex [-i] regexp ...
487# # pattern match on Referer header
488# # Referer is highly unreliable, so use with care
489# acl aclname ident username ...
490# acl aclname ident_regex [-i] pattern ...
491# # string match on ident output.
492# # use REQUIRED to accept any non-null ident.
493# acl aclname src_as number ...
494# acl aclname dst_as number ...
495# # Except for access control, AS numbers can be used for
496# # routing of requests to specific caches. Here's an
497# # example for routing all requests for AS#1241 and only
498# # those to mycache.mydomain.net:
499# # acl asexample dst_as 1241
500# # cache_peer_access mycache.mydomain.net allow asexample
501# # cache_peer_access mycache_mydomain.net deny all
502#
503# acl aclname proxy_auth [-i] username ...
504# acl aclname proxy_auth_regex [-i] pattern ...
505# # list of valid usernames
506# # use REQUIRED to accept any valid username.
507# #
508# # NOTE: when a Proxy-Authentication header is sent but it is not
509# # needed during ACL checking the username is NOT logged
510# # in access.log.
511# #
512# # NOTE: proxy_auth requires a EXTERNAL authentication program
513# # to check username/password combinations (see
514# # auth_param directive).
515# #
516# # NOTE: proxy_auth can't be used in a transparent proxy as
517# # the browser needs to be configured for using a proxy in order
518# # to respond to proxy authentication.
519#
520# acl aclname snmp_community string ...
521# # A community string to limit access to your SNMP Agent
522# # Example:
523# #
524# # acl snmppublic snmp_community public
525#
526# acl aclname maxconn number
527# # This will be matched when the client's IP address has
528# # more than <number> HTTP connections established.
529#
530# acl aclname max_user_ip [-s] number
531# # This will be matched when the user attempts to log in from more
532# # than <number> different ip addresses. The authenticate_ip_ttl
533# # parameter controls the timeout on the ip entries.
534# # If -s is specified the limit is strict, denying browsing
535# # from any further IP addresses until the ttl has expired. Without
536# # -s Squid will just annoy the user by "randomly" denying requests.
537# # (the counter is reset each time the limit is reached and a
538# # request is denied)
539# # NOTE: in acceleration mode or where there is mesh of child proxies,
540# # clients may appear to come from multiple addresses if they are
541# # going through proxy farms, so a limit of 1 may cause user problems.
542#
543# acl aclname req_mime_type mime-type ...
544# # regex match against the mime type of the request generated
545# # by the client. Can be used to detect file upload or some
546# # types HTTP tunneling requests.
547# # NOTE: This does NOT match the reply. You cannot use this
548# # to match the returned file type.
549#
550# acl aclname req_header header-name [-i] any\.regex\.here
551# # regex match against any of the known request headers. May be
552# # thought of as a superset of "browser", "referer" and "mime-type"
553# # ACLs.
554#
555# acl aclname rep_mime_type mime-type ...
556# # regex match against the mime type of the reply received by
557# # squid. Can be used to detect file download or some
558# # types HTTP tunneling requests.
559# # NOTE: This has no effect in http_access rules. It only has
560# # effect in rules that affect the reply data stream such as
561# # http_reply_access.
562#
563# acl aclname rep_header header-name [-i] any\.regex\.here
564# # regex match against any of the known reply headers. May be
565# # thought of as a superset of "browser", "referer" and "mime-type"
566# # ACLs.
567# #
568# # Example:
569# #
570# # acl many_spaces rep_header Content-Disposition -i [[:space:]]{3,}
571#
572# acl aclname external class_name [arguments...]
573# # external ACL lookup via a helper class defined by the
574# # external_acl_type directive.
575#
576# acl aclname urlgroup group1 ...
577# # match against the urlgroup as indicated by redirectors
578#
579# acl aclname user_cert attribute values...
580# # match against attributes in a user SSL certificate
581# # attribute is one of DN/C/O/CN/L/ST
582#
583# acl aclname ca_cert attribute values...
584# # match against attributes a users issuing CA SSL certificate
585# # attribute is one of DN/C/O/CN/L/ST
586#
587# acl aclname ext_user username ...
588# acl aclname ext_user_regex [-i] pattern ...
589# # string match on username returned by external acl helper
590# # use REQUIRED to accept any non-null user name.
591#
592#Examples:
593#acl macaddress arp 09:00:2b:23:45:67
594#acl myexample dst_as 1241
595#acl password proxy_auth REQUIRED
596#acl fileupload req_mime_type -i ^multipart/form-data$
597#acl javascript rep_mime_type -i ^application/x-javascript$
598#
599#Recommended minimum configuration:
600acl all src all
601acl manager proto cache_object
602acl localhost src 127.0.0.1/32
603acl to_localhost dst 127.0.0.0/8 0.0.0.0/32
604#
605# Example rule allowing access from your local networks.
606# Adapt to list your (internal) IP networks from where browsing
607# should be allowed
608acl localnet src 10.0.0.0/8 # RFC1918 possible internal network
609acl localnet src 172.16.0.0/12 # RFC1918 possible internal network
610acl localnet src 192.168.0.0/16 # RFC1918 possible internal network
611#
612acl SSL_ports port 443
613acl Safe_ports port 80 # http
614acl Safe_ports port 21 # ftp
615acl Safe_ports port 443 # https
616acl Safe_ports port 70 # gopher
617acl Safe_ports port 210 # wais
618acl Safe_ports port 1025-65535 # unregistered ports
619acl Safe_ports port 280 # http-mgmt
620acl Safe_ports port 488 # gss-http
621acl Safe_ports port 591 # filemaker
622acl Safe_ports port 777 # multiling http
623acl CONNECT method CONNECT
624
625
626# TAG: http_access
627# Allowing or Denying access based on defined access lists
628#
629# Access to the HTTP port:
630# http_access allow|deny [!]aclname ...
631#
632# NOTE on default values:
633#
634# If there are no "access" lines present, the default is to deny
635# the request.
636#
637# If none of the "access" lines cause a match, the default is the
638# opposite of the last line in the list. If the last line was
639# deny, the default is allow. Conversely, if the last line
640# is allow, the default will be deny. For these reasons, it is a
641# good idea to have an "deny all" or "allow all" entry at the end
642# of your access lists to avoid potential confusion.
643#
644#Default:
645# http_access deny all
646#
647#Recommended minimum configuration:
648#
649# Only allow cachemgr access from localhost
650http_access allow manager localhost
651http_access deny manager
652# Deny requests to unknown ports
653http_access deny !Safe_ports
654# Deny CONNECT to other than SSL ports
655http_access deny CONNECT !SSL_ports
656#
657# We strongly recommend the following be uncommented to protect innocent
658# web applications running on the proxy server who think the only
659# one who can access services on "localhost" is a local user
660#http_access deny to_localhost
661#
662# INSERT YOUR OWN RULE(S) HERE TO ALLOW ACCESS FROM YOUR CLIENTS
663
664# Example rule allowing access from your local networks.
665# Adapt localnet in the ACL section to list your (internal) IP networks
666# from where browsing should be allowed
667http_access allow localnet
668
669# And finally deny all other access to this proxy
670# http_access deny all
671
672http_access allow all
673http_access
674
675
676# TAG: http_access2
677# Allowing or Denying access based on defined access lists
678#
679# Identical to http_access, but runs after redirectors. If not set
680# then only http_access is used.
681#
682#Default:
683# none
684
685# TAG: http_reply_access
686# Allow replies to client requests. This is complementary to http_access.
687#
688# http_reply_access allow|deny [!] aclname ...
689#
690# NOTE: if there are no access lines present, the default is to allow
691# all replies
692#
693# If none of the access lines cause a match the opposite of the
694# last line will apply. Thus it is good practice to end the rules
695# with an "allow all" or "deny all" entry.
696#
697#Default:
698# http_reply_access allow all
699
700# TAG: icp_access
701# Allowing or Denying access to the ICP port based on defined
702# access lists
703#
704# icp_access allow|deny [!]aclname ...
705#
706# See http_access for details
707#
708#Default:
709# icp_access deny all
710#
711#Allow ICP queries from local networks only
712icp_access allow localnet
713icp_access deny all
714
715# TAG: htcp_access
716# Allowing or Denying access to the HTCP port based on defined
717# access lists
718#
719# htcp_access allow|deny [!]aclname ...
720#
721# See http_access for details
722#
723# NOTE: The default if no htcp_access lines are present is to
724# deny all traffic. This default may cause problems with peers
725# using the htcp or htcp-oldsquid options.
726#
727#Default:
728# htcp_access deny all
729#
730#Allow HTCP queries from local networks only
731# htcp_access allow localnet
732# htcp_access deny all
733
734# TAG: htcp_clr_access
735# Allowing or Denying access to purge content using HTCP based
736# on defined access lists
737#
738# htcp_clr_access allow|deny [!]aclname ...
739#
740# See http_access for details
741#
742##Allow HTCP CLR requests from trusted peers
743#acl htcp_clr_peer src 172.16.1.2
744#htcp_clr_access allow htcp_clr_peer
745#
746#Default:
747# htcp_clr_access deny all
748
749# TAG: miss_access
750# Use to force your neighbors to use you as a sibling instead of
751# a parent. For example:
752#
753# acl localclients src 172.16.0.0/16
754# miss_access allow localclients
755# miss_access deny !localclients
756#
757# This means only your local clients are allowed to fetch
758# MISSES and all other clients can only fetch HITS.
759#
760# By default, allow all clients who passed the http_access rules
761# to fetch MISSES from us.
762#
763#Default setting:
764# miss_access allow all
765
766# TAG: ident_lookup_access
767# A list of ACL elements which, if matched, cause an ident
768# (RFC931) lookup to be performed for this request. For
769# example, you might choose to always perform ident lookups
770# for your main multi-user Unix boxes, but not for your Macs
771# and PCs. By default, ident lookups are not performed for
772# any requests.
773#
774# To enable ident lookups for specific client addresses, you
775# can follow this example:
776#
777# acl ident_aware_hosts src 198.168.1.0/255.255.255.0
778# ident_lookup_access allow ident_aware_hosts
779# ident_lookup_access deny all
780#
781# Only src type ACL checks are fully supported. A src_domain
782# ACL might work at times, but it will not always provide
783# the correct result.
784#
785#Default:
786# ident_lookup_access deny all
787
788# TAG: reply_body_max_size bytes deny acl acl...
789# This option specifies the maximum size of a reply body in bytes.
790# It can be used to prevent users from downloading very large files,
791# such as MP3's and movies. When the reply headers are received,
792# the reply_body_max_size lines are processed, and the first line with
793# a result of "deny" is used as the maximum body size for this reply.
794# This size is checked twice. First when we get the reply headers,
795# we check the content-length value. If the content length value exists
796# and is larger than the allowed size, the request is denied and the
797# user receives an error message that says "the request or reply
798# is too large." If there is no content-length, and the reply
799# size exceeds this limit, the client's connection is just closed
800# and they will receive a partial reply.
801#
802# WARNING: downstream caches probably can not detect a partial reply
803# if there is no content-length header, so they will cache
804# partial responses and give them out as hits. You should NOT
805# use this option if you have downstream caches.
806#
807# If you set this parameter to zero (the default), there will be
808# no limit imposed.
809#
810#Default:
811# reply_body_max_size 0 allow all
812
813# TAG: authenticate_ip_shortcircuit_access
814# Access list determining when shortcicuiting the authentication process
815# based on source IP cached credentials is acceptable. Use this to deny
816# using the ip auth cache on requests from child proxies or other source
817# ip's having multiple users.
818#
819# See also authenticate_ip_shortcircuit_ttl directive
820#
821#Default:
822# none
823
824
825# OPTIONS FOR X-Forwarded-For
826# -----------------------------------------------------------------------------
827
828# TAG: follow_x_forwarded_for
829# Allowing or Denying the X-Forwarded-For header to be followed to
830# find the original source of a request.
831#
832# Requests may pass through a chain of several other proxies
833# before reaching us. The X-Forwarded-For header will contain a
834# comma-separated list of the IP addresses in the chain, with the
835# rightmost address being the most recent.
836#
837# If a request reaches us from a source that is allowed by this
838# configuration item, then we consult the X-Forwarded-For header
839# to see where that host received the request from. If the
840# X-Forwarded-For header contains multiple addresses, and if
841# acl_uses_indirect_client is on, then we continue backtracking
842# until we reach an address for which we are not allowed to
843# follow the X-Forwarded-For header, or until we reach the first
844# address in the list. (If acl_uses_indirect_client is off, then
845# it's impossible to backtrack through more than one level of
846# X-Forwarded-For addresses.)
847#
848# The end result of this process is an IP address that we will
849# refer to as the indirect client address. This address may
850# be treated as the client address for access control, delay
851# pools and logging, depending on the acl_uses_indirect_client,
852# delay_pool_uses_indirect_client and log_uses_indirect_client
853# options.
854#
855# SECURITY CONSIDERATIONS:
856#
857# Any host for which we follow the X-Forwarded-For header
858# can place incorrect information in the header, and Squid
859# will use the incorrect information as if it were the
860# source address of the request. This may enable remote
861# hosts to bypass any access control restrictions that are
862# based on the client's source addresses.
863#
864# For example:
865#
866# acl localhost src 127.0.0.1
867# acl my_other_proxy srcdomain .proxy.example.com
868# follow_x_forwarded_for allow localhost
869# follow_x_forwarded_for allow my_other_proxy
870#
871#Default:
872# follow_x_forwarded_for deny all
873
874# TAG: acl_uses_indirect_client on|off
875# Controls whether the indirect client address
876# (see follow_x_forwarded_for) is used instead of the
877# direct client address in acl matching.
878#
879#Default:
880# acl_uses_indirect_client on
881
882# TAG: delay_pool_uses_indirect_client on|off
883# Controls whether the indirect client address
884# (see follow_x_forwarded_for) is used instead of the
885# direct client address in delay pools.
886#
887#Default:
888# delay_pool_uses_indirect_client on
889
890# TAG: log_uses_indirect_client on|off
891# Controls whether the indirect client address
892# (see follow_x_forwarded_for) is used instead of the
893# direct client address in the access log.
894#
895#Default:
896# log_uses_indirect_client on
897
898
899# SSL OPTIONS
900# -----------------------------------------------------------------------------
901
902# TAG: ssl_unclean_shutdown
903# Note: This option is only available if Squid is rebuilt with the
904# --enable-ssl option
905#
906# Some browsers (especially MSIE) bugs out on SSL shutdown
907# messages.
908#
909#Default:
910# ssl_unclean_shutdown off
911
912# TAG: ssl_engine
913# Note: This option is only available if Squid is rebuilt with the
914# --enable-ssl option
915#
916# The OpenSSL engine to use. You will need to set this if you
917# would like to use hardware SSL acceleration for example.
918#
919#Default:
920# none
921
922# TAG: sslproxy_client_certificate
923# Note: This option is only available if Squid is rebuilt with the
924# --enable-ssl option
925#
926# Client SSL Certificate to use when proxying https:// URLs
927#
928#Default:
929# none
930
931# TAG: sslproxy_client_key
932# Note: This option is only available if Squid is rebuilt with the
933# --enable-ssl option
934#
935# Client SSL Key to use when proxying https:// URLs
936#
937#Default:
938# none
939
940# TAG: sslproxy_version
941# Note: This option is only available if Squid is rebuilt with the
942# --enable-ssl option
943#
944# SSL version level to use when proxying https:// URLs
945#
946#Default:
947# sslproxy_version 1
948
949# TAG: sslproxy_options
950# Note: This option is only available if Squid is rebuilt with the
951# --enable-ssl option
952#
953# SSL engine options to use when proxying https:// URLs
954#
955#Default:
956# none
957
958# TAG: sslproxy_cipher
959# Note: This option is only available if Squid is rebuilt with the
960# --enable-ssl option
961#
962# SSL cipher list to use when proxying https:// URLs
963#
964#Default:
965# none
966
967# TAG: sslproxy_cafile
968# Note: This option is only available if Squid is rebuilt with the
969# --enable-ssl option
970#
971# file containing CA certificates to use when verifying server
972# certificates while proxying https:// URLs
973#
974#Default:
975# none
976
977# TAG: sslproxy_capath
978# Note: This option is only available if Squid is rebuilt with the
979# --enable-ssl option
980#
981# directory containing CA certificates to use when verifying
982# server certificates while proxying https:// URLs
983#
984#Default:
985# none
986
987# TAG: sslproxy_flags
988# Note: This option is only available if Squid is rebuilt with the
989# --enable-ssl option
990#
991# Various flags modifying the use of SSL while proxying https:// URLs:
992# DONT_VERIFY_PEER Accept certificates even if they fail to
993# verify.
994# NO_DEFAULT_CA Don't use the default CA list built in
995# to OpenSSL.
996#
997#Default:
998# none
999
1000# TAG: sslpassword_program
1001# Note: This option is only available if Squid is rebuilt with the
1002# --enable-ssl option
1003#
1004# Specify a program used for entering SSL key passphrases
1005# when using encrypted SSL certificate keys. If not specified
1006# keys must either be unencrypted, or Squid started with the -N
1007# option to allow it to query interactively for the passphrase.
1008#
1009#Default:
1010# none
1011
1012
1013# NETWORK OPTIONS
1014# -----------------------------------------------------------------------------
1015
1016# TAG: http_port
1017# Usage: port [options]
1018# hostname:port [options]
1019# 1.2.3.4:port [options]
1020#
1021# The socket addresses where Squid will listen for HTTP client
1022# requests. You may specify multiple socket addresses.
1023# There are three forms: port alone, hostname with port, and
1024# IP address with port. If you specify a hostname or IP
1025# address, Squid binds the socket to that specific
1026# address. This replaces the old 'tcp_incoming_address'
1027# option. Most likely, you do not need to bind to a specific
1028# address, so you can use the port number alone.
1029#
1030# If you are running Squid in accelerator mode, you
1031# probably want to listen on port 80 also, or instead.
1032#
1033# The -I command line option will override the *first* port
1034# specified here.
1035#
1036# You may specify multiple socket addresses on multiple lines.
1037#
1038# Options:
1039#
1040# transparent Support for transparent interception of
1041# outgoing requests without browser settings.
1042#
1043# tproxy Support Linux TPROXY for spoofing outgoing
1044# connections using the client IP address.
1045#
1046# accel Accelerator mode. See also the related vhost,
1047# vport and defaultsite directives.
1048#
1049# defaultsite=domainname
1050# What to use for the Host: header if it is not present
1051# in a request. Determines what site (not origin server)
1052# accelerators should consider the default.
1053# Defaults to visible_hostname:port if not set
1054# May be combined with vport=NN to override the port number.
1055# Implies accel.
1056#
1057# vhost Accelerator mode using Host header for virtual
1058# domain support. Implies accel.
1059#
1060# vport Accelerator with IP based virtual host support.
1061# Implies accel.
1062#
1063# vport=NN As above, but uses specified port number rather
1064# than the http_port number. Implies accel.
1065#
1066# allow-direct Allow direct forwarding in accelerator mode. Normally
1067# accelerated requests is denied direct forwarding as it
1068# never_direct was used.
1069#
1070# urlgroup= Default urlgroup to mark requests with (see
1071# also acl urlgroup and url_rewrite_program)
1072#
1073# protocol= Protocol to reconstruct accelerated requests with.
1074# Defaults to http.
1075#
1076# no-connection-auth
1077# Prevent forwarding of Microsoft connection oriented
1078# authentication (NTLM, Negotiate and Kerberos)
1079#
1080# act-as-origin
1081# Act is if this Squid is the origin server.
1082# This currently means generate own Date: and
1083# Expires: headers. Implies accel.
