Squid Server Configuration on Linux 3.0

#     WELCOME TO SQUID 2

#     ------------------

#

#     This is the default Squid configuration file. You may wish

#     to look at the Squid home page (http://www.squid-cache.org/)

#     for the FAQ and other documentation.

#

#     The default Squid config file shows what the defaults for

#     various options happen to be.  If you don't need to change the

#     default, you shouldn't uncomment the line.  Doing so may cause

#     run-time problems.  In some cases "none" refers to no default

#     setting at all, while in other cases it refers to a valid

#     option - the comments for that keyword indicate if this is the

#     case.

#

# This line was added to make rpm -U Do The Right Thing (interim fix)

# NETWORK OPTIONS

# -----------------------------------------------------------------------------

#  TAG: http_port

#     Usage:      port

#           hostname:port

#           1.2.3.4:port

#

#     The socket addresses where Squid will listen for HTTP client

#     requests.  You may specify multiple socket addresses.

#     There are three forms: port alone, hostname with port, and

#     IP address with port.  If you specify a hostname or IP

#     address, Squid binds the socket to that specific

#     address.  This replaces the old 'tcp_incoming_address'

#     option.  Most likely, you do not need to bind to a specific

#     address, so you can use the port number alone.

#

#     The default port number is 3128.

#

#     If you are running Squid in accelerator mode, you

#     probably want to listen on port 80 also, or instead.

#

#     The -a command line option will override the *first* port

#     number listed here.   That option will NOT override an IP

#     address, however.

#

#     You may specify multiple socket addresses on multiple lines.

#

#     If you run Squid on a dual-homed machine with an internal

#     and an external interface we recommend you to specify the

#     internal address:port in http_port. This way Squid will only be

#     visible on the internal address.

#

#Default:

 http_port 3128

#  TAG: https_port

#        Usage:  [ip:]port cert=certificate.pem [key=key.pem] [options...]

#

#        The socket address where Squid will listen for HTTPS client

#        requests.

#

#        This is really only useful for situations where you are running

#        squid in accelerator mode and you want to do the SSL work at the

#        accelerator level.

#

#     You may specify multiple socket addresses on multiple lines,

#     each with their own SSL certificate and/or options.

#

#     Options:

#

#        cert=    Path to SSL certificate (PEM format)

#

#        key=           Path to SSL private key file (PEM format)

#                 if not specified, the certificate file is

#                 assumed to be a combined certificate and

#                 key file

#

#        version= The version of SSL/TLS supported

#                     1 automatic (default)

#                     2 SSLv2 only

#                     3 SSLv3 only

#                     4 TLSv1 only

#

#        cipher=  Colon separated list of supported ciphers

#

#        options= Varions SSL engine options. The most important

#                 being:

#                     NO_SSLv2  Disallow the use of SSLv2

#                     NO_SSLv3  Disallow the use of SSLv3

#                     NO_TLSv1  Disallow the use of TLSv1

#                 See src/ssl_support.c or OpenSSL documentation

#                 for a more complete list.

#

#Default:

# none

#  TAG: ssl_unclean_shutdown

#     Some browsers (especially MSIE) bugs out on SSL shutdown

#     messages.

#

#Default:

# ssl_unclean_shutdown off

#  TAG: icp_port

#     The port number where Squid sends and receives ICP queries to

#     and from neighbor caches.  Default is 3130.  To disable use

#     "0".  May be overridden with -u on the command line.

#

#Default:

# icp_port 3130

#  TAG: htcp_port

# Note: This option is only available if Squid is rebuilt with the

#       --enable-htcp option

#

#     The port number where Squid sends and receives HTCP queries to

#     and from neighbor caches.  Default is 4827.  To disable use

#     "0".

#

#Default:

# htcp_port 4827

#  TAG: mcast_groups

#     This tag specifies a list of multicast groups which your server

#     should join to receive multicasted ICP queries.

#

#     NOTE!  Be very careful what you put here!  Be sure you

#     understand the difference between an ICP _query_ and an ICP

#     _reply_.  This option is to be set only if you want to RECEIVE

#     multicast queries.  Do NOT set this option to SEND multicast

#     ICP (use cache_peer for that).  ICP replies are always sent via

#     unicast, so this option does not affect whether or not you will

#     receive replies from multicast group members.

#

#     You must be very careful to NOT use a multicast address which

#     is already in use by another group of caches.

#

#     If you are unsure about multicast, please read the Multicast

#     chapter in the Squid FAQ (http://www.squid-cache.org/FAQ/).

#

#     Usage: mcast_groups 239.128.16.128 224.0.1.20

#

#     By default, Squid doesn't listen on any multicast groups.

#

#Default:

# none

#  TAG: udp_incoming_address

#  TAG: udp_outgoing_address

#     udp_incoming_address    is used for the ICP socket receiving packets

#                       from other caches.

#     udp_outgoing_address    is used for ICP packets sent out to other

#                       caches.

#

#     The default behavior is to not bind to any specific address.

#

#     A udp_incoming_address value of 0.0.0.0 indicates Squid

#     should listen for UDP messages on all available interfaces.

#

#     If udp_outgoing_address is set to 255.255.255.255 (the default)

#     it will use the same socket as udp_incoming_address. Only

#     change this if you want to have ICP queries sent using another

#     address than where this Squid listens for ICP queries from other

#     caches.

#

#     NOTE, udp_incoming_address and udp_outgoing_address can not

#     have the same value since they both use port 3130.

#

#Default:

# udp_incoming_address 0.0.0.0

# udp_outgoing_address 255.255.255.255

# OPTIONS WHICH AFFECT THE NEIGHBOR SELECTION ALGORITHM

# -----------------------------------------------------------------------------

#  TAG: cache_peer

#     To specify other caches in a hierarchy, use the format:

#

#           cache_peer hostname type http_port icp_port

#

#     For example,

#

#     #                                        proxy  icp

#     #          hostname             type     port   port  options

#     #          -------------------- -------- ----- -----  -----------

#     cache_peer parent.foo.net       parent    3128  3130  [proxy-only]

#     cache_peer sib1.foo.net         sibling   3128  3130  [proxy-only]

#     cache_peer sib2.foo.net         sibling   3128  3130  [proxy-only]

#

#           type:  either 'parent', 'sibling', or 'multicast'.

#

#     proxy_port:  The port number where the cache listens for proxy

#                requests.

#

#       icp_port:  Used for querying neighbor caches about

#                objects.  To have a non-ICP neighbor

#                specify '7' for the ICP port and make sure the

#                neighbor machine has the UDP echo port

#                enabled in its /etc/inetd.conf file.

#

#         options: proxy-only

#                weight=n

#                ttl=n

#                no-query

#                default

#                round-robin

#                multicast-responder

#                closest-only

#                no-digest

#                no-netdb-exchange

#                no-delay

#                login=user:password | PASS | *:password

#                connect-timeout=nn

#                digest-url=url

#                allow-miss

#                max-conn

#                htcp

#                carp-load-factor

#

#                use 'proxy-only' to specify objects fetched

#                from this cache should not be saved locally.

#

#                use 'weight=n' to specify a weighted parent.

#                The weight must be an integer.  The default weight

#                is 1, larger weights are favored more.

#

#                use 'ttl=n' to specify a IP multicast TTL to use

#                when sending an ICP queries to this address.

#                Only useful when sending to a multicast group.

#                Because we don't accept ICP replies from random

#                hosts, you must configure other group members as

#                peers with the 'multicast-responder' option below.

#

#                use 'no-query' to NOT send ICP queries to this

#                neighbor.

#

#                use 'default' if this is a parent cache which can

#                be used as a "last-resort." You should probably

#                only use 'default' in situations where you cannot

#                use ICP with your parent cache(s).

#

#                use 'round-robin' to define a set of parents which

#                should be used in a round-robin fashion in the

#                absence of any ICP queries.

#

#                'multicast-responder' indicates the named peer

#                is a member of a multicast group.  ICP queries will

#                not be sent directly to the peer, but ICP replies

#                will be accepted from it.

#

#                'closest-only' indicates that, for ICP_OP_MISS

#                replies, we'll only forward CLOSEST_PARENT_MISSes

#                and never FIRST_PARENT_MISSes.

#

#                use 'no-digest' to NOT request cache digests from

#                this neighbor.

#

#                'no-netdb-exchange' disables requesting ICMP

#                RTT database (NetDB) from the neighbor.

#

#                use 'no-delay' to prevent access to this neighbor

#                from influencing the delay pools.

#

#                use 'login=user:password' if this is a personal/workgroup

#                proxy and your parent requires proxy authentication.

#                Note: The string can include URL escapes (i.e. %20 for

#                spaces). This also means % must be written as %%.

#

#                use 'login=PASS' if users must authenticate against

#                the upstream proxy. This will pass the users credentials

#                as they are to the peer proxy. This only works for the

#                Basic HTTP authentication sheme. Note: To combine this

#                with proxy_auth both proxies must share the same user

#                database as HTTP only allows for one proxy login.