1084#
1085# http11 Enables HTTP/1.1 support to clients. The HTTP/1.1
1086# support is still incomplete with an internal HTTP/1.0
1087# hop, but should work with most clients. The main
1088# HTTP/1.1 features missing due to this is forwarding
1089# of requests using chunked transfer encoding (results
1090# in 411) and forwarding of 1xx responses (silently
1091# dropped)
1092#
1093# name= Specifies a internal name for the port. Defaults to
1094# the port specification (port or addr:port)
1095#
1096# tcpkeepalive[=idle,interval,timeout]
1097# Enable TCP keepalive probes of idle connections
1098# idle is the initial time before TCP starts probing
1099# the connection, interval how often to probe, and
1100# timeout the time before giving up.
1101#
1102# If you run Squid on a dual-homed machine with an internal
1103# and an external interface we recommend you to specify the
1104# internal address:port in http_port. This way Squid will only be
1105# visible on the internal address.
1106#
1107# Squid normally listens to port 3128
1108http_port 3128
1109
1110# TAG: https_port
1111# Note: This option is only available if Squid is rebuilt with the
1112# --enable-ssl option
1113#
1114# Usage: [ip:]port cert=certificate.pem [key=key.pem] [options...]
1115#
1116# The socket address where Squid will listen for HTTPS client
1117# requests.
1118#
1119# This is really only useful for situations where you are running
1120# squid in accelerator mode and you want to do the SSL work at the
1121# accelerator level.
1122#
1123# You may specify multiple socket addresses on multiple lines,
1124# each with their own SSL certificate and/or options.
1125#
1126# Options:
1127#
1128# In addition to the options specified for http_port the folling
1129# SSL related options is supported:
1130#
1131# cert= Path to SSL certificate (PEM format).
1132#
1133# key= Path to SSL private key file (PEM format)
1134# if not specified, the certificate file is
1135# assumed to be a combined certificate and
1136# key file.
1137#
1138# version= The version of SSL/TLS supported
1139# 1 automatic (default)
1140# 2 SSLv2 only
1141# 3 SSLv3 only
1142# 4 TLSv1 only
1143#
1144# cipher= Colon separated list of supported ciphers.
1145#
1146# options= Various SSL engine options. The most important
1147# being:
1148# NO_SSLv2 Disallow the use of SSLv2
1149# NO_SSLv3 Disallow the use of SSLv3
1150# NO_TLSv1 Disallow the use of TLSv1
1151# SINGLE_DH_USE Always create a new key when using
1152# temporary/ephemeral DH key exchanges
1153# See src/ssl_support.c or OpenSSL SSL_CTX_set_options
1154# documentation for a complete list of options.
1155#
1156# clientca= File containing the list of CAs to use when
1157# requesting a client certificate.
1158#
1159# cafile= File containing additional CA certificates to
1160# use when verifying client certificates. If unset
1161# clientca will be used.
1162#
1163# capath= Directory containing additional CA certificates
1164# and CRL lists to use when verifying client certificates.
1165#
1166# crlfile= File of additional CRL lists to use when verifying
1167# the client certificate, in addition to CRLs stored in
1168# the capath. Implies VERIFY_CRL flag below.
1169#
1170# dhparams= File containing DH parameters for temporary/ephemeral
1171# DH key exchanges.
1172#
1173# sslflags= Various flags modifying the use of SSL:
1174# DELAYED_AUTH
1175# Don't request client certificates
1176# immediately, but wait until acl processing
1177# requires a certificate (not yet implemented).
1178# NO_DEFAULT_CA
1179# Don't use the default CA lists built in
1180# to OpenSSL.
1181# NO_SESSION_REUSE
1182# Don't allow for session reuse. Each connection
1183# will result in a new SSL session.
1184# VERIFY_CRL
1185# Verify CRL lists when accepting client
1186# certificates.
1187# VERIFY_CRL_ALL
1188# Verify CRL lists for all certificates in the
1189# client certificate chain.
1190#
1191# sslcontext= SSL session ID context identifier.
1192#
1193#
1194#Default:
1195# none
1196
1197# TAG: tcp_outgoing_tos
1198# Allows you to select a TOS/Diffserv value to mark outgoing
1199# connections with, based on the username or source address
1200# making the request.
1201#
1202# tcp_outgoing_tos ds-field [!]aclname ...
1203#
1204# Example where normal_service_net uses the TOS value 0x00
1205# and good_service_net uses 0x20
1206#
1207# acl normal_service_net src 10.0.0.0/255.255.255.0
1208# acl good_service_net src 10.0.1.0/255.255.255.0
1209# tcp_outgoing_tos 0x00 normal_service_net
1210# tcp_outgoing_tos 0x20 good_service_net
1211#
1212# TOS/DSCP values really only have local significance - so you should
1213# know what you're specifying. For more information, see RFC2474 and
1214# RFC3260.
1215#
1216# The TOS/DSCP byte must be exactly that - a octet value 0 - 255, or
1217# "default" to use whatever default your host has. Note that in
1218# practice often only values 0 - 63 is usable as the two highest bits
1219# have been redefined for use by ECN (RFC3168).
1220#
1221# Processing proceeds in the order specified, and stops at first fully
1222# matching line.
1223#
1224# Note: The use of this directive using client dependent ACLs is
1225# incompatible with the use of server side persistent connections. To
1226# ensure correct results it is best to set server_persisten_connections
1227# to off when using this directive in such configurations.
1228#
1229#Default:
1230# none
1231
1232# TAG: tcp_outgoing_address
1233# Allows you to map requests to different outgoing IP addresses
1234# based on the username or source address of the user making
1235# the request.
1236#
1237# tcp_outgoing_address ipaddr [[!]aclname] ...
1238#
1239# Example where requests from 10.0.0.0/24 will be forwarded
1240# with source address 10.1.0.1, 10.0.2.0/24 forwarded with
1241# source address 10.1.0.2 and the rest will be forwarded with
1242# source address 10.1.0.3.
1243#
1244# acl normal_service_net src 10.0.0.0/24
1245# acl good_service_net src 10.0.1.0/24 10.0.2.0/24
1246# tcp_outgoing_address 10.1.0.1 normal_service_net
1247# tcp_outgoing_address 10.1.0.2 good_service_net
1248# tcp_outgoing_address 10.1.0.3
1249#
1250# Processing proceeds in the order specified, and stops at first fully
1251# matching line.
1252#
1253# Note: The use of this directive using client dependent ACLs is
1254# incompatible with the use of server side persistent connections. To
1255# ensure correct results it is best to set server_persistent_connections
1256# to off when using this directive in such configurations.
1257#
1258#Default:
1259# none
1260
1261# TAG: zph_mode
1262# This option enables packet level marking of HIT/MISS responses,
1263# either using IP TOS or socket priority.
1264# off Feature disabled
1265# tos Set the IP TOS/Diffserv field
1266# priority Set the socket priority (may get mapped to TOS by OS,
1267# otherwise only usable in local rulesets)
1268# option Embed the mark in an IP option field. See also
1269# zph_option.
1270#
1271# See also tcp_outgoing_tos for details/requirements about TOS usage.
1272#
1273#Default:
1274# zph_mode off
1275
1276# TAG: zph_local
1277# Allows you to select a TOS/Diffserv/Priority value to mark local hits.
1278# Default: 0 (disabled).
1279#
1280#Default:
1281# zph_local 0
1282
1283# TAG: zph_sibling
1284# Allows you to select a TOS/Diffserv/Priority value to mark sibling hits.
1285# Default: 0 (disabled).
1286#
1287#Default:
1288# zph_sibling 0
1289
1290# TAG: zph_parent
1291# Allows you to select a TOS/Diffserv/Priority value to mark parent hits.
1292# Default: 0 (disabled).
1293#
1294#Default:
1295# zph_parent 0
1296
1297# TAG: zph_option
1298# The IP option to use when zph_mode is set to "option". Defaults to
1299# 136 which is officially registered as "SATNET Stream ID".
1300#
1301#Default:
1302# zph_option 136
1303
1304
1305# OPTIONS WHICH AFFECT THE NEIGHBOR SELECTION ALGORITHM
1306# -----------------------------------------------------------------------------
1307
1308# TAG: cache_peer
1309# To specify other caches in a hierarchy, use the format:
1310#
1311# cache_peer hostname type http-port icp-port [options]
1312#
1313# For example,
1314#
1315# # proxy icp
1316# # hostname type port port options
1317# # -------------------- -------- ----- ----- -----------
1318# cache_peer parent.foo.net parent 3128 3130 proxy-only default
1319# cache_peer sib1.foo.net sibling 3128 3130 proxy-only
1320# cache_peer sib2.foo.net sibling 3128 3130 proxy-only
1321#
1322# type: either 'parent', 'sibling', or 'multicast'.
1323#
1324# proxy-port: The port number where the cache listens for proxy
1325# requests.
1326#
1327# icp-port: Used for querying neighbor caches about
1328# objects. To have a non-ICP neighbor
1329# specify '7' for the ICP port and make sure the
1330# neighbor machine has the UDP echo port
1331# enabled in its /etc/inetd.conf file.
1332# NOTE: Also requires icp_port option enabled to send/receive
1333# requests via this method.
1334#
1335# options: proxy-only
1336# weight=n
1337# ttl=n
1338# no-query
1339# default
1340# round-robin
1341# carp
1342# multicast-responder
1343# multicast-siblings
1344# closest-only
1345# no-digest
1346# no-netdb-exchange
1347# no-delay
1348# login=user:password | PASS | *:password
1349# connect-timeout=nn
1350# digest-url=url
1351# allow-miss
1352# max-conn=n
1353# htcp
1354# htcp-oldsquid
1355# originserver
1356# userhash
1357# sourcehash
1358# name=xxx
1359# monitorurl=url
1360# monitorsize=sizespec
1361# monitorinterval=seconds
1362# monitortimeout=seconds
1363# forceddomain=name
1364# ssl
1365# sslcert=/path/to/ssl/certificate
1366# sslkey=/path/to/ssl/key
1367# sslversion=1|2|3|4
1368# sslcipher=...
1369# ssloptions=...
1370# front-end-https[=on|auto]
1371# connection-auth[=on|off|auto]
1372# idle=n
1373# http11
1374#
1375# use 'proxy-only' to specify objects fetched
1376# from this cache should not be saved locally.
1377#
1378# use 'weight=n' to affect the selection of a peer
1379# during any weighted peer-selection mechanisms.
1380# The weight must be an integer; default is 1,
1381# larger weights are favored more.
1382# This option does not affect parent selection if a peering
1383# protocol is not in use.
1384#
1385# use 'ttl=n' to specify a IP multicast TTL to use
1386# when sending an ICP queries to this address.
1387# Only useful when sending to a multicast group.
1388# Because we don't accept ICP replies from random
1389# hosts, you must configure other group members as
1390# peers with the 'multicast-responder' option below.
1391#
1392# use 'no-query' to NOT send ICP queries to this
1393# neighbor.
1394#
1395# use 'default' if this is a parent cache which can
1396# be used as a "last-resort" if a peer cannot be located
1397# by any of the peer-selection mechanisms.
1398# If specified more than once, only the first is used.
1399#
1400# use 'round-robin' to define a set of parents which
1401# should be used in a round-robin fashion in the
1402# absence of any ICP queries.
1403#
1404# use 'carp' to define a set of parents which should
1405# be used as a CARP array. The requests will be
1406# distributed among the parents based on the CARP load
1407# balancing hash function based on their weight.
1408#
1409# 'multicast-responder' indicates the named peer
1410# is a member of a multicast group. ICP queries will
1411# not be sent directly to the peer, but ICP replies
1412# will be accepted from it.
1413#
1414# the 'multicast-siblings' option is meant to be used
1415# only for cache peers of type "multicast". It instructs
1416# Squid that ALL members of this multicast group have
1417# "sibling" relationship with it, not "parent". This is
1418# an optimization that avoids useless multicast queries
1419# to a multicast group when the requested object would
1420# be fetched only from a "parent" cache, anyway. It's
1421# useful, e.g., when configuring a pool of redundant
1422# Squid proxies, being members of the same
1423# multicast group.
1424#
1425# 'closest-only' indicates that, for ICP_OP_MISS
1426# replies, we'll only forward CLOSEST_PARENT_MISSes
1427# and never FIRST_PARENT_MISSes.
1428#
1429# use 'no-digest' to NOT request cache digests from
1430# this neighbor.
1431#
1432# 'no-netdb-exchange' disables requesting ICMP
1433# RTT database (NetDB) from the neighbor.
1434#
1435# use 'no-delay' to prevent access to this neighbor
1436# from influencing the delay pools.
1437#
1438# use 'login=user:password' if this is a personal/workgroup
1439# proxy and your parent requires proxy authentication.
1440# Note: The string can include URL escapes (i.e. %20 for
1441# spaces). This also means % must be written as %%.
1442#
1443# use 'login=PASS' if users must authenticate against
1444# the upstream proxy or in the case of a reverse proxy
1445# configuration, the origin web server. This will pass
1446# the users credentials as they are to the peer.
1447# Note: To combine this with local authentication the Basic
1448# authentication scheme must be used, and both servers must
1449# share the same user database as HTTP only allows for
1450# a single login (one for proxy, one for origin server).
1451# Also be warned this will expose your users proxy
1452# password to the peer. USE WITH CAUTION
1453#
1454# use 'login=*:password' to pass the username to the
1455# upstream cache, but with a fixed password. This is meant
1456# to be used when the peer is in another administrative
1457# domain, but it is still needed to identify each user.
1458# The star can optionally be followed by some extra
1459# information which is added to the username. This can
1460# be used to identify this proxy to the peer, similar to
1461# the login=username:password option above.
1462#
1463# use 'connect-timeout=nn' to specify a peer
1464# specific connect timeout (also see the
1465# peer_connect_timeout directive)
1466#
1467# use 'digest-url=url' to tell Squid to fetch the cache
1468# digest (if digests are enabled) for this host from
1469# the specified URL rather than the Squid default
1470# location.
1471#
1472# use 'allow-miss' to disable Squid's use of only-if-cached
1473# when forwarding requests to siblings. This is primarily
1474# useful when icp_hit_stale is used by the sibling. To
1475# extensive use of this option may result in forwarding
1476# loops, and you should avoid having two-way peerings
1477# with this option. (for example to deny peer usage on
1478# requests from peer by denying cache_peer_access if the
1479# source is a peer)
1480#
1481# use 'max-conn=n' to limit the amount of connections Squid
1482# may open to this peer.
1483#
1484# use 'htcp' to send HTCP, instead of ICP, queries
1485# to the neighbor. You probably also want to
1486# set the "icp port" to 4827 instead of 3130.
1487# You must also allow this Squid htcp_access and
1488# http_access in the peer Squid configuration.
1489#
1490# use 'htcp-oldsquid' to send HTCP to old Squid versions
1491# You must also allow this Squid htcp_access and
1492# http_access in the peer Squid configuration.
1493#
1494# 'originserver' causes this parent peer to be contacted as
1495# a origin server. Meant to be used in accelerator setups.
1496#
1497# use 'userhash' to load-balance amongst a set of parents
1498# based on the client proxy_auth or ident username.
1499#
1500# use 'sourcehash' to load-balance amongst a set of parents
1501# based on the client source ip.
1502#
1503# use 'name=xxx' if you have multiple peers on the same
1504# host but different ports. This name can be used to
1505# differentiate the peers in cache_peer_access and similar
1506# directives.
1507#
1508# use 'monitorurl=url' to have periodically request a given
1509# URL from the peer, and only consider the peer as alive
1510# if this monitoring is successful (default none)
1511#
1512# use 'monitorsize=min[-max]' to limit the size range of
1513# 'monitorurl' replies considered valid. Defaults to 0 to
1514# accept any size replies as valid.
1515#
1516# use 'monitorinterval=seconds' to change frequency of
1517# how often the peer is monitored with 'monitorurl'
1518# (default 300 for a 5 minute interval). If set to 0
1519# then monitoring is disabled even if a URL is defined.
1520#
1521# use 'monitortimeout=seconds' to change the timeout of
1522# 'monitorurl'. Defaults to 'monitorinterval'.
1523#
1524# use 'forceddomain=name' to forcibly set the Host header
1525# of requests forwarded to this peer. Useful in accelerator
1526# setups where the server (peer) expects a certain domain
1527# name and using redirectors to feed this domain name
1528# is not feasible.
1529#
1530# use 'ssl' to indicate connections to this peer should
1531# be SSL/TLS encrypted.
1532#
1533# use 'sslcert=/path/to/ssl/certificate' to specify a client
1534# SSL certificate to use when connecting to this peer.
1535#
1536# use 'sslkey=/path/to/ssl/key' to specify the private SSL
1537# key corresponding to sslcert above. If 'sslkey' is not
1538# specified 'sslcert' is assumed to reference a
1539# combined file containing both the certificate and the key.
1540#
1541# use sslversion=1|2|3|4 to specify the SSL version to use
1542# when connecting to this peer
1543# 1 = automatic (default)
1544# 2 = SSL v2 only
1545# 3 = SSL v3 only
1546# 4 = TLS v1 only
1547#
1548# use sslcipher=... to specify the list of valid SSL ciphers
1549# to use when connecting to this peer.
1550#
1551# use ssloptions=... to specify various SSL engine options:
1552# NO_SSLv2 Disallow the use of SSLv2
1553# NO_SSLv3 Disallow the use of SSLv3
1554# NO_TLSv1 Disallow the use of TLSv1
1555# See src/ssl_support.c or the OpenSSL documentation for
1556# a more complete list.
1557#
1558# use sslcafile=... to specify a file containing
1559# additional CA certificates to use when verifying the
1560# peer certificate.
1561#
1562# use sslcapath=... to specify a directory containing
1563# additional CA certificates to use when verifying the
1564# peer certificate.
1565#
1566# use sslcrlfile=... to specify a certificate revocation
1567# list file to use when verifying the peer certificate.
1568#
1569# use sslflags=... to specify various flags modifying the
1570# SSL implementation:
1571# DONT_VERIFY_PEER
1572# Accept certificates even if they fail to
1573# verify.
1574# NO_DEFAULT_CA
1575# Don't use the default CA list built in
1576# to OpenSSL.
1577#
1578# use ssldomain= to specify the peer name as advertised
1579# in it's certificate. Used for verifying the correctness
1580# of the received peer certificate. If not specified the
1581# peer hostname will be used.
1582#
1583# use front-end-https to enable the "Front-End-Https: On"
1584# header needed when using Squid as a SSL frontend in front
1585# of Microsoft OWA. See MS KB document Q307347 for details
1586# on this header. If set to auto the header will
1587# only be added if the request is forwarded as a https://
1588# URL.
1589#
1590# use connection-auth=off to tell Squid that this peer does
1591# not support Microsoft connection oriented authentication,
1592# and any such challenges received from there should be
1593# ignored. Default is auto to automatically determine the
1594# status of the peer.
1595#
1596# use idle=n to specify a minimum number of idle connections
1597# that should be kept open to this peer.
1598#
1599# use http11 to send requests using HTTP/1.1 to this peer.
1600# Note: The HTTP/1.1 support is still incomplete, with an
1601# internal HTTP/1.0 hop. As result 1xx responses will not
1602# be forwarded.
1603#
1604#Default:
1605# none
1606
1607# TAG: cache_peer_domain
1608# Use to limit the domains for which a neighbor cache will be
1609# queried. Usage:
1610#
1611# cache_peer_domain cache-host domain [domain ...]
1612# cache_peer_domain cache-host !domain
1613#
1614# For example, specifying
1615#
1616# cache_peer_domain parent.foo.net .edu
1617#
1618# has the effect such that UDP query packets are sent to
1619# 'bigserver' only when the requested object exists on a
1620# server in the .edu domain. Prefixing the domain name
1621# with '!' means the cache will be queried for objects
1622# NOT in that domain.
1623#
1624# NOTE: * Any number of domains may be given for a cache-host,
1625# either on the same or separate lines.
1626# * When multiple domains are given for a particular
1627# cache-host, the first matched domain is applied.
1628# * Cache hosts with no domain restrictions are queried
1629# for all requests.
1630# * There are no defaults.
1631# * There is also a 'cache_peer_access' tag in the ACL
1632# section.
1633#
1634#Default:
1635# none
1636
1637# TAG: cache_peer_access
1638# Similar to 'cache_peer_domain' but provides more flexibility by
1639# using ACL elements.
1640#
1641# cache_peer_access cache-host allow|deny [!]aclname ...
1642#
1643# The syntax is identical to 'http_access' and the other lists of
1644# ACL elements. See the comments for 'http_access' below, or
1645# the Squid FAQ (http://www.squid-cache.org/FAQ/FAQ-10.html).
1646#
1647#Default:
1648# none
1649
1650# TAG: neighbor_type_domain
1651# usage: neighbor_type_domain neighbor parent|sibling domain domain ...
1652#
1653# Modifying the neighbor type for specific domains is now
1654# possible. You can treat some domains differently than the the
1655# default neighbor type specified on the 'cache_peer' line.
1656# Normally it should only be necessary to list domains which
1657# should be treated differently because the default neighbor type
1658# applies for hostnames which do not match domains listed here.
1659#
1660#EXAMPLE:
1661# cache_peer cache.foo.org parent 3128 3130
1662# neighbor_type_domain cache.foo.org sibling .com .net
1663# neighbor_type_domain cache.foo.org sibling .au .de
1664#
1665#Default:
1666# none
1667
1668# TAG: dead_peer_timeout (seconds)
1669# This controls how long Squid waits to declare a peer cache
1670# as "dead." If there are no ICP replies received in this
1671# amount of time, Squid will declare the peer dead and not
1672# expect to receive any further ICP replies. However, it
1673# continues to send ICP queries, and will mark the peer as
1674# alive upon receipt of the first subsequent ICP reply.
1675#
1676# This timeout also affects when Squid expects to receive ICP
1677# replies from peers. If more than 'dead_peer' seconds have
1678# passed since the last ICP reply was received, Squid will not
1679# expect to receive an ICP reply on the next query. Thus, if
1680# your time between requests is greater than this timeout, you
1681# will see a lot of requests sent DIRECT to origin servers
1682# instead of to your parents.
1683#
1684#Default:
1685# dead_peer_timeout 10 seconds
1686
1687# TAG: hierarchy_stoplist
1688# A list of words which, if found in a URL, cause the object to
1689# be handled directly by this cache. In other words, use this
1690# to not query neighbor caches for certain objects. You may
1691# list this option multiple times. Note: never_direct overrides
1692# this option.
1693#We recommend you to use at least the following line.
1694hierarchy_stoplist cgi-bin ?
1695
1696
1697# MEMORY CACHE OPTIONS
1698# -----------------------------------------------------------------------------
1699
1700# TAG: cache_mem (bytes)
1701# NOTE: THIS PARAMETER DOES NOT SPECIFY THE MAXIMUM PROCESS SIZE.
1702# IT ONLY PLACES A LIMIT ON HOW MUCH ADDITIONAL MEMORY SQUID WILL
1703# USE AS A MEMORY CACHE OF OBJECTS. SQUID USES MEMORY FOR OTHER
1704# THINGS AS WELL. SEE THE SQUID FAQ SECTION 8 FOR DETAILS.
1705#
1706# 'cache_mem' specifies the ideal amount of memory to be used
1707# for:
1708# * In-Transit objects
1709# * Hot Objects
1710# * Negative-Cached objects
1711#
1712# Data for these objects are stored in 4 KB blocks. This
1713# parameter specifies the ideal upper limit on the total size of
1714# 4 KB blocks allocated. In-Transit objects take the highest
1715# priority.
1716#
1717# In-transit objects have priority over the others. When
1718# additional space is needed for incoming data, negative-cached
1719# and hot objects will be released. In other words, the
1720# negative-cached and hot objects will fill up any unused space
1721# not needed for in-transit objects.
1722#
1723# If circumstances require, this limit will be exceeded.
1724# Specifically, if your incoming request rate requires more than
1725# 'cache_mem' of memory to hold in-transit objects, Squid will
1726# exceed this limit to satisfy the new requests. When the load
1727# decreases, blocks will be freed until the high-water mark is
1728# reached. Thereafter, blocks will be used to store hot
1729# objects.
1730#
1731#Default:
1732# cache_mem 16 MB
1733
1734# TAG: maximum_object_size_in_memory (bytes)
1735# Objects greater than this size will not be attempted to kept in
1736# the memory cache. This should be set high enough to keep objects
1737# accessed frequently in memory to improve performance whilst low
1738# enough to keep larger objects from hoarding cache_mem.
1739#
1740#Default:
1741# maximum_object_size_in_memory 8 KB
1742
1743# TAG: memory_replacement_policy
1744# The memory replacement policy parameter determines which
1745# objects are purged from memory when memory space is needed.
1746#
1747# See cache_replacement_policy for details.
1748#
1749#Default:
1750# memory_replacement_policy lru
1751
1752
1753# DISK CACHE OPTIONS
1754# -----------------------------------------------------------------------------
1755
1756# TAG: cache_replacement_policy
1757# The cache replacement policy parameter determines which
1758# objects are evicted (replaced) when disk space is needed.
1759#
1760# lru : Squid's original list based LRU policy
1761# heap GDSF : Greedy-Dual Size Frequency
1762# heap LFUDA: Least Frequently Used with Dynamic Aging
1763# heap LRU : LRU policy implemented using a heap
1764#
1765# Applies to any cache_dir lines listed below this.
1766#
1767# The LRU policies keeps recently referenced objects.
1768#
1769# The heap GDSF policy optimizes object hit rate by keeping smaller
1770# popular objects in cache so it has a better chance of getting a
1771# hit. It achieves a lower byte hit rate than LFUDA though since
1772# it evicts larger (possibly popular) objects.
1773#
1774# The heap LFUDA policy keeps popular objects in cache regardless of
1775# their size and thus optimizes byte hit rate at the expense of
1776# hit rate since one large, popular object will prevent many
1777# smaller, slightly less popular objects from being cached.
1778#
1779# Both policies utilize a dynamic aging mechanism that prevents
1780# cache pollution that can otherwise occur with frequency-based
1781# replacement policies.
1782#
1783# NOTE: if using the LFUDA replacement policy you should increase
1784# the value of maximum_object_size above its default of 4096 KB to
1785# to maximize the potential byte hit rate improvement of LFUDA.
1786#
1787# For more information about the GDSF and LFUDA cache replacement
1788# policies see http://www.hpl.hp.com/techreports/1999/HPL-1999-69.html
1789# and http://fog.hpl.external.hp.com/techreports/98/HPL-98-173.html.
1790#
1791#Default:
1792# cache_replacement_policy lru
1793
1794# TAG: cache_dir
1795# Usage:
1796#
1797# cache_dir Type Directory-Name Fs-specific-data [options]
1798#
1799# You can specify multiple cache_dir lines to spread the
1800# cache among different disk partitions.
1801#
1802# Type specifies the kind of storage system to use. Only "ufs"
1803# is built by default. To enable any of the other storage systems
1804# see the --enable-storeio configure option.
1805#
1806# 'Directory' is a top-level directory where cache swap
1807# files will be stored. If you want to use an entire disk
1808# for caching, this can be the mount-point directory.
1809# The directory must exist and be writable by the Squid
1810# process. Squid will NOT create this directory for you.
1811# Only using COSS, a raw disk device or a stripe file can
1812# be specified, but the configuration of the "cache_swap_log"
1813# tag is mandatory.
1814#
1815# The ufs store type:
1816#
1817# "ufs" is the old well-known Squid storage format that has always
1818# been there.
1819#
1820# cache_dir ufs Directory-Name Mbytes L1 L2 [options]
1821#
1822# 'Mbytes' is the amount of disk space (MB) to use under this
1823# directory. The default is 100 MB. Change this to suit your
1824# configuration. Do NOT put the size of your disk drive here.
1825# Instead, if you want Squid to use the entire disk drive,
1826# subtract 20% and use that value.
1827#
1828# 'Level-1' is the number of first-level subdirectories which
1829# will be created under the 'Directory'. The default is 16.
1830#
1831# 'Level-2' is the number of second-level subdirectories which
1832# will be created under each first-level directory. The default
1833# is 256.
1834#
1835# The aufs store type:
1836#
1837# "aufs" uses the same storage format as "ufs", utilizing
1838# POSIX-threads to avoid blocking the main Squid process on
1839# disk-I/O. This was formerly known in Squid as async-io.
1840#
1841# cache_dir aufs Directory-Name Mbytes L1 L2 [options]
1842#
1843# see argument descriptions under ufs above
1844#
1845# The diskd store type:
1846#
1847# "diskd" uses the same storage format as "ufs", utilizing a
1848# separate process to avoid blocking the main Squid process on
1849# disk-I/O.
1850#
1851# cache_dir diskd Directory-Name Mbytes L1 L2 [options] [Q1=n] [Q2=n]
1852#
1853# see argument descriptions under ufs above
1854#
1855# Q1 specifies the number of unacknowledged I/O requests when Squid
1856# stops opening new files. If this many messages are in the queues,
1857# Squid won't open new files. Default is 64
1858#
1859# Q2 specifies the number of unacknowledged messages when Squid
1860# starts blocking. If this many messages are in the queues,
1861# Squid blocks until it receives some replies. Default is 72
1862#
1863# When Q1 < Q2 (the default), the cache directory is optimized
1864# for lower response time at the expense of a decrease in hit
1865# ratio. If Q1 > Q2, the cache directory is optimized for
1866# higher hit ratio at the expense of an increase in response
1867# time.
1868#
1869# The coss store type:
1870#
1871# block-size=n defines the "block size" for COSS cache_dir's.
1872# Squid uses file numbers as block numbers. Since file numbers
1873# are limited to 24 bits, the block size determines the maximum
1874# size of the COSS partition. The default is 512 bytes, which
1875# leads to a maximum cache_dir size of 512<<24, or 8 GB. Note
1876# you should not change the COSS block size after Squid
1877# has written some objects to the cache_dir.
1878#
1879# overwrite-percent=n defines the percentage of disk that COSS
1880# must write to before a given object will be moved to the
1881# current stripe. A value of "n" closer to 100 will cause COSS
1882# to waste less disk space by having multiple copies of an object
1883# on disk, but will increase the chances of overwriting a popular
1884# object as COSS overwrites stripes. A value of "n" close to 0
1885# will cause COSS to keep all current objects in the current COSS
1886# stripe at the expense of the hit rate. The default value of 50
1887# will allow any given object to be stored on disk a maximum of
1888# 2 times.
1889#
1890# max-stripe-waste=n defines the maximum amount of space that COSS
1891# will waste in a given stripe (in bytes). When COSS writes data
1892# to disk, it will potentially waste up to "max-size" worth of disk
1893# space for each 1MB of data written. If "max-size" is set to a
1894# large value (ie >256k), this could potentially result in large
1895# amounts of wasted disk space. Setting this value to a lower value
1896# (ie 64k or 32k) will result in a COSS disk refusing to cache
1897# larger objects until the COSS stripe has been filled to within
1898# "max-stripe-waste" of the maximum size (1MB).
1899#
1900# membufs=n defines the number of "memory-only" stripes that COSS
1901# will use. When an cache hit is performed on a COSS stripe before
1902# COSS has reached the overwrite-percent value for that object,
1903# COSS will use a series of memory buffers to hold the object in
1904# while the data is sent to the client. This will define the maximum
1905# number of memory-only buffers that COSS will use. The default value
1906# is 10, which will use a maximum of 10MB of memory for buffers.
1907#
1908# maxfullbufs=n defines the maximum number of stripes a COSS partition
1909# will have in memory waiting to be freed (either because the disk is
1910# under load and the stripe is unwritten, or because clients are still
1911# transferring data from objects using the memory). In order to try
1912# and maintain a good hit rate under load, COSS will reserve the last
1913# 2 full stripes for object hits. (ie a COSS cache_dir will reject
1914# new objects when the number of full stripes is 2 less than maxfullbufs)
1915#
1916# The null store type:
1917#
1918# no options are allowed or required
1919#
1920# Common options:
1921#
1922# no-store, no new objects should be stored to this cache_dir
1923#
1924# min-size=n, refers to the min object size this storedir will accept.
1925# It's used to restrict a storedir to only store large objects
1926# (e.g. aufs) while other storedirs are optimized for smaller objects
1927# (e.g. COSS). Defaults to 0.
1928#
1929# max-size=n, refers to the max object size this storedir supports.
1930# It is used to initially choose the storedir to dump the object.
1931# Note: To make optimal use of the max-size limits you should order
1932# the cache_dir lines with the smallest max-size value first and the
1933# ones with no max-size specification last.
1934#
1935# Note that for coss, max-size must be less than COSS_MEMBUF_SZ
1936# (hard coded at 1 MB).
1937#
1938#Default:
1939 cache_dir ufs c:/squid/var/cache 100 16 256
1940
1941# TAG: store_dir_select_algorithm
1942# Set this to 'round-robin' as an alternative.
1943#
1944#Default:
1945# store_dir_select_algorithm least-load
1946
1947# TAG: max_open_disk_fds
1948# To avoid having disk as the I/O bottleneck Squid can optionally
1949# bypass the on-disk cache if more than this amount of disk file
1950# descriptors are open.
1951#
1952# A value of 0 indicates no limit.
1953#
1954#Default:
1955# max_open_disk_fds 0
1956
1957# TAG: minimum_object_size (bytes)
1958# Objects smaller than this size will NOT be saved on disk. The
1959# value is specified in kilobytes, and the default is 0 KB, which
1960# means there is no minimum.
1961#
1962#Default:
1963# minimum_object_size 0 KB
1964
1965# TAG: maximum_object_size (bytes)
1966# Objects larger than this size will NOT be saved on disk. The
1967# value is specified in kilobytes, and the default is 4MB. If
1968# you wish to get a high BYTES hit ratio, you should probably
1969# increase this (one 32 MB object hit counts for 3200 10KB
1970# hits). If you wish to increase speed more than your want to
1971# save bandwidth you should leave this low.
1972#
1973# NOTE: if using the LFUDA replacement policy you should increase
1974# this value to maximize the byte hit rate improvement of LFUDA!
1975# See replacement_policy below for a discussion of this policy.
1976#
1977#Default:
1978# maximum_object_size 4096 KB
1979
1980# TAG: cache_swap_low (percent, 0-100)
1981# TAG: cache_swap_high (percent, 0-100)
1982#
1983# The low- and high-water marks for cache object replacement.
1984# Replacement begins when the swap (disk) usage is above the
1985# low-water mark and attempts to maintain utilization near the
1986# low-water mark. As swap utilization gets close to high-water
1987# mark object eviction becomes more aggressive. If utilization is
1988# close to the low-water mark less replacement is done each time.
1989#
1990# Defaults are 90% and 95%. If you have a large cache, 5% could be
1991# hundreds of MB. If this is the case you may wish to set these
1992# numbers closer together.
1993#
1994#Default:
1995# cache_swap_low 90
1996# cache_swap_high 95
1997
1998# TAG: update_headers on|off
1999# By default Squid updates stored HTTP headers when receiving
2000# a 304 response. Set this to off if you want to disable this
2001# for disk I/O performance reasons. Disabling this VIOLATES the
2002# HTTP standard, and could make you liable for problems which it
2003# causes.
2004#
2005#Default:
2006# update_headers on
2007
2008
2009# LOGFILE OPTIONS
2010# -----------------------------------------------------------------------------
2011
2012# TAG: logformat
2013# Usage:
2014#
2015# logformat <name> <format specification>
2016#
2017# Defines an access log format.
2018#
2019# The <format specification> is a string with embedded % format codes
2020#
2021# % format codes all follow the same basic structure where all but
2022# the formatcode is optional. Output strings are automatically escaped
2023# as required according to their context and the output format
2024# modifiers are usually not needed, but can be specified if an explicit
2025# output format is desired.
2026#
2027# % ["|[|'|#] [-] [[0]width] [{argument}] formatcode
2028#
2029# " output in quoted string format
2030# [ output in squid text log format as used by log_mime_hdrs
2031# # output in URL quoted format
2032# ' output as-is
2033#
2034# - left aligned
2035# width field width. If starting with 0 the
2036# output is zero padded
2037# {arg} argument such as header name etc
2038#
2039# Format codes:
2040#
2041# >a Client source IP address
2042# >A Client FQDN
2043# >p Client source port
2044# <A Server IP address or peer name
2045# la Local IP address (http_port)
2046# lp Local port number (http_port)
2047# oa Our outgoing IP address (tcp_outgoing_address)
2048# ts Seconds since epoch
2049# tu subsecond time (milliseconds)
2050# tl Local time. Optional strftime format argument
2051# default %d/%b/%Y:%H:%M:%S %z
2052# tg GMT time. Optional strftime format argument
2053# default %d/%b/%Y:%H:%M:%S %z
2054# tr Response time (milliseconds)
2055# >h Request header. Optional header name argument
2056# on the format header[:[separator]element]
2057# <h Reply header. Optional header name argument
2058# as for >h
2059# un User name
2060# ul User name from authentication
2061# ui User name from ident
2062# us User name from SSL
2063# ue User name from external acl helper
2064# Hs HTTP status code
2065# Ss Squid request status (TCP_MISS etc)
2066# Sh Squid hierarchy status (DEFAULT_PARENT etc)
2067# mt MIME content type
2068# rm Request method (GET/POST etc)
2069# ru Request URL
2070# rp Request URL-Path excluding hostname
2071# rv Request protocol version
2072# ea Log string returned by external acl
2073# <st Reply size including HTTP headers
2074# >st Request size including HTTP headers
2075# st Request+Reply size including HTTP headers
2076# sn Unique sequence number per log line entry
2077# % a literal % character
2078#
2079# The default formats available (which do not need re-defining) are:
2080#
2081#logformat squid %ts.%03tu %6tr %>a %Ss/%03Hs %<st %rm %ru %un %Sh/%<A %mt
2082#logformat squidmime %ts.%03tu %6tr %>a %Ss/%03Hs %<st %rm %ru %un %Sh/%<A %mt [%>h] [%<h]
2083#logformat common %>a %ui %un [%tl] "%rm %ru HTTP/%rv" %Hs %<st %Ss:%Sh
2084#logformat combined %>a %ui %un [%tl] "%rm %ru HTTP/%rv" %Hs %<st "%{Referer}>h" "%{User-Agent}>h" %Ss:%Sh
2085#
2086#Default:
2087# none
2088
2089# TAG: access_log
2090# These files log client request activities. Has a line every HTTP or
2091# ICP request. The format is:
2092# access_log <filepath> [<logformat name> [acl acl ...]]
2093# access_log none [acl acl ...]]
2094#
2095# Will log to the specified file using the specified format (which
2096# must be defined in a logformat directive) those entries which match
2097# ALL the acl's specified (which must be defined in acl clauses).
2098# If no acl is specified, all requests will be logged to this file.
2099#
2100# To disable logging of a request use the filepath "none", in which case
2101# a logformat name should not be specified.
2102#
2103# To log the request via syslog specify a filepath of "syslog":
2104#
2105# access_log syslog[:facility.priority] [format [acl1 [acl2 ....]]]
2106# where facility could be any of:
2107# authpriv, daemon, local0 .. local7 or user.
2108#
2109# And priority could be any of:
2110# err, warning, notice, info, debug.
2111access_log c:/squid/var/logs/access.log squid
2112
2113# TAG: log_access allow|deny acl acl...
2114# This options allows you to control which requests gets logged
2115# to access.log (see access_log directive). Requests denied for
2116# logging will also not be accounted for in performance counters.
2117#
2118#Default:
2119# none
2120
2121# TAG: logfile_daemon
2122# Specify the path to the logfile-writing daemon. This daemon is
2123# used to write the access and store logs, if configured.
2124#
2125#Default:
2126# logfile_daemon c:/squid/libexec/logfile-daemon.exe
2127
2128# TAG: cache_log
2129# Cache logging file. This is where general information about
2130# your cache's behavior goes. You can increase the amount of data
2131# logged to this file with the "debug_options" tag below.
2132#
2133#Default:
2134# cache_log c:/squid/var/logs/cache.log
2135
2136# TAG: cache_store_log
2137# Logs the activities of the storage manager. Shows which
2138# objects are ejected from the cache, and which objects are
2139# saved and for how long. To disable, enter "none". There are
2140# not really utilities to analyze this data, so you can safely
2141# disable it.
2142#
2143#Default:
2144# cache_store_log c:/squid/var/logs/store.log
2145
2146# TAG: cache_swap_state
2147# Location for the cache "swap.state" file. This index file holds
2148# the metadata of objects saved on disk. It is used to rebuild
2149# the cache during startup. Normally this file resides in each
2150# 'cache_dir' directory, but you may specify an alternate
2151# pathname here. Note you must give a full filename, not just
2152# a directory. Since this is the index for the whole object
2153# list you CANNOT periodically rotate it!
2154#
2155# If %s can be used in the file name it will be replaced with a
2156# a representation of the cache_dir name where each / is replaced
2157# with '.'. This is needed to allow adding/removing cache_dir
2158# lines when cache_swap_log is being used.
2159#
2160# If have more than one 'cache_dir', and %s is not used in the name
2161# these swap logs will have names such as:
2162#
2163# cache_swap_log.00
2164# cache_swap_log.01
2165# cache_swap_log.02
2166#
2167# The numbered extension (which is added automatically)
2168# corresponds to the order of the 'cache_dir' lines in this
2169# configuration file. If you change the order of the 'cache_dir'
2170# lines in this file, these index files will NOT correspond to
2171# the correct 'cache_dir' entry (unless you manually rename
2172# them). We recommend you do NOT use this option. It is
2173# better to keep these index files in each 'cache_dir' directory.
2174#
2175#Default:
2176# none
2177
2178# TAG: logfile_rotate
2179# Specifies the number of logfile rotations to make when you
2180# type 'squid -k rotate'. The default is 10, which will rotate
2181# with extensions 0 through 9. Setting logfile_rotate to 0 will
2182# disable the file name rotation, but the logfiles are still closed
2183# and re-opened. This will enable you to rename the logfiles
2184# yourself just before sending the rotate signal.
2185#
2186# Note, the 'squid -k rotate' command normally sends a USR1
2187# signal to the running squid process. In certain situations
2188# (e.g. on Linux with Async I/O), USR1 is used for other
2189# purposes, so -k rotate uses another signal. It is best to get
2190# in the habit of using 'squid -k rotate' instead of 'kill -USR1
2191# <pid>'.
2192#
2193#Default:
2194# logfile_rotate 10
2195
2196# TAG: emulate_httpd_log on|off
2197# The Cache can emulate the log file format which many 'httpd'
2198# programs use. To disable/enable this emulation, set
2199# emulate_httpd_log to 'off' or 'on'. The default
2200# is to use the native log format since it includes useful
2201# information Squid-specific log analyzers use.
2202#
2203#Default:
2204# emulate_httpd_log off
2205
2206# TAG: log_ip_on_direct on|off
2207# Log the destination IP address in the hierarchy log tag when going
2208# direct. Earlier Squid versions logged the hostname here. If you
2209# prefer the old way set this to off.
2210#
2211#Default:
2212# log_ip_on_direct on
2213
2214# TAG: mime_table
2215# Pathname to Squid's MIME table. You shouldn't need to change
2216# this, but the default file contains examples and formatting
2217# information if you do.
2218#
2219#Default:
2220# mime_table c:/squid/etc/mime.conf
2221
2222# TAG: log_mime_hdrs on|off
2223# The Cache can record both the request and the response MIME
2224# headers for each HTTP transaction. The headers are encoded
2225# safely and will appear as two bracketed fields at the end of
2226# the access log (for either the native or httpd-emulated log
2227# formats). To enable this logging set log_mime_hdrs to 'on'.
2228#
2229#Default:
2230# log_mime_hdrs off
2231
2232# TAG: useragent_log
2233# Squid will write the User-Agent field from HTTP requests
2234# to the filename specified here. By default useragent_log
2235# is disabled.
2236#
2237#Default:
2238# none
2239
2240# TAG: referer_log
2241# Squid will write the Referer field from HTTP requests to the
2242# filename specified here. By default referer_log is disabled.
2243# Note that "referer" is actually a misspelling of "referrer"
2244# however the misspelt version has been accepted into the HTTP RFCs
2245# and we accept both.
2246#
2247#Default:
2248# none
2249
2250# TAG: pid_filename
2251# A filename to write the process-id to. To disable, enter "none".
2252#
2253#Default:
2254# pid_filename c:/squid/var/logs/squid.pid
2255
2256# TAG: debug_options
2257# Logging options are set as section,level where each source file
2258# is assigned a unique section. Lower levels result in less
2259# output, Full debugging (level 9) can result in a very large
2260# log file, so be careful. The magic word "ALL" sets debugging
2261# levels for all sections. We recommend normally running with
2262# "ALL,1".
2263#
2264#Default:
2265# debug_options ALL,1
2266
2267# TAG: log_fqdn on|off
2268# Turn this on if you wish to log fully qualified domain names
2269# in the access.log. To do this Squid does a DNS lookup of all
2270# IP's connecting to it. This can (in some situations) increase
2271# latency, which makes your cache seem slower for interactive
2272# browsing.
2273#
2274#Default:
2275# log_fqdn off
2276
2277# TAG: client_netmask
2278# A netmask for client addresses in logfiles and cachemgr output.
2279# Change this to protect the privacy of your cache clients.
2280# A netmask of 255.255.255.0 will log all IP's in that range with
2281# the last digit set to '0'.
2282#
2283#Default:
2284# client_netmask 255.255.255.255
2285
2286# TAG: forward_log
2287# Note: This option is only available if Squid is rebuilt with the
2288# --enable-forward-log option
2289#
2290# Logs the server-side requests.
2291#
2292# This is currently work in progress.
2293#
2294#Default:
2295# none
2296
2297# TAG: strip_query_terms
2298# By default, Squid strips query terms from requested URLs before
2299# logging. This protects your user's privacy.
2300#
2301#Default:
2302# strip_query_terms on
2303
2304# TAG: buffered_logs on|off
2305# cache.log log file is written with stdio functions, and as such
2306# it can be buffered or unbuffered. By default it will be unbuffered.
2307# Buffering it can speed up the writing slightly (though you are
2308# unlikely to need to worry unless you run with tons of debugging
2309# enabled in which case performance will suffer badly anyway..).
2310#
2311#Default:
2312# buffered_logs off
2313
2314# TAG: netdb_filename
2315# A filename where Squid stores it's netdb state between restarts.
2316# To disable, enter "none".
2317#
2318#Default:
2319# netdb_filename c:/squid/var/logs/netdb.state
2320
2321
2322# OPTIONS FOR FTP GATEWAYING
2323# -----------------------------------------------------------------------------
2324
2325# TAG: ftp_user
2326# If you want the anonymous login password to be more informative
2327# (and enable the use of picky ftp servers), set this to something
2328# reasonable for your domain, like wwwuser@somewhere.net
2329#
2330# The reason why this is domainless by default is the
2331# request can be made on the behalf of a user in any domain,
2332# depending on how the cache is used.
2333# Some ftp server also validate the email address is valid
2334# (for example perl.com).
2335#
2336#Default:
2337# ftp_user Squid@
2338
2339# TAG: ftp_list_width
2340# Sets the width of ftp listings. This should be set to fit in
2341# the width of a standard browser. Setting this too small
2342# can cut off long filenames when browsing ftp sites.
2343#
2344#Default:
2345# ftp_list_width 32
2346
2347# TAG: ftp_passive
2348# If your firewall does not allow Squid to use passive
2349# connections, turn off this option.
2350#
2351#Default:
2352# ftp_passive on
2353
2354# TAG: ftp_sanitycheck
2355# For security and data integrity reasons Squid by default performs
2356# sanity checks of the addresses of FTP data connections ensure the
2357# data connection is to the requested server. If you need to allow
2358# FTP connections to servers using another IP address for the data
2359# connection turn this off.
2360#
2361#Default:
2362# ftp_sanitycheck on
2363
2364# TAG: ftp_telnet_protocol
2365# The FTP protocol is officially defined to use the telnet protocol
2366# as transport channel for the control connection. However, many
2367# implementations are broken and does not respect this aspect of
2368# the FTP protocol.
2369#
2370# If you have trouble accessing files with ASCII code 255 in the
2371# path or similar problems involving this ASCII code you can
2372# try setting this directive to off. If that helps, report to the
2373# operator of the FTP server in question that their FTP server
2374# is broken and does not follow the FTP standard.
2375#
2376#Default:
2377# ftp_telnet_protocol on
2378
2379
2380# OPTIONS FOR EXTERNAL SUPPORT PROGRAMS
2381# -----------------------------------------------------------------------------
2382
2383# TAG: diskd_program
2384# Specify the location of the diskd executable.
2385# Note this is only useful if you have compiled in
2386# diskd as one of the store io modules.
2387#
2388#Default:
2389# diskd_program c:/squid/libexec/diskd-daemon.exe
2390
2391# TAG: unlinkd_program
2392# Specify the location of the executable for file deletion process.
2393#
2394#Default:
2395# unlinkd_program c:/squid/libexec/unlinkd.exe
2396
2397# TAG: pinger_program
2398# Note: This option is only available if Squid is rebuilt with the
2399# --enable-icmp option
2400#
2401# Specify the location of the executable for the pinger process.
2402#
2403#Default:
2404# pinger_program c:/squid/libexec/pinger.exe
2405
2406
2407# OPTIONS FOR URL REWRITING
2408# -----------------------------------------------------------------------------
2409
2410# TAG: storeurl_rewrite_program
2411# Specify the location of the executable for the Store URL rewriter.
2412# The Store URL rewriter allows URLs to be "normalised" ; mapping
2413# multiple URLs to a single URL representation for cache operations.
2414#
2415# For example, if you request an object at:
2416#
2417# http://srv1.example.com/image.gif
2418#
2419# and a subsequent request for:
2420#
2421# http://srv2.example.com/image.gif
2422#
2423# then Squid will treat these both as different URLs and cache them
2424# seperately.
2425#
2426# This is almost the normal case, but an increasing number of sites
2427# distribute the same content between multiple frontend hosts.
2428# The Store URL rewriter allows you to rewrite these URLs to one URL
2429# to use for cache operations, but not -fetches-. Fetches are still
2430# made from the original site, but stored with the store URL rewritten
2431# URL as the store key.
2432#
2433# For each requested URL rewriter will receive on line with the format
2434#
2435# URL <SP> client_ip "/" fqdn <SP> user <SP> method <SP> urlgroup
2436# [<SP> kvpairs] <NL>
2437#
2438# In the future, the rewriter interface will be extended with
2439# key=value pairs ("kvpairs" shown above). Rewriter programs
2440# should be prepared to receive and possibly ignore additional
2441# whitespace-separated tokens on each input line.
2442#
2443# And the rewriter may return a rewritten URL. The other components of
2444# the request line does not need to be returned (ignored if they are).
2445#
2446# By default, a Store URL rewriter is not used.
2447#
2448# Please note - the normal URL rewriter rewrites Squid's _destination_
2449# URL - ie, what it fetches. The Store URL rewriter rewrites Squid's
2450# _store_ URL - ie, what it uses to store and retrieve objects.
2451#
2452#Default:
2453# none
2454
2455# TAG: storeurl_rewrite_children
2456#
2457#
2458#Default:
2459# storeurl_rewrite_children 5
2460
2461# TAG: storeurl_rewrite_concurrency
2462#
2463#
2464#Default:
2465# storeurl_rewrite_concurrency 0
2466
2467# TAG: url_rewrite_program
2468# Specify the location of the executable for the URL rewriter.
2469# Since they can perform almost any function there isn't one included.
2470#
2471# For each requested URL rewriter will receive on line with the format
2472#
2473# URL <SP> client_ip "/" fqdn <SP> user <SP> method <SP> urlgroup
2474# [<SP> kvpairs] <NL>
2475#
2476# In the future, the rewriter interface will be extended with
2477# key=value pairs ("kvpairs" shown above). Rewriter programs
2478# should be prepared to receive and possibly ignore additional
2479# whitespace-separated tokens on each input line.
2480#
2481# And the rewriter may return a rewritten URL. The other components of
2482# the request line does not need to be returned (ignored if they are).
2483#
2484# The rewriter can also indicate that a client-side redirect should
2485# be performed to the new URL. This is done by prefixing the returned
2486# URL with "301:" (moved permanently) or 302: (moved temporarily).
2487#
2488# It can also return a "urlgroup" that can subsequently be matched
2489# in cache_peer_access and similar ACL driven rules. An urlgroup is
2490# returned by prefixing the returned URL with "!urlgroup!".
2491#
2492# By default, a URL rewriter is not used.
2493#
2494#Default:
2495# none
2496
2497# TAG: url_rewrite_children
2498# The number of redirector processes to spawn. If you start
2499# too few Squid will have to wait for them to process a backlog of
2500# URLs, slowing it down. If you start too many they will use RAM
2501# and other system resources.
2502#
2503#Default:
2504# url_rewrite_children 5
2505
2506# TAG: url_rewrite_concurrency
2507# The number of requests each redirector helper can handle in
2508# parallel. Defaults to 0 which indicates the redirector
2509# is a old-style single threaded redirector.
2510#
2511# When this directive is set to a value >= 1 then the protocol
2512# used to communicate with the helper is modified to include
2513# a request ID in front of the request/response. The request
2514# ID from the request must be echoed back with the response
2515# to that request.
2516#
2517#Default:
2518# url_rewrite_concurrency 0
2519
2520# TAG: url_rewrite_host_header
2521# By default Squid rewrites any Host: header in redirected
2522# requests. If you are running an accelerator this may
2523# not be a wanted effect of a redirector.
2524#
2525# WARNING: Entries are cached on the result of the URL rewriting
2526# process, so be careful if you have domain-virtual hosts.
2527#
2528#Default:
2529# url_rewrite_host_header on
2530
2531# TAG: url_rewrite_access
2532# If defined, this access list specifies which requests are
2533# sent to the redirector processes. By default all requests
2534# are sent.
2535#
2536#Default:
2537# none
2538
2539# TAG: storeurl_access
2540#
2541#
2542#Default:
2543# none
2544
2545# TAG: redirector_bypass
2546# When this is 'on', a request will not go through the
2547# redirector if all redirectors are busy. If this is 'off'
2548# and the redirector queue grows too large, Squid will exit
2549# with a FATAL error and ask you to increase the number of
2550# redirectors. You should only enable this if the redirectors
2551# are not critical to your caching system. If you use
2552# redirectors for access control, and you enable this option,
2553# users may have access to pages they should not
2554# be allowed to request.
2555#
2556#Default:
2557# redirector_bypass off
2558
2559# TAG: location_rewrite_program
2560# Specify the location of the executable for the Location rewriter,
2561# used to rewrite server generated redirects. Usually used in
2562# conjunction with a url_rewrite_program
2563#
2564# For each Location header received the location rewriter will receive
2565# one line with the format:
2566#
2567# location URL <SP> requested URL <SP> urlgroup <NL>
2568#
2569# And the rewriter may return a rewritten Location URL or a blank line.
2570# The other components of the request line does not need to be returned
2571# (ignored if they are).
2572#
2573# By default, a Location rewriter is not used.
2574#
2575#Default:
2576# none
2577
2578# TAG: location_rewrite_children
2579# The number of location rewriting processes to spawn. If you start
2580# too few Squid will have to wait for them to process a backlog of
2581# URLs, slowing it down. If you start too many they will use RAM
2582# and other system resources.
2583#
2584#Default:
2585# location_rewrite_children 5
2586
2587# TAG: location_rewrite_concurrency
2588# The number of requests each Location rewriter helper can handle in
2589# parallel. Defaults to 0 which indicates that the helper
2590# is a old-style singlethreaded helper.
2591#
2592#Default:
2593# location_rewrite_concurrency 0
2594
2595# TAG: location_rewrite_access
2596# If defined, this access list specifies which requests are
2597# sent to the location rewriting processes. By default all Location
2598# headers are sent.
2599#
2600#Default:
2601# none
2602
2603
2604# OPTIONS FOR TUNING THE CACHE
2605# -----------------------------------------------------------------------------
2606
2607# TAG: cache
2608# A list of ACL elements which, if matched, cause the request to
2609# not be satisfied from the cache and the reply to not be cached.
2610# In other words, use this to force certain objects to never be cached.
2611#
2612# You must use the word 'DENY' to indicate the ACL names which should
2613# NOT be cached.
2614#
2615# Default is to allow all to be cached.
2616#
2617#Default:
2618# none
2619
2620# TAG: max_stale time-units
2621# This option puts an upper limit on how stale content Squid
2622# will serve from the cache if cache validation fails.
2623#
2624#Default:
2625# max_stale 1 week
2626
2627# TAG: refresh_pattern
2628# usage: refresh_pattern [-i] regex min percent max [options]
2629#
2630# By default, regular expressions are CASE-SENSITIVE. To make
2631# them case-insensitive, use the -i option.
2632#
2633# 'Min' is the time (in minutes) an object without an explicit
2634# expiry time should be considered fresh. The recommended
2635# value is 0, any higher values may cause dynamic applications
2636# to be erroneously cached unless the application designer
2637# has taken the appropriate actions.
2638#
2639# 'Percent' is a percentage of the objects age (time since last
2640# modification age) an object without explicit expiry time
2641# will be considered fresh.
2642#
2643# 'Max' is an upper limit on how long objects without an explicit
2644# expiry time will be considered fresh.
2645#
2646# options: override-expire
2647# override-lastmod
2648# reload-into-ims
2649# ignore-reload
2650# ignore-no-cache
2651# ignore-private
2652# ignore-auth
2653# stale-while-revalidate=NN
2654# ignore-stale-while-revalidate
2655# max-stale=NN
2656# negative-ttl=NN
2657#
2658# override-expire enforces min age even if the server
2659# sent an explicit expiry time (e.g., with the
2660# Expires: header or Cache-Control: max-age). Doing this
2661# VIOLATES the HTTP standard. Enabling this feature
2662# could make you liable for problems which it causes.
2663#
2664# Note: this does not enforce staleness - it only extends
2665# freshness / min. If the server returns a Expires time which
2666# is longer than your max time, Squid will still consider
2667# the object fresh for that period of time.
2668#
2669# override-lastmod enforces min age even on objects
2670# that were modified recently.
2671#
2672# reload-into-ims changes client no-cache or ``reload''
2673# to If-Modified-Since requests. Doing this VIOLATES the
2674# HTTP standard. Enabling this feature could make you
2675# liable for problems which it causes.
2676#
2677# ignore-reload ignores a client no-cache or ``reload''
2678# header. Doing this VIOLATES the HTTP standard. Enabling
2679# this feature could make you liable for problems which
2680# it causes.
2681#
2682# ignore-no-cache ignores any ``Pragma: no-cache'' and
2683# ``Cache-control: no-cache'' headers received from a server.
2684# The HTTP RFC never allows the use of this (Pragma) header
2685# from a server, only a client, though plenty of servers
2686# send it anyway.
2687#
2688# ignore-private ignores any ``Cache-control: private''
2689# headers received from a server. Doing this VIOLATES
2690# the HTTP standard. Enabling this feature could make you
2691# liable for problems which it causes.
2692#
2693# ignore-auth caches responses to requests with authorization,
2694# as if the originserver had sent ``Cache-control: public''
2695# in the response header. Doing this VIOLATES the HTTP standard.
2696# Enabling this feature could make you liable for problems which
2697# it causes.
2698#
2699# stale-while-revalidate=NN makes Squid perform an asyncronous
2700# cache validation if the object isn't more stale than NN.
2701# Doing this VIOLATES the HTTP standard. Enabling this
2702# feature could make you liable for problems which it
2703# causes.
2704#
2705# ignore-stale-while-revalidate makes Squid ignore any 'Cache-Control:
2706# stale-while-revalidate=NN' headers received from a server. Can be
2707# combined with stale-while-revalidate=NN to override the server provided
2708# value.
2709#
2710# max-stale=NN provided a maximum staleness factor. Squid won't
2711# serve objects more stale than this even if it failed to
2712# validate the object.
2713#
2714# negative-ttl=NN overrides the global negative_ttl parameter
2715# selectively for URLs matching this pattern (in seconds).
2716#
2717# Basically a cached object is:
2718#
2719# FRESH if expires < now, else STALE
2720# STALE if age > max
2721# FRESH if lm-factor < percent, else STALE
2722# FRESH if age < min
2723# else STALE
2724#
2725# The refresh_pattern lines are checked in the order listed here.
2726# The first entry which matches is used. If none of the entries
2727# match the default will be used.
2728#
2729# Note, you must uncomment all the default lines if you want
2730# to change one. The default setting is only active if none is
2731# used.
2732#
2733#Suggested default:
2734refresh_pattern ^ftp: 1440 20% 10080
2735refresh_pattern ^gopher: 1440 0% 1440
2736refresh_pattern -i (/cgi-bin/|\?) 0 0% 0
2737refresh_pattern . 0 20% 4320
2738
2739# TAG: quick_abort_min (KB)
2740# TAG: quick_abort_max (KB)
2741# TAG: quick_abort_pct (percent)
2742# The cache by default continues downloading aborted requests
2743# which are almost completed (less than 16 KB remaining). This
2744# may be undesirable on slow (e.g. SLIP) links and/or very busy
2745# caches. Impatient users may tie up file descriptors and
2746# bandwidth by repeatedly requesting and immediately aborting
2747# downloads.
2748#
2749# When the user aborts a request, Squid will check the
2750# quick_abort values to the amount of data transfered until
2751# then.
2752#
2753# If the transfer has less than 'quick_abort_min' KB remaining,
2754# it will finish the retrieval.
2755#
2756# If the transfer has more than 'quick_abort_max' KB remaining,
2757# it will abort the retrieval.
2758#
2759# If more than 'quick_abort_pct' of the transfer has completed,
2760# it will finish the retrieval.
2761#
2762# If you do not want any retrieval to continue after the client
2763# has aborted, set both 'quick_abort_min' and 'quick_abort_max'
2764# to '0 KB'.
2765#
2766# If you want retrievals to always continue if they are being
2767# cached set 'quick_abort_min' to '-1 KB'.
2768#
2769#Default:
2770# quick_abort_min 16 KB
2771# quick_abort_max 16 KB
2772# quick_abort_pct 95
2773
2774# TAG: read_ahead_gap buffer-size
2775# The amount of data the cache will buffer ahead of what has been
2776# sent to the client when retrieving an object from another server.
2777#
2778#Default:
2779# read_ahead_gap 16 KB
2780
2781# TAG: negative_ttl time-units
2782# Time-to-Live (TTL) for failed requests. Certain types of
2783# failures (such as "connection refused" and "404 Not Found") are
2784# negatively-cached for a configurable amount of time. The
2785# default is 5 minutes. Note that this is different from
2786# negative caching of DNS lookups.
2787#
2788#Default:
2789# negative_ttl 5 minutes
2790
2791# TAG: positive_dns_ttl time-units
2792# Upper limit on how long Squid will cache positive DNS responses.
2793# Default is 6 hours (360 minutes). This directive must be set
2794# larger than negative_dns_ttl.
2795#
2796#Default:
2797# positive_dns_ttl 6 hours
2798
2799# TAG: negative_dns_ttl time-units
2800# Time-to-Live (TTL) for negative caching of failed DNS lookups.
2801# This also sets the lower cache limit on positive lookups.
2802# Minimum value is 1 second, and it is not recommendable to go
2803# much below 10 seconds.
2804#
2805#Default:
2806# negative_dns_ttl 1 minute
2807
2808# TAG: range_offset_limit (bytes)
2809# Sets a upper limit on how far into the the file a Range request
2810# may be to cause Squid to prefetch the whole file. If beyond this
2811# limit Squid forwards the Range request as it is and the result
2812# is NOT cached.
2813#
2814# This is to stop a far ahead range request (lets say start at 17MB)
2815# from making Squid fetch the whole object up to that point before
2816# sending anything to the client.
2817#
2818# A value of -1 causes Squid to always fetch the object from the
2819# beginning so it may cache the result. (2.0 style)
2820#
2821# A value of 0 causes Squid to never fetch more than the
2822# client requested. (default)
2823#
2824#Default:
2825# range_offset_limit 0 KB
2826
2827# TAG: minimum_expiry_time (seconds)
2828# The minimum caching time according to (Expires - Date)
2829# Headers Squid honors if the object can't be revalidated
2830# defaults to 60 seconds. In reverse proxy enorinments it
2831# might be desirable to honor shorter object lifetimes. It
2832# is most likely better to make your server return a
2833# meaningful Last-Modified header however.
2834#
2835#Default:
2836# minimum_expiry_time 60 seconds
2837
2838# TAG: store_avg_object_size (kbytes)
2839# Average object size, used to estimate number of objects your
2840# cache can hold. The default is 13 KB.
2841#
2842#Default:
2843# store_avg_object_size 13 KB
2844
2845# TAG: store_objects_per_bucket
2846# Target number of objects per bucket in the store hash table.
2847# Lowering this value increases the total number of buckets and
2848# also the storage maintenance rate. The default is 20.
2849#
2850#Default:
2851# store_objects_per_bucket 20
2852
2853
2854# HTTP OPTIONS
2855# -----------------------------------------------------------------------------
2856
2857# TAG: request_header_max_size (KB)
2858# This specifies the maximum size for HTTP headers in a request.
2859# Request headers are usually relatively small (about 512 bytes).
2860# Placing a limit on the request header size will catch certain
2861# bugs (for example with persistent connections) and possibly
2862# buffer-overflow or denial-of-service attacks.
2863#
2864#Default:
2865# request_header_max_size 20 KB
2866
2867# TAG: reply_header_max_size (KB)
2868# This specifies the maximum size for HTTP headers in a reply.
2869# Reply headers are usually relatively small (about 512 bytes).
2870# Placing a limit on the reply header size will catch certain
2871# bugs (for example with persistent connections) and possibly
2872# buffer-overflow or denial-of-service attacks.
2873#
2874#Default:
2875# reply_header_max_size 20 KB
2876
2877# TAG: request_body_max_size (KB)
2878# This specifies the maximum size for an HTTP request body.
2879# In other words, the maximum size of a PUT/POST request.
2880# A user who attempts to send a request with a body larger
2881# than this limit receives an "Invalid Request" error message.
2882# If you set this parameter to a zero (the default), there will
2883# be no limit imposed.
2884#
2885#Default:
2886# request_body_max_size 0 KB
2887
2888# TAG: broken_posts
2889# A list of ACL elements which, if matched, causes Squid to send
2890# an extra CRLF pair after the body of a PUT/POST request.
2891#
2892# Some HTTP servers has broken implementations of PUT/POST,
2893# and rely on an extra CRLF pair sent by some WWW clients.
2894#
2895# Quote from RFC2616 section 4.1 on this matter:
2896#
2897# Note: certain buggy HTTP/1.0 client implementations generate an
2898# extra CRLF's after a POST request. To restate what is explicitly
2899# forbidden by the BNF, an HTTP/1.1 client must not preface or follow
2900# a request with an extra CRLF.
2901#
2902#Example:
2903# acl buggy_server url_regex ^http://....
2904# broken_posts allow buggy_server
2905#
2906#Default:
2907# none
2908
2909# TAG: upgrade_http0.9
2910# This access list controls when HTTP/0.9 responses is upgraded
2911# to our current HTTP version. The default is to always upgrade.
2912#
2913# Some applications expect to be able to respond with non-HTTP
2914# responses and clients gets confused if the response is upgraded.
2915# For example SHOUTcast servers used for mp3 streaming.
2916#
2917# To enable some flexibility in detection of such applications
2918# the first line of the response is available in the internal header
2919# X-HTTP09-First-Line for use in the rep_header acl.
2920#
2921# Don't upgrade ShoutCast responses to HTTP
2922acl shoutcast rep_header X-HTTP09-First-Line ^ICY.[0-9]
2923upgrade_http0.9 deny shoutcast
2924
2925# TAG: via on|off
2926# If set (default), Squid will include a Via header in requests and
2927# replies as required by RFC2616.
2928#
2929#Default:
2930# via on
2931
2932# TAG: cache_vary
2933# When 'cache_vary' is set to off, response that have a
2934# Vary header will not be stored in the cache.
2935#
2936#Default:
2937# cache_vary on
2938
2939# TAG: broken_vary_encoding
2940# Many servers have broken support for on-the-fly Content-Encoding,
2941# returning the same ETag on both plain and gzip:ed variants.
2942# Vary replies matching this access list will have the cache split
2943# on the Accept-Encoding header of the request and not trusting the
2944# ETag to be unique.
2945#
2946# Apache mod_gzip and mod_deflate known to be broken so don't trust
2947# Apache to signal ETag correctly on such responses
2948acl apache rep_header Server ^Apache
2949broken_vary_encoding allow apache
2950
2951# TAG: collapsed_forwarding (on|off)
2952# This option enables multiple requests for the same URI to be
2953# processed as one request. Normally disabled to avoid increased
2954# latency on dynamic content, but there can be benefit from enabling
2955# this in accelerator setups where the web servers are the bottleneck
2956# and reliable and returns mostly cacheable information.
2957#
2958#Default:
2959# collapsed_forwarding off
2960
2961# TAG: refresh_stale_hit (time)
2962# This option changes the refresh algorithm to allow concurrent
2963# requests while an object is being refreshed to be processed as
2964# cache hits if the object expired less than X seconds ago. Default
2965# is 0 to disable this feature. This option is mostly interesting
2966# in accelerator setups where a few objects is accessed very
2967# frequently.
2968#
2969#Default:
2970# refresh_stale_hit 0 seconds
2971
2972# TAG: ie_refresh on|off
2973# Microsoft Internet Explorer up until version 5.5 Service
2974# Pack 1 has an issue with transparent proxies, wherein it
2975# is impossible to force a refresh. Turning this on provides
2976# a partial fix to the problem, by causing all IMS-REFRESH
2977# requests from older IE versions to check the origin server
2978# for fresh content. This reduces hit ratio by some amount
2979# (~10% in my experience), but allows users to actually get
2980# fresh content when they want it. Note because Squid
2981# cannot tell if the user is using 5.5 or 5.5SP1, the behavior
2982# of 5.5 is unchanged from old versions of Squid (i.e. a
2983# forced refresh is impossible). Newer versions of IE will,
2984# hopefully, continue to have the new behavior and will be
2985# handled based on that assumption. This option defaults to
2986# the old Squid behavior, which is better for hit ratios but
2987# worse for clients using IE, if they need to be able to
2988# force fresh content.
2989#
2990#Default:
2991# ie_refresh off
2992
2993# TAG: vary_ignore_expire on|off
2994# Many HTTP servers supporting Vary gives such objects
2995# immediate expiry time with no cache-control header
2996# when requested by a HTTP/1.0 client. This option
2997# enables Squid to ignore such expiry times until
2998# HTTP/1.1 is fully implemented.
2999# WARNING: This may eventually cause some varying
3000# objects not intended for caching to get cached.
3001#
3002#Default:
3003# vary_ignore_expire off
3004
3005# TAG: extension_methods
3006# Squid only knows about standardized HTTP request methods.
3007# You can add up to 20 additional "extension" methods here.
3008#
3009#Default:
3010# none
3011
3012# TAG: request_entities
3013# Squid defaults to deny GET and HEAD requests with request entities,
3014# as the meaning of such requests are undefined in the HTTP standard
3015# even if not explicitly forbidden.
3016#
3017# Set this directive to on if you have clients which insists
3018# on sending request entities in GET or HEAD requests. But be warned
3019# that there is server software (both proxies and web servers) which
3020# can fail to properly process this kind of request which may make you
3021# vulnerable to cache pollution attacks if enabled.
3022#
3023#Default:
3024# request_entities off
3025
3026# TAG: header_access
3027# Usage: header_access header_name allow|deny [!]aclname ...
3028#
3029# WARNING: Doing this VIOLATES the HTTP standard. Enabling
3030# this feature could make you liable for problems which it
3031# causes.
3032#
3033# This option replaces the old 'anonymize_headers' and the
3034# older 'http_anonymizer' option with something that is much
3035# more configurable. This new method creates a list of ACLs
3036# for each header, allowing you very fine-tuned header
3037# mangling.
3038#
3039# You can only specify known headers for the header name.
3040# Other headers are reclassified as 'Other'. You can also
3041# refer to all the headers with 'All'.
3042#
3043# For example, to achieve the same behavior as the old
3044# 'http_anonymizer standard' option, you should use:
3045#
3046# header_access From deny all
3047# header_access Referer deny all
3048# header_access Server deny all
3049# header_access User-Agent deny all
3050# header_access WWW-Authenticate deny all
3051# header_access Link deny all
3052#
3053# Or, to reproduce the old 'http_anonymizer paranoid' feature
3054# you should use:
3055#
3056# header_access Allow allow all
3057# header_access Authorization allow all
3058# header_access WWW-Authenticate allow all
3059# header_access Proxy-Authorization allow all
3060# header_access Proxy-Authenticate allow all
3061# header_access Cache-Control allow all
3062# header_access Content-Encoding allow all
3063# header_access Content-Length allow all
3064# header_access Content-Type allow all
3065# header_access Date allow all
3066# header_access Expires allow all
3067# header_access Host allow all
3068# header_access If-Modified-Since allow all
3069# header_access Last-Modified allow all
3070# header_access Location allow all
3071# header_access Pragma allow all
3072# header_access Accept allow all
3073# header_access Accept-Charset allow all
3074# header_access Accept-Encoding allow all
3075# header_access Accept-Language allow all
3076# header_access Content-Language allow all
3077# header_access Mime-Version allow all
3078# header_access Retry-After allow all
3079# header_access Title allow all
3080# header_access Connection allow all
3081# header_access Proxy-Connection allow all
3082# header_access All deny all
3083#
3084# By default, all headers are allowed (no anonymizing is
3085# performed).
3086#
3087#Default:
3088# none
3089
3090# TAG: header_replace
3091# Usage: header_replace header_name message
3092# Example: header_replace User-Agent Nutscrape/1.0 (CP/M; 8-bit)
3093#
3094# This option allows you to change the contents of headers
3095# denied with header_access above, by replacing them with
3096# some fixed string. This replaces the old fake_user_agent
3097# option.
3098#
3099# By default, headers are removed if denied.
3100#
3101#Default:
3102# none
3103
3104# TAG: relaxed_header_parser on|off|warn
3105# In the default "on" setting Squid accepts certain forms
3106# of non-compliant HTTP messages where it is unambiguous
3107# what the sending application intended even if the message
3108# is not correctly formatted. The messages is then normalized
3109# to the correct form when forwarded by Squid.
3110#
3111# If set to "warn" then a warning will be emitted in cache.log
3112# each time such HTTP error is encountered.
3113#
3114# If set to "off" then such HTTP errors will cause the request
3115# or response to be rejected.
3116#
3117#Default:
3118# relaxed_header_parser on
3119
3120# TAG: server_http11 on|off
3121# This option enables the use ot HTTP/1.1 on outgoing "direct" requests.
3122# See also the http11 cache_peer option.
3123# Note: The HTTP/1.1 support is still incomplete, with an
3124# internal HTTP/1.0 hop. As result 1xx responses will not
3125# be forwarded.
3126#
3127#Default:
3128# server_http11 off
3129
3130# TAG: ignore_expect_100 on|off
3131# This option makes Squid ignore any Expect: 100-continue header present
3132# in the request.
3133# Note: Enabling this is a HTTP protocol violation, but some client may
3134# not handle it well..
3135#
3136#Default:
3137# ignore_expect_100 off
3138
3139# TAG: external_refresh_check
3140# This option defines an external helper for determining whether to
3141# refresh a stale response. It will be called when Squid receives a
3142# request for a cached response that is stale; the helper can either
3143# confirm that the response is stale with a STALE response, or
3144# extend the freshness of the response (thereby avoiding a refresh
3145# check) with a FRESH response, along with a freshness=nnn keyword.
3146#
3147# external_refresh_check [options] FORMAT.. /path/to/helper [helper_args]
3148#
3149# If present, helper_args will be passed to the helper on the command
3150# line verbatim.
3151#
3152# Options:
3153#
3154# children=n Number of processes to spawn to service external
3155# refresh checks (default 5).
3156# concurrency=n Concurrency level per process. Only used with
3157# helpers capable of processing more than one query
3158# at a time.
3159#
3160# When using the concurrency option, the protocol is changed by introducing
3161# a query channel tag infront of the request/response. The query channel
3162# tag is a number between 0 and concurrency-1.
3163#
3164# FORMAT specifications:
3165#
3166# %CACHE_URI The URI of the cached response
3167# %RES{Header} HTTP response header value
3168# %AGE The age of the cached response
3169#
3170# The request sent to the helper consists of the data in the format
3171# specification in the order specified.
3172#
3173# The helper receives lines per the above format specification, and
3174# returns lines starting with OK or ERR indicating the validity of
3175# the request and optionally followed by additional keywords with
3176# more details. URL escaping is used to protect each value in both
3177# requests and responses.
3178#
3179# General result syntax:
3180#
3181# FRESH / STALE keyword=value ...
3182#
3183# Defined keywords:
3184#
3185# freshness=nnn The number of seconds to extend the freshness of
3186# the response by.
3187# log=string String to be logged in access.log. Available as
3188# %ef in logformat specifications.
3189# res{Header}=value
3190# Value to update response headers with. If already
3191# present, the supplied value completely replaces
3192# the cached value.
3193#
3194# In the event of a helper-related error (e.g., overload), Squid
3195# will always default to STALE.
3196#
3197#Default:
3198# none
3199
3200
3201# TIMEOUTS
3202# -----------------------------------------------------------------------------
3203
3204# TAG: forward_timeout time-units
3205# This parameter specifies how long Squid should at most attempt in
3206# finding a forwarding path for the request before giving up.
3207#
3208#Default:
3209# forward_timeout 4 minutes
3210
3211# TAG: connect_timeout time-units
3212# This parameter specifies how long to wait for the TCP connect to
3213# the requested server or peer to complete before Squid should
3214# attempt to find another path where to forward the request.
3215#
3216#Default:
3217# connect_timeout 1 minute
3218
3219# TAG: peer_connect_timeout time-units
3220# This parameter specifies how long to wait for a pending TCP
3221# connection to a peer cache. The default is 30 seconds. You
3222# may also set different timeout values for individual neighbors
3223# with the 'connect-timeout' option on a 'cache_peer' line.
3224#
3225#Default:
3226# peer_connect_timeout 30 seconds
3227
3228# TAG: read_timeout time-units
3229# The read_timeout is applied on server-side connections. After
3230# each successful read(), the timeout will be extended by this
3231# amount. If no data is read again after this amount of time,
3232# the request is aborted and logged with ERR_READ_TIMEOUT. The
3233# default is 15 minutes.
3234#
3235#Default:
3236# read_timeout 15 minutes
3237
3238# TAG: request_timeout
3239# How long to wait for an HTTP request after initial
3240# connection establishment.
3241#
3242#Default:
3243# request_timeout 5 minutes
3244
3245# TAG: persistent_request_timeout
3246# How long to wait for the next HTTP request on a persistent
3247# connection after the previous request completes.
3248#
3249#Default:
3250# persistent_request_timeout 2 minutes
3251
3252# TAG: client_lifetime time-units
3253# The maximum amount of time a client (browser) is allowed to
3254# remain connected to the cache process. This protects the Cache
3255# from having a lot of sockets (and hence file descriptors) tied up
3256# in a CLOSE_WAIT state from remote clients that go away without
3257# properly shutting down (either because of a network failure or
3258# because of a poor client implementation). The default is one
3259# day, 1440 minutes.
3260#
3261# NOTE: The default value is intended to be much larger than any
3262# client would ever need to be connected to your cache. You
3263# should probably change client_lifetime only as a last resort.
3264# If you seem to have many client connections tying up
3265# filedescriptors, we recommend first tuning the read_timeout,
3266# request_timeout, persistent_request_timeout and quick_abort values.
3267#
3268#Default:
3269# client_lifetime 1 day
3270
3271# TAG: half_closed_clients
3272# Some clients may shutdown the sending side of their TCP
3273# connections, while leaving their receiving sides open. Sometimes,
3274# Squid can not tell the difference between a half-closed and a
3275# fully-closed TCP connection. By default, half-closed client
3276# connections are kept open until a read(2) or write(2) on the
3277# socket returns an error. Change this option to 'off' and Squid
3278# will immediately close client connections when read(2) returns
3279# "no more data to read."
3280#
3281#Default:
3282# half_closed_clients on
3283
3284# TAG: pconn_timeout
3285# Timeout for idle persistent connections to servers and other
3286# proxies.
3287#
3288#Default:
3289# pconn_timeout 1 minute
3290
3291# TAG: ident_timeout
3292# Maximum time to wait for IDENT lookups to complete.
3293#
3294# If this is too high, and you enabled IDENT lookups from untrusted
3295# users, you might be susceptible to denial-of-service by having
3296# many ident requests going at once.
3297#
3298#Default:
3299# ident_timeout 10 seconds
3300
3301# TAG: shutdown_lifetime time-units
3302# When SIGTERM or SIGHUP is received, the cache is put into
3303# "shutdown pending" mode until all active sockets are closed.
3304# This value is the lifetime to set for all open descriptors
3305# during shutdown mode. Any active clients after this many
3306# seconds will receive a 'timeout' message.
3307#
3308#Default:
3309# shutdown_lifetime 30 seconds
3310
3311
3312# ADMINISTRATIVE PARAMETERS
3313# -----------------------------------------------------------------------------
3314
3315# TAG: cache_mgr
3316# Email-address of local cache manager who will receive
3317# mail if the cache dies. The default is "webmaster".
3318#
3319#Default:
3320# cache_mgr webmaster
3321
3322# TAG: mail_from
3323# From: email-address for mail sent when the cache dies.
3324# The default is to use 'appname@unique_hostname'.
3325# Default appname value is "squid", can be changed into
3326# src/globals.h before building squid.
3327#
3328#Default:
3329# none
3330
3331# TAG: mail_program
3332# Email program used to send mail if the cache dies.
3333# The default is "mail". The specified program must comply
3334# with the standard Unix mail syntax:
3335# mail-program recipient < mailfile
3336#
3337# Optional command line options can be specified.
3338#
3339#Default:
3340# mail_program mail
3341
3342# TAG: cache_effective_user
3343# If you start Squid as root, it will change its effective/real
3344# UID/GID to the user specified below. The default is to change
3345# to UID to nobody. If you define cache_effective_user, but not
3346# cache_effective_group, Squid sets the GID to the effective
3347# user's default group ID (taken from the password file) and
3348# supplementary group list from the from groups membership of
3349# cache_effective_user.
3350#
3351#Default:
3352# cache_effective_user nobody
3353
3354# TAG: cache_effective_group
3355# If you want Squid to run with a specific GID regardless of
3356# the group memberships of the effective user then set this
3357# to the group (or GID) you want Squid to run as. When set
3358# all other group privileges of the effective user is ignored
3359# and only this GID is effective. If Squid is not started as
3360# root the user starting Squid must be member of the specified
3361# group.
3362#
3363#Default:
3364# none
3365
3366# TAG: httpd_suppress_version_string on|off
3367# Suppress Squid version string info in HTTP headers and HTML error pages.
3368#
3369#Default:
3370# httpd_suppress_version_string off
3371
3372# TAG: visible_hostname
3373# If you want to present a special hostname in error messages, etc,
3374# define this. Otherwise, the return value of gethostname()
3375# will be used. If you have multiple caches in a cluster and
3376# get errors about IP-forwarding you must set them to have individual
3377# names with this setting.
3378#
3379#Default:
3380# none
3381
3382# TAG: unique_hostname
3383# If you want to have multiple machines with the same
3384# 'visible_hostname' you must give each machine a different
3385# 'unique_hostname' so forwarding loops can be detected.
3386#
3387#Default:
3388# none
3389
3390# TAG: hostname_aliases
3391# A list of other DNS names your cache has.
3392#
3393#Default:
3394# none
3395
3396# TAG: umask
3397# Minimum umask which should be enforced while the proxy
3398# is running, in addition to the umask set at startup.
3399#
3400# Note: Should start with a 0 to indicate the normal octal
3401# representation of umasks
3402#
3403#Default:
3404# umask 027
3405
3406
3407# OPTIONS FOR THE CACHE REGISTRATION SERVICE
3408# -----------------------------------------------------------------------------
3409#
3410# This section contains parameters for the (optional) cache
3411# announcement service. This service is provided to help
3412# cache administrators locate one another in order to join or
3413# create cache hierarchies.
3414#
3415# An 'announcement' message is sent (via UDP) to the registration
3416# service by Squid. By default, the announcement message is NOT
3417# SENT unless you enable it with 'announce_period' below.
3418#
3419# The announcement message includes your hostname, plus the
3420# following information from this configuration file:
3421#
3422# http_port
3423# icp_port
3424# cache_mgr
3425#
3426# All current information is processed regularly and made
3427# available on the Web at http://www.ircache.net/Cache/Tracker/.
3428
3429# TAG: announce_period
3430# This is how frequently to send cache announcements. The
3431# default is `0' which disables sending the announcement
3432# messages.
3433#
3434# To enable announcing your cache, just uncomment the line
3435# below.
3436#
3437#Default:
3438# announce_period 0
3439#
3440#To enable announcing your cache, just uncomment the line below.
3441#announce_period 1 day
3442
3443# TAG: announce_host
3444# TAG: announce_file
3445# TAG: announce_port
3446# announce_host and announce_port set the hostname and port
3447# number where the registration message will be sent.
3448#
3449# Hostname will default to 'tracker.ircache.net' and port will
3450# default default to 3131. If the 'filename' argument is given,
3451# the contents of that file will be included in the announce
3452# message.
3453#
3454#Default:
3455# announce_host tracker.ircache.net
3456# announce_port 3131
3457
3458
3459# HTTPD-ACCELERATOR OPTIONS
3460# -----------------------------------------------------------------------------
3461
3462# TAG: httpd_accel_no_pmtu_disc on|off
3463# In many setups of transparently intercepting proxies Path-MTU
3464# discovery can not work on traffic towards the clients. This is
3465# the case when the intercepting device does not fully track
3466# connections and fails to forward ICMP must fragment messages
3467# to the cache server.
3468#
3469# If you have such setup and experience that certain clients
3470# sporadically hang or never complete requests set this to on.
3471#
3472#Default:
3473# httpd_accel_no_pmtu_disc off
3474
3475
3476# DELAY POOL PARAMETERS
3477# -----------------------------------------------------------------------------
3478
3479# TAG: delay_pools
3480# Note: This option is only available if Squid is rebuilt with the
3481# --enable-delay-pools option
3482#
3483# This represents the number of delay pools to be used. For example,
3484# if you have one class 2 delay pool and one class 3 delays pool, you
3485# have a total of 2 delay pools.
3486#
3487#Default:
3488# delay_pools 0
3489
3490# TAG: delay_class
3491# Note: This option is only available if Squid is rebuilt with the
3492# --enable-delay-pools option
3493#
3494# This defines the class of each delay pool. There must be exactly one
3495# delay_class line for each delay pool. For example, to define two
3496# delay pools, one of class 2 and one of class 3, the settings above
3497# and here would be:
3498#
3499#Example:
3500# delay_pools 2 # 2 delay pools
3501# delay_class 1 2 # pool 1 is a class 2 pool
3502# delay_class 2 3 # pool 2 is a class 3 pool
3503#
3504# The delay pool classes are:
3505#
3506# class 1 Everything is limited by a single aggregate
3507# bucket.
3508#
3509# class 2 Everything is limited by a single aggregate
3510# bucket as well as an "individual" bucket chosen
3511# from bits 25 through 32 of the IP address.
3512#
3513# class 3 Everything is limited by a single aggregate
3514# bucket as well as a "network" bucket chosen
3515# from bits 17 through 24 of the IP address and a
3516# "individual" bucket chosen from bits 17 through
3517# 32 of the IP address.
3518#
3519# NOTE: If an IP address is a.b.c.d
3520# -> bits 25 through 32 are "d"
3521# -> bits 17 through 24 are "c"
3522# -> bits 17 through 32 are "c * 256 + d"
3523#
3524#Default:
3525# none
3526
3527# TAG: delay_access
3528# Note: This option is only available if Squid is rebuilt with the
3529# --enable-delay-pools option
3530#
3531# This is used to determine which delay pool a request falls into.
3532#
3533# delay_access is sorted per pool and the matching starts with pool 1,
3534# then pool 2, ..., and finally pool N. The first delay pool where the
3535# request is allowed is selected for the request. If it does not allow
3536# the request to any pool then the request is not delayed (default).
3537#
3538# For example, if you want some_big_clients in delay
3539# pool 1 and lotsa_little_clients in delay pool 2:
3540#
3541#Example:
3542# delay_access 1 allow some_big_clients
3543# delay_access 1 deny all
3544# delay_access 2 allow lotsa_little_clients
3545# delay_access 2 deny all
3546#
3547#Default:
3548# none
3549
3550# TAG: delay_parameters
3551# Note: This option is only available if Squid is rebuilt with the
3552# --enable-delay-pools option
3553#
3554# This defines the parameters for a delay pool. Each delay pool has
3555# a number of "buckets" associated with it, as explained in the
3556# description of delay_class. For a class 1 delay pool, the syntax is:
3557#
3558#delay_parameters pool aggregate
3559#
3560# For a class 2 delay pool:
3561#
3562#delay_parameters pool aggregate individual
3563#
3564# For a class 3 delay pool:
3565#
3566#delay_parameters pool aggregate network individual
3567#
3568# The variables here are:
3569#
3570# pool a pool number - ie, a number between 1 and the
3571# number specified in delay_pools as used in
3572# delay_class lines.
3573#
3574# aggregate the "delay parameters" for the aggregate bucket
3575# (class 1, 2, 3).
3576#
3577# individual the "delay parameters" for the individual
3578# buckets (class 2, 3).
3579#
3580# network the "delay parameters" for the network buckets
3581# (class 3).
3582#
3583# A pair of delay parameters is written restore/maximum, where restore is
3584# the number of bytes (not bits - modem and network speeds are usually
3585# quoted in bits) per second placed into the bucket, and maximum is the
3586# maximum number of bytes which can be in the bucket at any time.
3587#
3588# For example, if delay pool number 1 is a class 2 delay pool as in the
3589# above example, and is being used to strictly limit each host to 64kbps
3590# (plus overheads), with no overall limit, the line is:
3591#
3592#delay_parameters 1 -1/-1 8000/8000
3593#
3594# Note that the figure -1 is used to represent "unlimited".
3595#
3596# And, if delay pool number 2 is a class 3 delay pool as in the above
3597# example, and you want to limit it to a total of 256kbps (strict limit)
3598# with each 8-bit network permitted 64kbps (strict limit) and each
3599# individual host permitted 4800bps with a bucket maximum size of 64kb
3600# to permit a decent web page to be downloaded at a decent speed
3601# (if the network is not being limited due to overuse) but slow down
3602# large downloads more significantly:
3603#
3604#delay_parameters 2 32000/32000 8000/8000 600/8000
3605#
3606# There must be one delay_parameters line for each delay pool.
3607#
3608#Default:
3609# none
3610
3611# TAG: delay_initial_bucket_level (percent, 0-100)
3612# Note: This option is only available if Squid is rebuilt with the
3613# --enable-delay-pools option
3614#
3615# The initial bucket percentage is used to determine how much is put
3616# in each bucket when squid starts, is reconfigured, or first notices
3617# a host accessing it (in class 2 and class 3, individual hosts and
3618# networks only have buckets associated with them once they have been
3619# "seen" by squid).
3620#
3621#Default:
3622# delay_initial_bucket_level 50
3623
3624
3625# WCCPv1 AND WCCPv2 CONFIGURATION OPTIONS
3626# -----------------------------------------------------------------------------
3627
3628# TAG: wccp_router
3629# Note: This option is only available if Squid is rebuilt with the
3630# --enable-wccp option
3631#
3632# TAG: wccp2_router
3633# Note: This option is only available if Squid is rebuilt with the
3634# --enable-wccpv2 option
3635#
3636# Use this option to define your WCCP ``home'' router for
3637# Squid.
3638#
3639# wccp_router supports a single WCCP(v1) router
3640#
3641# wccp2_router supports multiple WCCPv2 routers
3642#
3643# only one of the two may be used at the same time and defines
3644# which version of WCCP to use.
3645#
3646#Default:
3647# wccp_router 0.0.0.0
3648
3649# TAG: wccp_version
3650# Note: This option is only available if Squid is rebuilt with the
3651# --enable-wccp option
3652#
3653# This directive is only relevant if you need to set up WCCP(v1)
3654# to some very old and end-of-life Cisco routers. In all other
3655# setups it must be left unset or at the default setting.
3656# It defines an internal version in the WCCP(v1) protocol,
3657# with version 4 being the officially documented protocol.
3658#
3659# According to some users, Cisco IOS 11.2 and earlier only
3660# support WCCP version 3. If you're using that or an earlier
3661# version of IOS, you may need to change this value to 3, otherwise
3662# do not specify this parameter.
3663#
3664#Default:
3665# wccp_version 4
3666
3667# TAG: wccp2_rebuild_wait
3668# Note: This option is only available if Squid is rebuilt with the
3669# --enable-wccpv2 option
3670#
3671# If this is enabled Squid will wait for the cache dir rebuild to finish
3672# before sending the first wccp2 HereIAm packet
3673#
3674#Default:
3675# wccp2_rebuild_wait on
3676
3677# TAG: wccp2_forwarding_method
3678# Note: This option is only available if Squid is rebuilt with the
3679# --enable-wccpv2 option
3680#
3681# WCCP2 allows the setting of forwarding methods between the
3682# router/switch and the cache. Valid values are as follows:
3683#
3684# 1 - GRE encapsulation (forward the packet in a GRE/WCCP tunnel)
3685# 2 - L2 redirect (forward the packet using Layer 2/MAC rewriting)
3686#
3687# Currently (as of IOS 12.4) cisco routers only support GRE.
3688# Cisco switches only support the L2 redirect assignment method.
3689#
3690#Default:
3691# wccp2_forwarding_method 1
3692
3693# TAG: wccp2_return_method
3694# Note: This option is only available if Squid is rebuilt with the
3695# --enable-wccpv2 option
3696#
3697# WCCP2 allows the setting of return methods between the
3698# router/switch and the cache for packets that the cache
3699# decides not to handle. Valid values are as follows:
3700#
3701# 1 - GRE encapsulation (forward the packet in a GRE/WCCP tunnel)
3702# 2 - L2 redirect (forward the packet using Layer 2/MAC rewriting)
3703#
3704# Currently (as of IOS 12.4) cisco routers only support GRE.
3705# Cisco switches only support the L2 redirect assignment.
3706#
3707# If the "ip wccp redirect exclude in" command has been
3708# enabled on the cache interface, then it is still safe for
3709# the proxy server to use a l2 redirect method even if this
3710# option is set to GRE.
3711#
3712#Default:
3713# wccp2_return_method 1
3714
3715# TAG: wccp2_assignment_method
3716# Note: This option is only available if Squid is rebuilt with the
3717# --enable-wccpv2 option
3718#
3719# WCCP2 allows the setting of methods to assign the WCCP hash
3720# Valid values are as follows:
3721#
3722# 1 - Hash assignment
3723# 2 - Mask assignment
3724#
3725# As a general rule, cisco routers support the hash assignment method
3726# and cisco switches support the mask assignment method.
3727#
3728#Default:
3729# wccp2_assignment_method 1
3730
3731# TAG: wccp2_service
3732# Note: This option is only available if Squid is rebuilt with the
3733# --enable-wccpv2 option
3734#
3735# WCCP2 allows for multiple traffic services. There are two
3736# types: "standard" and "dynamic". The standard type defines
3737# one service id - http (id 0). The dynamic service ids can be from
3738# 51 to 255 inclusive. In order to use a dynamic service id
3739# one must define the type of traffic to be redirected; this is done
3740# using the wccp2_service_info option.
3741#
3742# The "standard" type does not require a wccp2_service_info option,
3743# just specifying the service id will suffice.
3744#
3745# MD5 service authentication can be enabled by adding
3746# "password=<password>" to the end of this service declaration.
3747#
3748# Examples:
3749#
3750# wccp2_service standard 0 # for the 'web-cache' standard service
3751# wccp2_service dynamic 80 # a dynamic service type which will be
3752# # fleshed out with subsequent options.
3753# wccp2_service standard 0 password=foo
3754#
3755#
3756#Default:
3757# wccp2_service standard 0
3758
3759# TAG: wccp2_service_info
3760# Note: This option is only available if Squid is rebuilt with the
3761# --enable-wccpv2 option
3762#
3763# Dynamic WCCPv2 services require further information to define the
3764# traffic you wish to have diverted.
3765#
3766# The format is:
3767#
3768# wccp2_service_info <id> protocol=<protocol> flags=<flag>,<flag>..
3769# priority=<priority> ports=<port>,<port>..
3770#
3771# The relevant WCCPv2 flags:
3772# + src_ip_hash, dst_ip_hash
3773# + source_port_hash, dst_port_hash
3774# + src_ip_alt_hash, dst_ip_alt_hash
3775# + src_port_alt_hash, dst_port_alt_hash
3776# + ports_source
3777#
3778# The port list can be one to eight entries.
3779#
3780# Example:
3781#
3782# wccp2_service_info 80 protocol=tcp flags=src_ip_hash,ports_source
3783# priority=240 ports=80
3784#
3785# Note: the service id must have been defined by a previous
3786# 'wccp2_service dynamic <id>' entry.
3787#
3788#Default:
3789# none
3790
3791# TAG: wccp2_weight
3792# Note: This option is only available if Squid is rebuilt with the
3793# --enable-wccpv2 option
3794#
3795# Each cache server gets assigned a set of the destination
3796# hash proportional to their weight.
3797#
3798#Default:
3799# wccp2_weight 10000
3800
3801# TAG: wccp_address
3802# Note: This option is only available if Squid is rebuilt with the
3803# --enable-wccp option
3804#
3805# TAG: wccp2_address
3806# Note: This option is only available if Squid is rebuilt with the
3807# --enable-wccpv2 option
3808#
3809# Use this option if you require WCCP to use a specific
3810# interface address.
3811#
3812# The default behavior is to not bind to any specific address.
3813#
3814#Default:
3815# wccp_address 0.0.0.0
3816# wccp2_address 0.0.0.0
3817
3818
3819# PERSISTENT CONNECTION HANDLING
3820# -----------------------------------------------------------------------------
3821#
3822# Also see "pconn_timeout" in the TIMEOUTS section
3823
3824# TAG: client_persistent_connections
3825# TAG: server_persistent_connections
3826# Persistent connection support for clients and servers. By
3827# default, Squid uses persistent connections (when allowed)
3828# with its clients and servers. You can use these options to
3829# disable persistent connections with clients and/or servers.
3830#
3831#Default:
3832# client_persistent_connections on
3833# server_persistent_connections on
3834
3835# TAG: persistent_connection_after_error
3836# With this directive the use of persistent connections after
3837# HTTP errors can be disabled. Useful if you have clients
3838# who fail to handle errors on persistent connections proper.
3839#
3840#Default:
3841# persistent_connection_after_error off
3842
3843# TAG: detect_broken_pconn
3844# Some servers have been found to incorrectly signal the use
3845# of HTTP/1.0 persistent connections even on replies not
3846# compatible, causing significant delays. This server problem
3847# has mostly been seen on redirects.
3848#
3849# By enabling this directive Squid attempts to detect such
3850# broken replies and automatically assume the reply is finished
3851# after 10 seconds timeout.
3852#
3853#Default:
3854# detect_broken_pconn off
3855
3856
3857# CACHE DIGEST OPTIONS
3858# -----------------------------------------------------------------------------
3859
3860# TAG: digest_generation
3861# This controls whether the server will generate a Cache Digest
3862# of its contents.
3863#
3864#Default:
3865# digest_generation on
3866
3867# TAG: digest_bits_per_entry
3868# This is the number of bits of the server's Cache Digest which
3869# will be associated with the Digest entry for a given HTTP
3870# Method and URL (public key) combination. The default is 5.
3871#
3872#Default:
3873# digest_bits_per_entry 5
3874
3875# TAG: digest_rebuild_period (seconds)
3876# This is the wait time between Cache Digest rebuilds.
3877#
3878#Default:
3879# digest_rebuild_period 1 hour
3880
3881# TAG: digest_rewrite_period (seconds)
3882# This is the wait time between Cache Digest writes to disk.
3883#
3884#Default:
3885# digest_rewrite_period 1 hour
3886
3887# TAG: digest_swapout_chunk_size (bytes)
3888# This is the number of bytes of the Cache Digest to write to
3889# disk at a time. It defaults to 4096 bytes (4KB), the Squid
3890# default swap page.
3891#
3892#Default:
3893# digest_swapout_chunk_size 4096 bytes
3894
3895# TAG: digest_rebuild_chunk_percentage (percent, 0-100)
3896# This is the percentage of the Cache Digest to be scanned at a
3897# time. By default it is set to 10% of the Cache Digest.
3898#
3899#Default:
3900# digest_rebuild_chunk_percentage 10
3901
3902
3903# SNMP OPTIONS
3904# -----------------------------------------------------------------------------
3905
3906# TAG: snmp_port
3907# Squid can now serve statistics and status information via SNMP.
3908# By default it listens to port 3401 on the machine. If you don't
3909# wish to use SNMP, set this to "0".
3910#
3911#Default:
3912# snmp_port 3401
3913
3914# TAG: snmp_access
3915# Allowing or denying access to the SNMP port.
3916#
3917# All access to the agent is denied by default.
3918# usage:
3919#
3920# snmp_access allow|deny [!]aclname ...
3921#
3922#Example:
3923# snmp_access allow snmppublic localhost
3924# snmp_access deny all
3925#
3926#Default:
3927# snmp_access deny all
3928
3929# TAG: snmp_incoming_address
3930# TAG: snmp_outgoing_address
3931# Just like 'udp_incoming_address' above, but for the SNMP port.
3932#
3933# snmp_incoming_address is used for the SNMP socket receiving
3934# messages from SNMP agents.
3935# snmp_outgoing_address is used for SNMP packets returned to SNMP
3936# agents.
3937#
3938# The default snmp_incoming_address (0.0.0.0) is to listen on all
3939# available network interfaces.
3940#
3941# If snmp_outgoing_address is set to 255.255.255.255 (the default)
3942# it will use the same socket as snmp_incoming_address. Only
3943# change this if you want to have SNMP replies sent using another
3944# address than where this Squid listens for SNMP queries.
3945#
3946# NOTE, snmp_incoming_address and snmp_outgoing_address can not have
3947# the same value since they both use port 3401.
3948#
3949#Default:
3950# snmp_incoming_address 0.0.0.0
3951# snmp_outgoing_address 255.255.255.255
3952
3953
3954# ICP OPTIONS
3955# -----------------------------------------------------------------------------
3956
3957# TAG: icp_port
3958# The port number where Squid sends and receives ICP queries to
3959# and from neighbor caches. Default is 3130. To disable use
3960# "0". May be overridden with -u on the command line.
3961#
3962#Default:
3963# icp_port 3130
3964
3965# TAG: htcp_port
3966# The port number where Squid sends and receives HTCP queries to
3967# and from neighbor caches. Default is 4827. To disable use
3968# "0".
3969#
3970#Default:
3971# htcp_port 4827
3972
3973# TAG: log_icp_queries on|off
3974# If set, ICP queries are logged to access.log. You may wish
3975# do disable this if your ICP load is VERY high to speed things
3976# up or to simplify log analysis.
3977#
3978#Default:
3979# log_icp_queries on
3980
3981# TAG: udp_incoming_address
3982# udp_incoming_address is used for UDP packets received from other
3983# caches.
3984#
3985# The default behavior is to not bind to any specific address.
3986#
3987# Only change this if you want to have all UDP queries received on
3988# a specific interface/address.
3989#
3990# NOTE: udp_incoming_address is used by the ICP, HTCP, and DNS
3991# modules. Altering it will affect all of them in the same manner.
3992#
3993# see also; udp_outgoing_address
3994#
3995# NOTE, udp_incoming_address and udp_outgoing_address can not
3996# have the same value since they both use the same port.
3997#
3998#Default:
3999# udp_incoming_address 0.0.0.0
4000
4001# TAG: udp_outgoing_address
4002# udp_outgoing_address is used for UDP packets sent out to other
4003# caches.
4004#
4005# The default behavior is to not bind to any specific address.
4006#
4007# Instead it will use the same socket as udp_incoming_address.
4008# Only change this if you want to have UDP queries sent using another
4009# address than where this Squid listens for UDP queries from other
4010# caches.
4011#
4012# NOTE: udp_outgoing_address is used by the ICP, HTCP, and DNS
4013# modules. Altering it will affect all of them in the same manner.
4014#
4015# see also; udp_incoming_address
4016#
4017# NOTE, udp_incoming_address and udp_outgoing_address can not
4018# have the same value since they both use the same port.
4019#
4020#Default:
4021# udp_outgoing_address 255.255.255.255
4022
4023# TAG: icp_hit_stale on|off
4024# If you want to return ICP_HIT for stale cache objects, set this
4025# option to 'on'. If you have sibling relationships with caches
4026# in other administrative domains, this should be 'off'. If you only
4027# have sibling relationships with caches under your control,
4028# it is probably okay to set this to 'on'.
4029# If set to 'on', your siblings should use the option "allow-miss"
4030# on their cache_peer lines for connecting to you.
4031#
4032#Default:
4033# icp_hit_stale off
4034
4035# TAG: minimum_direct_hops
4036# If using the ICMP pinging stuff, do direct fetches for sites
4037# which are no more than this many hops away.
4038#
4039#Default:
4040# minimum_direct_hops 4
4041
4042# TAG: minimum_direct_rtt
4043# If using the ICMP pinging stuff, do direct fetches for sites
4044# which are no more than this many rtt milliseconds away.
4045#
4046#Default:
4047# minimum_direct_rtt 400
4048
4049# TAG: netdb_low
4050# TAG: netdb_high
4051# The low and high water marks for the ICMP measurement
4052# database. These are counts, not percents. The defaults are
4053# 900 and 1000. When the high water mark is reached, database
4054# entries will be deleted until the low mark is reached.
4055#
4056#Default:
4057# netdb_low 900
4058# netdb_high 1000
4059
4060# TAG: netdb_ping_period
4061# The minimum period for measuring a site. There will be at
4062# least this much delay between successive pings to the same
4063# network. The default is five minutes.
4064#
4065#Default:
4066# netdb_ping_period 5 minutes
4067
4068# TAG: query_icmp on|off
4069# If you want to ask your peers to include ICMP data in their ICP
4070# replies, enable this option.
4071#
4072# If your peer has configured Squid (during compilation) with
4073# '--enable-icmp' that peer will send ICMP pings to origin server
4074# sites of the URLs it receives. If you enable this option the
4075# ICP replies from that peer will include the ICMP data (if available).
4076# Then, when choosing a parent cache, Squid will choose the parent with
4077# the minimal RTT to the origin server. When this happens, the
4078# hierarchy field of the access.log will be
4079# "CLOSEST_PARENT_MISS". This option is off by default.
4080#
4081#Default:
4082# query_icmp off
4083
4084# TAG: test_reachability on|off
4085# When this is 'on', ICP MISS replies will be ICP_MISS_NOFETCH
4086# instead of ICP_MISS if the target host is NOT in the ICMP
4087# database, or has a zero RTT.
4088#
4089#Default:
4090# test_reachability off
4091
4092# TAG: icp_query_timeout (msec)
4093# Normally Squid will automatically determine an optimal ICP
4094# query timeout value based on the round-trip-time of recent ICP
4095# queries. If you want to override the value determined by
4096# Squid, set this 'icp_query_timeout' to a non-zero value. This
4097# value is specified in MILLISECONDS, so, to use a 2-second
4098# timeout (the old default), you would write:
4099#
4100# icp_query_timeout 2000
4101#
4102#Default:
4103# icp_query_timeout 0
4104
4105# TAG: maximum_icp_query_timeout (msec)
4106# Normally the ICP query timeout is determined dynamically. But
4107# sometimes it can lead to very large values (say 5 seconds).
4108# Use this option to put an upper limit on the dynamic timeout
4109# value. Do NOT use this option to always use a fixed (instead
4110# of a dynamic) timeout value. To set a fixed timeout see the
4111# 'icp_query_timeout' directive.
4112#
4113#Default:
4114# maximum_icp_query_timeout 2000
4115
4116# TAG: minimum_icp_query_timeout (msec)
4117# Normally the ICP query timeout is determined dynamically. But
4118# sometimes it can lead to very small timeouts, even lower than
4119# the normal latency variance on your link due to traffic.
4120# Use this option to put an lower limit on the dynamic timeout
4121# value. Do NOT use this option to always use a fixed (instead
4122# of a dynamic) timeout value. To set a fixed timeout see the
4123# 'icp_query_timeout' directive.
4124#
4125#Default:
4126# minimum_icp_query_timeout 5
4127
4128
4129# MULTICAST ICP OPTIONS
4130# -----------------------------------------------------------------------------
4131
4132# TAG: mcast_groups
4133# This tag specifies a list of multicast groups which your server
4134# should join to receive multicasted ICP queries.
4135#
4136# NOTE! Be very careful what you put here! Be sure you
4137# understand the difference between an ICP _query_ and an ICP
4138# _reply_. This option is to be set only if you want to RECEIVE
4139# multicast queries. Do NOT set this option to SEND multicast
4140# ICP (use cache_peer for that). ICP replies are always sent via
4141# unicast, so this option does not affect whether or not you will
4142# receive replies from multicast group members.
4143#
4144# You must be very careful to NOT use a multicast address which
4145# is already in use by another group of caches.
4146#
4147# If you are unsure about multicast, please read the Multicast
4148# chapter in the Squid FAQ (http://www.squid-cache.org/FAQ/).
4149#
4150# Usage: mcast_groups 239.128.16.128 224.0.1.20
4151#
4152# By default, Squid doesn't listen on any multicast groups.
4153#
4154#Default:
4155# none
4156
4157# TAG: mcast_miss_addr
4158# Note: This option is only available if Squid is rebuilt with the
4159# --enable-multicast-miss option
4160#
4161# If you enable this option, every "cache miss" URL will
4162# be sent out on the specified multicast address.
4163#
4164# Do not enable this option unless you are are absolutely
4165# certain you understand what you are doing.
4166#
4167#Default:
4168# mcast_miss_addr 255.255.255.255
4169
4170# TAG: mcast_miss_ttl
4171# Note: This option is only available if Squid is rebuilt with the
4172# --enable-multicast-miss option
4173#
4174# This is the time-to-live value for packets multicasted
4175# when multicasting off cache miss URLs is enabled. By
4176# default this is set to 'site scope', i.e. 16.
4177#
4178#Default:
4179# mcast_miss_ttl 16
4180
4181# TAG: mcast_miss_port
4182# Note: This option is only available if Squid is rebuilt with the
4183# --enable-multicast-miss option
4184#
4185# This is the port number to be used in conjunction with
4186# 'mcast_miss_addr'.
4187#
4188#Default:
4189# mcast_miss_port 3135
4190
4191# TAG: mcast_miss_encode_key
4192# Note: This option is only available if Squid is rebuilt with the
4193# --enable-multicast-miss option
4194#
4195# The URLs that are sent in the multicast miss stream are
4196# encrypted. This is the encryption key.
4197#
4198#Default:
4199# mcast_miss_encode_key XXXXXXXXXXXXXXXX
4200
4201# TAG: mcast_icp_query_timeout (msec)
4202# For multicast peers, Squid regularly sends out ICP "probes" to
4203# count how many other peers are listening on the given multicast
4204# address. This value specifies how long Squid should wait to
4205# count all the replies. The default is 2000 msec, or 2
4206# seconds.
4207#
4208#Default:
4209# mcast_icp_query_timeout 2000
4210
4211
4212# INTERNAL ICON OPTIONS
4213# -----------------------------------------------------------------------------
4214
4215# TAG: icon_directory
4216# Where the icons are stored. These are normally kept in
4217# c:/squid/share/icons
4218#
4219#Default:
4220# icon_directory c:/squid/share/icons
4221
4222# TAG: global_internal_static
4223# This directive controls is Squid should intercept all requests for
4224# /squid-internal-static/ no matter which host the URL is requesting
4225# (default on setting), or if nothing special should be done for
4226# such URLs (off setting). The purpose of this directive is to make
4227# icons etc work better in complex cache hierarchies where it may
4228# not always be possible for all corners in the cache mesh to reach
4229# the server generating a directory listing.
4230#
4231#Default:
4232# global_internal_static on
4233
4234# TAG: short_icon_urls
4235# If this is enabled Squid will use short URLs for icons.
4236#
4237# If off the URLs for icons will always be absolute URLs
4238# including the proxy name and port.
4239#
4240#Default:
4241# short_icon_urls off
4242
4243
4244# ERROR PAGE OPTIONS
4245# -----------------------------------------------------------------------------
4246
4247# TAG: error_directory
4248# If you wish to create your own versions of the default
4249# (English) error files, either to customize them to suit your
4250# language or company copy the template English files to another
4251# directory and point this tag at them.
4252#
4253# The squid developers are interested in making squid available in
4254# a wide variety of languages. If you are making translations for a
4255# langauge that Squid does not currently provide please consider
4256# contributing your translation back to the project.
4257#
4258#Default:
4259# error_directory c:/squid/share/errors/English
4260
4261# TAG: error_map
4262# Map errors to custom messages
4263#
4264# error_map message_url http_status ...
4265#
4266# http_status ... is a list of HTTP status codes or Squid error
4267# messages.
4268#
4269# Use in accelerators to substitute the error messages returned
4270# by servers with other custom errors.
4271#
4272# error_map http://your.server/error/404.shtml 404
4273#
4274# Requests for error messages is a GET request for the configured
4275# URL with the following special headers
4276#
4277# X-Error-Status: The received HTTP status code (i.e. 404)
4278# X-Request-URI: The requested URI where the error occurred
4279#
4280# In Addition the following headers are forwarded from the client
4281# request:
4282#
4283# User-Agent, Cookie, X-Forwarded-For, Via, Authorization,
4284# Accept, Referer
4285#
4286# And the following headers from the server reply:
4287#
4288# Server, Via, Location, Content-Location
4289#
4290# The reply returned to the client will carry the original HTTP
4291# headers from the real error message, but with the reply body
4292# of the configured error message.
4293#
4294#
4295#Default:
4296# none
4297
4298# TAG: err_html_text
4299# HTML text to include in error messages. Make this a "mailto"
4300# URL to your admin address, or maybe just a link to your
4301# organizations Web page.
4302#
4303# To include this in your error messages, you must rewrite
4304# the error template files (found in the "errors" directory).
4305# Wherever you want the 'err_html_text' line to appear,
4306# insert a %L tag in the error template file.
4307#
4308#Default:
4309# none
4310
4311# TAG: deny_info
4312# Usage: deny_info err_page_name acl
4313# or deny_info http://... acl
4314# Example: deny_info ERR_CUSTOM_ACCESS_DENIED bad_guys
4315#
4316# This can be used to return a ERR_ page for requests which
4317# do not pass the 'http_access' rules. Squid remembers the last
4318# acl it evaluated in http_access, and if a 'deny_info' line exists
4319# for that ACL Squid returns a corresponding error page.
4320#
4321# The acl is typically the last acl on the http_access deny line which
4322# denied access. The exceptions to this rule are:
4323# - When Squid needs to request authentication credentials. It's then
4324# the first authentication related acl encountered
4325# - When none of the http_access lines matches. It's then the last
4326# acl processed on the last http_access line.
4327#
4328# You may use ERR_ pages that come with Squid or create your own pages
4329# and put them into the configured errors/ directory.
4330#
4331# Alternatively you can specify an error URL. The browsers will
4332# get redirected (302) to the specified URL. %s in the redirection
4333# URL will be replaced by the requested URL.
4334#
4335# Alternatively you can tell Squid to reset the TCP connection
4336# by specifying TCP_RESET.
4337#
4338#Default:
4339# none
4340
4341
4342# OPTIONS INFLUENCING REQUEST FORWARDING
4343# -----------------------------------------------------------------------------
4344
4345# TAG: nonhierarchical_direct
4346# By default, Squid will send any non-hierarchical requests
4347# (matching hierarchy_stoplist or not cacheable request type) direct
4348# to origin servers.
4349#
4350# If you set this to off, Squid will prefer to send these
4351# requests to parents.
4352#
4353# Note that in most configurations, by turning this off you will only
4354# add latency to these request without any improvement in global hit
4355# ratio.
4356#
4357# If you are inside an firewall see never_direct instead of
4358# this directive.
4359#
4360#Default:
4361# nonhierarchical_direct on
4362
4363# TAG: prefer_direct
4364# Normally Squid tries to use parents for most requests. If you for some
4365# reason like it to first try going direct and only use a parent if
4366# going direct fails set this to on.
4367#
4368# By combining nonhierarchical_direct off and prefer_direct on you
4369# can set up Squid to use a parent as a backup path if going direct
4370# fails.
4371#
4372# Note: If you want Squid to use parents for all requests see
4373# the never_direct directive. prefer_direct only modifies how Squid
4374# acts on cacheable requests.
4375#
4376#Default:
4377# prefer_direct off
4378
4379# TAG: ignore_ims_on_miss on|off
4380# This options makes Squid ignore If-Modified-Since on
4381# cache misses. This is useful while the cache is
4382# mostly empty to more quickly have the cache populated.
4383#
4384#Default:
4385# ignore_ims_on_miss off
4386
4387# TAG: always_direct
4388# Usage: always_direct allow|deny [!]aclname ...
4389#
4390# Here you can use ACL elements to specify requests which should
4391# ALWAYS be forwarded by Squid to the origin servers without using
4392# any peers. For example, to always directly forward requests for
4393# local servers ignoring any parents or siblings you may have use
4394# something like:
4395#
4396# acl local-servers dstdomain my.domain.net
4397# always_direct allow local-servers
4398#
4399# To always forward FTP requests directly, use
4400#
4401# acl FTP proto FTP
4402# always_direct allow FTP
4403#
4404# NOTE: There is a similar, but opposite option named
4405# 'never_direct'. You need to be aware that "always_direct deny
4406# foo" is NOT the same thing as "never_direct allow foo". You
4407# may need to use a deny rule to exclude a more-specific case of
4408# some other rule. Example:
4409#
4410# acl local-external dstdomain external.foo.net
4411# acl local-servers dstdomain .foo.net
4412# always_direct deny local-external
4413# always_direct allow local-servers
4414#
4415# NOTE: If your goal is to make the client forward the request
4416# directly to the origin server bypassing Squid then this needs
4417# to be done in the client configuration. Squid configuration
4418# can only tell Squid how Squid should fetch the object.
4419#
4420# NOTE: This directive is not related to caching. The replies
4421# is cached as usual even if you use always_direct. To not cache
4422# the replies see no_cache.
4423#
4424# This option replaces some v1.1 options such as local_domain
4425# and local_ip.
4426#
4427#Default:
4428# none
4429
4430# TAG: never_direct
4431# Usage: never_direct allow|deny [!]aclname ...
4432#
4433# never_direct is the opposite of always_direct. Please read
4434# the description for always_direct if you have not already.
4435#
4436# With 'never_direct' you can use ACL elements to specify
4437# requests which should NEVER be forwarded directly to origin
4438# servers. For example, to force the use of a proxy for all
4439# requests, except those in your local domain use something like:
4440#
4441# acl local-servers dstdomain .foo.net
4442# acl all src 0.0.0.0/0.0.0.0
4443# never_direct deny local-servers
4444# never_direct allow all
4445#
4446# or if Squid is inside a firewall and there are local intranet
4447# servers inside the firewall use something like:
4448#
4449# acl local-intranet dstdomain .foo.net
4450# acl local-external dstdomain external.foo.net
4451# always_direct deny local-external
4452# always_direct allow local-intranet
4453# never_direct allow all
4454#
4455# This option replaces some v1.1 options such as inside_firewall
4456# and firewall_ip.
4457#
4458#Default:
4459# none
4460
4461
4462# ADVANCED NETWORKING OPTIONS
4463# -----------------------------------------------------------------------------
4464
4465# TAG: max_filedescriptors
4466# The maximum number of filedescriptors supported.
4467#
4468# The default "0" means Squid inherits the current ulimit setting.
4469#
4470# Note: Changing this requires a restart of Squid. Also
4471# not all comm loops supports values larger than --with-maxfd.
4472#
4473#Default:
4474# max_filedescriptors 0
4475
4476# TAG: accept_filter
4477# FreeBSD:
4478#
4479# The name of an accept(2) filter to install on Squid's
4480# listen socket(s). This feature is perhaps specific to
4481# FreeBSD and requires support in the kernel.
4482#
4483# The 'httpready' filter delays delivering new connections
4484# to Squid until a full HTTP request has been received.
4485# See the accf_http(9) man page for details.
4486#
4487# The 'dataready' filter delays delivering new connections
4488# to Squid until there is some data to process.
4489# See the accf_dataready(9) man page for details.
4490#
4491# Linux:
4492#
4493# The 'data' filter delays delivering of new connections
4494# to Squid until there is some data to process by TCP_ACCEPT_DEFER.
4495# You may optionally specify a number of seconds to wait by
4496# 'data=N' where N is the number of seconds. Defaults to 30
4497# if not specified. See the tcp(7) man page for details.
4498#EXAMPLE:
4499## FreeBSD
4500#accept_filter httpready
4501## Linux
4502#accept_filter data
4503#
4504#Default:
4505# none
4506
4507# TAG: tcp_recv_bufsize (bytes)
4508# Size of receive buffer to set for TCP sockets. Probably just
4509# as easy to change your kernel's default. Set to zero to use
4510# the default buffer size.
4511#
4512#Default:
4513# tcp_recv_bufsize 0 bytes
4514
4515# TAG: incoming_rate
4516# This directive controls how aggressive Squid should accept new
4517# connections compared to processing existing connections.
4518# The lower number the more frequent Squid will look for new
4519# incoming requests.
4520#
4521#Default:
4522# incoming_rate 30
4523
4524
4525# DNS OPTIONS
4526# -----------------------------------------------------------------------------
4527
4528# TAG: check_hostnames
4529# For security and stability reasons Squid by default checks
4530# hostnames for Internet standard RFC compliance. If you do not want
4531# Squid to perform these checks then turn this directive off.
4532#
4533#Default:
4534# check_hostnames on
4535
4536# TAG: allow_underscore
4537# Underscore characters is not strictly allowed in Internet hostnames
4538# but nevertheless used by many sites. Set this to off if you want
4539# Squid to be strict about the standard.
4540# This check is performed only when check_hostnames is set to on.
4541#
4542#Default:
4543# allow_underscore on
4544
4545# TAG: cache_dns_program
4546# Note: This option is only available if Squid is rebuilt with the
4547# --disable-internal-dns option
4548#
4549# Specify the location of the executable for dnslookup process.
4550#
4551#Default:
4552# cache_dns_program c:/squid/libexec/dnsserver.exe
4553
4554# TAG: dns_children
4555# Note: This option is only available if Squid is rebuilt with the
4556# --disable-internal-dns option
4557#
4558# The number of processes spawn to service DNS name lookups.
4559# For heavily loaded caches on large servers, you should
4560# probably increase this value to at least 10. The maximum
4561# is 32. The default is 5.
4562#
4563# You must have at least one dnsserver process.
4564#
4565#Default:
4566# dns_children 5
4567
4568# TAG: dns_retransmit_interval
4569# Initial retransmit interval for DNS queries. The interval is
4570# doubled each time all configured DNS servers have been tried.
4571#
4572#
4573#Default:
4574# dns_retransmit_interval 5 seconds
4575
4576# TAG: dns_timeout
4577# DNS Query timeout. If no response is received to a DNS query
4578# within this time all DNS servers for the queried domain
4579# are assumed to be unavailable.
4580#
4581#Default:
4582# dns_timeout 2 minutes
4583
4584# TAG: dns_defnames on|off
4585# Normally the RES_DEFNAMES resolver option is disabled
4586# (see res_init(3)). This prevents caches in a hierarchy
4587# from interpreting single-component hostnames locally. To allow
4588# Squid to handle single-component names, enable this option.
4589#
4590#Default:
4591# dns_defnames off
4592
4593# TAG: dns_nameservers
4594# Use this if you want to specify a list of DNS name servers
4595# (IP addresses) to use instead of those given in your
4596# /etc/resolv.conf file.
4597# On Windows platforms, if no value is specified here or in
4598# the /etc/resolv.conf file, the list of DNS name servers are
4599# taken from the Windows registry, both static and dynamic DHCP
4600# configurations are supported.
4601#
4602# Example: dns_nameservers 10.0.0.1 192.172.0.4
4603#
4604#Default:
4605# none
4606
4607# TAG: hosts_file
4608# Location of the host-local IP name-address associations
4609# database. Most Operating Systems have such a file on different
4610# default locations:
4611# - Un*X & Linux: /etc/hosts
4612# - Windows NT/2000: %SystemRoot%\system32\drivers\etc\hosts
4613# (%SystemRoot% value install default is c:\winnt)
4614# - Windows XP/2003: %SystemRoot%\system32\drivers\etc\hosts
4615# (%SystemRoot% value install default is c:\windows)
4616# - Windows 9x/Me: %windir%\hosts
4617# (%windir% value is usually c:\windows)
4618# - Cygwin: /etc/hosts
4619#
4620# The file contains newline-separated definitions, in the
4621# form ip_address_in_dotted_form name [name ...] names are
4622# whitespace-separated. Lines beginning with an hash (#)
4623# character are comments.
4624#
4625# The file is checked at startup and upon configuration.
4626# If set to 'none', it won't be checked.
4627# If append_domain is used, that domain will be added to
4628# domain-local (i.e. not containing any dot character) host
4629# definitions.
4630#
4631#Default:
4632# none
4633
4634# TAG: dns_testnames
4635# The DNS tests exit as soon as the first site is successfully looked up
4636#
4637# This test can be disabled with the -D command line option.
4638#
4639#Default:
4640# dns_testnames netscape.com internic.net nlanr.net microsoft.com
4641
4642# TAG: append_domain
4643# Appends local domain name to hostnames without any dots in
4644# them. append_domain must begin with a period.
4645#
4646# Be warned there are now Internet names with no dots in
4647# them using only top-domain names, so setting this may
4648# cause some Internet sites to become unavailable.
4649#
4650#Example:
4651# append_domain .yourdomain.com
4652#
4653#Default:
4654# none
4655
4656# TAG: ignore_unknown_nameservers
4657# By default Squid checks that DNS responses are received
4658# from the same IP addresses they are sent to. If they
4659# don't match, Squid ignores the response and writes a warning
4660# message to cache.log. You can allow responses from unknown
4661# nameservers by setting this option to 'off'.
4662#
4663#Default:
4664# ignore_unknown_nameservers on
4665
4666# TAG: ipcache_size (number of entries)
4667# TAG: ipcache_low (percent)
4668# TAG: ipcache_high (percent)
4669# The size, low-, and high-water marks for the IP cache.
4670#
4671#Default:
4672# ipcache_size 1024
4673# ipcache_low 90
4674# ipcache_high 95
4675
4676# TAG: fqdncache_size (number of entries)
4677# Maximum number of FQDN cache entries.
4678#
4679#Default:
4680# fqdncache_size 1024
4681
4682
4683# MISCELLANEOUS
4684# -----------------------------------------------------------------------------
4685
4686# TAG: memory_pools on|off
4687# If set, Squid will keep pools of allocated (but unused) memory
4688# available for future use. If memory is a premium on your
4689# system and you believe your malloc library outperforms Squid
4690# routines, disable this.
4691#
4692#Default:
4693# memory_pools on
4694
4695# TAG: memory_pools_limit (bytes)
4696# Used only with memory_pools on:
4697# memory_pools_limit 50 MB
4698#
4699# If set to a non-zero value, Squid will keep at most the specified
4700# limit of allocated (but unused) memory in memory pools. All free()
4701# requests that exceed this limit will be handled by your malloc
4702# library. Squid does not pre-allocate any memory, just safe-keeps
4703# objects that otherwise would be free()d. Thus, it is safe to set
4704# memory_pools_limit to a reasonably high value even if your
4705# configuration will use less memory.
4706#
4707# If set to zero, Squid will keep all memory it can. That is, there
4708# will be no limit on the total amount of memory used for safe-keeping.
4709#
4710# To disable memory allocation optimization, do not set
4711# memory_pools_limit to 0. Set memory_pools to "off" instead.
4712#
4713# An overhead for maintaining memory pools is not taken into account
4714# when the limit is checked. This overhead is close to four bytes per
4715# object kept. However, pools may actually _save_ memory because of
4716# reduced memory thrashing in your malloc library.
4717#
4718#Default:
4719# memory_pools_limit 5 MB
4720
4721# TAG: forwarded_for on|off
4722# If set, Squid will include your system's IP address or name
4723# in the HTTP requests it forwards. By default it looks like
4724# this:
4725#
4726# X-Forwarded-For: 192.1.2.3
4727#
4728# If you disable this, it will appear as
4729#
4730# X-Forwarded-For: unknown
4731#
4732#Default:
4733# forwarded_for on
4734
4735# TAG: cachemgr_passwd
4736# Specify passwords for cachemgr operations.
4737#
4738# Usage: cachemgr_passwd password action action ...
4739#
4740# Some valid actions are (see cache manager menu for a full list):
4741# 5min
4742# 60min
4743# asndb
4744# authenticator
4745# cbdata
4746# client_list
4747# comm_incoming
4748# config *
4749# counters
4750# delay
4751# digest_stats
4752# dns
4753# events
4754# filedescriptors
4755# fqdncache
4756# histograms
4757# http_headers
4758# info
4759# io
4760# ipcache
4761# mem
4762# menu
4763# netdb
4764# non_peers
4765# objects
4766# offline_toggle *
4767# pconn
4768# peer_select
4769# reconfigure *
4770# redirector
4771# refresh
4772# server_list
4773# shutdown *
4774# store_digest
4775# storedir
4776# utilization
4777# via_headers
4778# vm_objects
4779#
4780# * Indicates actions which will not be performed without a
4781# valid password, others can be performed if not listed here.
4782#
4783# To disable an action, set the password to "disable".
4784# To allow performing an action without a password, set the
4785# password to "none".
4786#
4787# Use the keyword "all" to set the same password for all actions.
4788#
4789#Example:
4790# cachemgr_passwd secret shutdown
4791# cachemgr_passwd lesssssssecret info stats/objects
4792# cachemgr_passwd disable all
4793#
4794#Default:
4795# none
4796
4797# TAG: client_db on|off
4798# If you want to disable collecting per-client statistics,
4799# turn off client_db here.
4800#
4801#Default:
4802# client_db on
4803
4804# TAG: reload_into_ims on|off
4805# When you enable this option, client no-cache or ``reload''
4806# requests will be changed to If-Modified-Since requests.
4807# Doing this VIOLATES the HTTP standard. Enabling this
4808# feature could make you liable for problems which it
4809# causes.
4810#
4811# see also refresh_pattern for a more selective approach.
4812#
4813#Default:
4814# reload_into_ims off
4815
4816# TAG: maximum_single_addr_tries
4817# This sets the maximum number of connection attempts for a
4818# host that only has one address (for multiple-address hosts,
4819# each address is tried once).
4820#
4821# The default value is one attempt, the (not recommended)
4822# maximum is 255 tries. A warning message will be generated
4823# if it is set to a value greater than ten.
4824#
4825# Note: This is in addition to the request re-forwarding which
4826# takes place if Squid fails to get a satisfying response.
4827#
4828#Default:
4829# maximum_single_addr_tries 1
4830
4831# TAG: retry_on_error
4832# If set to on Squid will automatically retry requests when
4833# receiving an error response. This is mainly useful if you
4834# are in a complex cache hierarchy to work around access
4835# control errors.
4836#
4837#Default:
4838# retry_on_error off
4839
4840# TAG: as_whois_server
4841# WHOIS server to query for AS numbers. NOTE: AS numbers are
4842# queried only when Squid starts up, not for every request.
4843#
4844#Default:
4845# as_whois_server whois.ra.net
4846# as_whois_server whois.ra.net
4847
4848# TAG: offline_mode
4849# Enable this option and Squid will never try to validate cached
4850# objects.
4851#
4852#Default:
4853# offline_mode off
4854
4855# TAG: uri_whitespace
4856# What to do with requests that have whitespace characters in the
4857# URI. Options:
4858#
4859# strip: The whitespace characters are stripped out of the URL.
4860# This is the behavior recommended by RFC2396.
4861# deny: The request is denied. The user receives an "Invalid
4862# Request" message.
4863# allow: The request is allowed and the URI is not changed. The
4864# whitespace characters remain in the URI. Note the
4865# whitespace is passed to redirector processes if they
4866# are in use.
4867# encode: The request is allowed and the whitespace characters are
4868# encoded according to RFC1738. This could be considered
4869# a violation of the HTTP/1.1
4870# RFC because proxies are not allowed to rewrite URI's.
4871# chop: The request is allowed and the URI is chopped at the
4872# first whitespace. This might also be considered a
4873# violation.
4874#
4875#Default:
4876# uri_whitespace strip
4877
4878# TAG: coredump_dir
4879# By default Squid leaves core files in the directory from where
4880# it was started. If you set 'coredump_dir' to a directory
4881# that exists, Squid will chdir() to that directory at startup
4882# and coredump files will be left there.
4883#
4884#Default:
4885# coredump_dir none
4886#
4887# Leave coredumps in the first cache dir
4888coredump_dir c:/squid/var/cache
4889
4890# TAG: chroot
4891# Use this to have Squid do a chroot() while initializing. This
4892# also causes Squid to fully drop root privileges after
4893# initializing. This means, for example, if you use a HTTP
4894# port less than 1024 and try to reconfigure, you will may get an
4895# error saying that Squid can not open the port.
4896#
4897#Default:
4898# none
4899
4900# TAG: balance_on_multiple_ip
4901# Some load balancing servers based on round robin DNS have been
4902# found not to preserve user session state across requests
4903# to different IP addresses.
4904#
4905# By default Squid rotates IP's per request. By disabling
4906# this directive only connection failure triggers rotation.
4907#
4908#Default:
4909# balance_on_multiple_ip on
4910
4911# TAG: pipeline_prefetch
4912# To boost the performance of pipelined requests to closer
4913# match that of a non-proxied environment Squid can try to fetch
4914# up to two requests in parallel from a pipeline.
4915#
4916# Defaults to off for bandwidth management and access logging
4917# reasons.
4918#
4919#Default:
4920# pipeline_prefetch off
4921
4922# TAG: high_response_time_warning (msec)
4923# If the one-minute median response time exceeds this value,
4924# Squid prints a WARNING with debug level 0 to get the
4925# administrators attention. The value is in milliseconds.
4926#
4927#Default:
4928# high_response_time_warning 0
4929
4930# TAG: high_page_fault_warning
4931# If the one-minute average page fault rate exceeds this
4932# value, Squid prints a WARNING with debug level 0 to get
4933# the administrators attention. The value is in page faults
4934# per second.
4935#
4936#Default:
4937# high_page_fault_warning 0
4938
4939# TAG: high_memory_warning
4940# If the memory usage (as determined by mallinfo) exceeds
4941# this amount, Squid prints a WARNING with debug level 0 to get
4942# the administrators attention.
4943#
4944#Default:
4945# high_memory_warning 0 KB
4946
4947# TAG: sleep_after_fork (microseconds)
4948# When this is set to a non-zero value, the main Squid process
4949# sleeps the specified number of microseconds after a fork()
4950# system call. This sleep may help the situation where your
4951# system reports fork() failures due to lack of (virtual)
4952# memory. Note, however, if you have a lot of child
4953# processes, these sleep delays will add up and your
4954# Squid will not service requests for some amount of time
4955# until all the child processes have been started.
4956# On Windows value less then 1000 (1 milliseconds) are
4957# rounded to 1000.
4958#
4959#Default:
4960# sleep_after_fork 0
4961
4962# TAG: zero_buffers on|off
4963# Squid by default will zero all buffers before using or reusing them.
4964# Setting this to 'off' will result in fixed-sized temporary buffers
4965# not being zero'ed. This may give a performance boost on certain
4966# platforms but it may result in undefined behaviour at the present
4967# time.
4968#
4969#Default:
4970# zero_buffers on
4971
4972# TAG: windows_ipaddrchangemonitor on|off
4973# On Windows Squid by default will monitor IP address changes and will
4974# reconfigure itself after any detected event. This is very useful for
4975# proxies connected to internet with dial-up interfaces.
4976# In some cases (a Proxy server acting as VPN gateway is one) it could be
4977# desiderable to disable this behaviour setting this to 'off'.
4978# Note: after changing this, Squid service must be restarted.
4979#
4980#Default:
4981# windows_ipaddrchangemonitor on