#                Also be warned this will expose your users proxy

#                password to the peer. USE WITH CAUTION

#

#                use 'login=*:password' to pass the username to the

#                upstream cache, but with a fixed password. This is meant

#                to be used when the peer is in another administrative

#                domain, but it is still needed to identify each user.

#                The star can optionally be followed by some extra

#                information which is added to the username. This can

#                be used to identify this proxy to the peer, similar to

#                the login=username:password option above.

#

#                use 'connect-timeout=nn' to specify a peer

#                specific connect timeout (also see the

#                peer_connect_timeout directive)

#

#                use 'digest-url=url' to tell Squid to fetch the cache

#                digest (if digests are enabled) for this host from

#                the specified URL rather than the Squid default

#                location.

#

#                use 'allow-miss' to disable Squid's use of only-if-cached

#                when forwarding requests to siblings. This is primarily

#                useful when icp_hit_stale is used by the sibling. To

#                extensive use of this option may result in forwarding

#                loops, and you should avoid having two-way peerings

#                with this option. (for example to deny peer usage on

#                requests from peer by denying cache_peer_access if the

#                source is a peer)

#

#                use 'max-conn' to limit the amount of connections Squid

#                may open to this peer.

#

#                use 'htcp' to send HTCP, instead of ICP, queries

#                to the neighbor.  You probably also want to

#                set the "icp port" to 4827 instead of 3130.

#

#                use 'carp-load-factor=f' to define a parent

#                cache as one participating in a CARP array.

#                The 'f' values for all CARP parents must add

#                up to 1.0.

#

#

#     NOTE: non-ICP/HTCP neighbors must be specified as 'parent'.

#

#Default:

# none

#  TAG: cache_peer_domain

#     Use to limit the domains for which a neighbor cache will be

#     queried.  Usage:

#

#     cache_peer_domain cache-host domain [domain ...]

#     cache_peer_domain cache-host !domain

#

#     For example, specifying

#

#           cache_peer_domain parent.foo.net    .edu

#

#     has the effect such that UDP query packets are sent to

#     'bigserver' only when the requested object exists on a

#     server in the .edu domain.  Prefixing the domainname

#     with '!' means the cache will be queried for objects

#     NOT in that domain.

#

#     NOTE: * Any number of domains may be given for a cache-host,

#             either on the same or separate lines.

#           * When multiple domains are given for a particular

#             cache-host, the first matched domain is applied.

#           * Cache hosts with no domain restrictions are queried

#             for all requests.

#           * There are no defaults.

#           * There is also a 'cache_peer_access' tag in the ACL

#             section.

#

#Default:

# none

#  TAG: neighbor_type_domain

#     usage: neighbor_type_domain neighbor parent|sibling domain domain ...

#

#     Modifying the neighbor type for specific domains is now

#     possible.  You can treat some domains differently than the the

#     default neighbor type specified on the 'cache_peer' line.

#     Normally it should only be necessary to list domains which

#     should be treated differently because the default neighbor type

#     applies for hostnames which do not match domains listed here.

#

#EXAMPLE:

#     cache_peer  parent cache.foo.org 3128 3130

#     neighbor_type_domain cache.foo.org sibling .com .net

#     neighbor_type_domain cache.foo.org sibling .au .de

#

#Default:

# none

#  TAG: icp_query_timeout     (msec)

#     Normally Squid will automatically determine an optimal ICP

#     query timeout value based on the round-trip-time of recent ICP

#     queries.  If you want to override the value determined by

#     Squid, set this 'icp_query_timeout' to a non-zero value.  This

#     value is specified in MILLISECONDS, so, to use a 2-second

#     timeout (the old default), you would write:

#

#           icp_query_timeout 2000

#

#Default:

# icp_query_timeout 0

#  TAG: maximum_icp_query_timeout   (msec)

#     Normally the ICP query timeout is determined dynamically.  But

#     sometimes it can lead to very large values (say 5 seconds).

#     Use this option to put an upper limit on the dynamic timeout

#     value.  Do NOT use this option to always use a fixed (instead

#     of a dynamic) timeout value. To set a fixed timeout see the

#     'icp_query_timeout' directive.

#

#Default:

# maximum_icp_query_timeout 2000

#  TAG: mcast_icp_query_timeout     (msec)

#     For Multicast peers, Squid regularly sends out ICP "probes" to

#     count how many other peers are listening on the given multicast

#     address.  This value specifies how long Squid should wait to

#     count all the replies.  The default is 2000 msec, or 2

#     seconds.

#

#Default:

# mcast_icp_query_timeout 2000

#  TAG: dead_peer_timeout     (seconds)

#     This controls how long Squid waits to declare a peer cache

#     as "dead."  If there are no ICP replies received in this

#     amount of time, Squid will declare the peer dead and not

#     expect to receive any further ICP replies.  However, it

#     continues to send ICP queries, and will mark the peer as

#     alive upon receipt of the first subsequent ICP reply.

#

#     This timeout also affects when Squid expects to receive ICP

#     replies from peers.  If more than 'dead_peer' seconds have

#     passed since the last ICP reply was received, Squid will not

#     expect to receive an ICP reply on the next query.  Thus, if

#     your time between requests is greater than this timeout, you

#     will see a lot of requests sent DIRECT to origin servers

#     instead of to your parents.

#

#Default:

# dead_peer_timeout 10 seconds

#  TAG: hierarchy_stoplist

#     A list of words which, if found in a URL, cause the object to

#     be handled directly by this cache.  In other words, use this

#     to not query neighbor caches for certain objects.  You may

#     list this option multiple times.

#We recommend you to use at least the following line.

hierarchy_stoplist cgi-bin ?

#  TAG: no_cache

#     A list of ACL elements which, if matched, cause the request to

#     not be satisfied from the cache and the reply to not be cached.

#     In other words, use this to force certain objects to never be cached.

#

#     You must use the word 'DENY' to indicate the ACL names which should

#     NOT be cached.

#

#We recommend you to use the following two lines.

acl QUERY urlpath_regex cgi-bin \?

no_cache deny QUERY

# OPTIONS WHICH AFFECT THE CACHE SIZE

# -----------------------------------------------------------------------------

#  TAG: cache_mem (bytes)

#     NOTE: THIS PARAMETER DOES NOT SPECIFY THE MAXIMUM PROCESS SIZE.

#     IT ONLY PLACES A LIMIT ON HOW MUCH ADDITIONAL MEMORY SQUID WILL

#     USE AS A MEMORY CACHE OF OBJECTS. SQUID USES MEMORY FOR OTHER

#     THINGS AS WELL. SEE THE SQUID FAQ SECTION 8 FOR DETAILS.

#

#     'cache_mem' specifies the ideal amount of memory to be used

#     for:

#           * In-Transit objects

#           * Hot Objects

#           * Negative-Cached objects

#

#     Data for these objects are stored in 4 KB blocks.  This

#     parameter specifies the ideal upper limit on the total size of

#     4 KB blocks allocated.  In-Transit objects take the highest

#     priority.

#

#     In-transit objects have priority over the others.  When

#     additional space is needed for incoming data, negative-cached

#     and hot objects will be released.  In other words, the

#     negative-cached and hot objects will fill up any unused space

#     not needed for in-transit objects.

#

#     If circumstances require, this limit will be exceeded.

#     Specifically, if your incoming request rate requires more than

#     'cache_mem' of memory to hold in-transit objects, Squid will

#     exceed this limit to satisfy the new requests.  When the load

#     decreases, blocks will be freed until the high-water mark is

#     reached.  Thereafter, blocks will be used to store hot

#     objects.

#

#Default:

# cache_mem 8 MB

#  TAG: cache_swap_low  (percent, 0-100)

#  TAG: cache_swap_high (percent, 0-100)

#

#     The low- and high-water marks for cache object replacement.

#     Replacement begins when the swap (disk) usage is above the

#     low-water mark and attempts to maintain utilization near the

#     low-water mark.  As swap utilization gets close to high-water

#     mark object eviction becomes more aggressive.  If utilization is

#     close to the low-water mark less replacement is done each time.

#

#     Defaults are 90% and 95%. If you have a large cache, 5% could be

#     hundreds of MB. If this is the case you may wish to set these

#     numbers closer together.

#

#Default:

# cache_swap_low 90

# cache_swap_high 95

#  TAG: maximum_object_size   (bytes)

#     Objects larger than this size will NOT be saved on disk.  The

#     value is specified in kilobytes, and the default is 4MB.  If

#     you wish to get a high BYTES hit ratio, you should probably

#     increase this (one 32 MB object hit counts for 3200 10KB

#     hits).  If you wish to increase speed more than your want to

#     save bandwidth you should leave this low.

#

#     NOTE: if using the LFUDA replacement policy you should increase

#     this value to maximize the byte hit rate improvement of LFUDA!

#     See replacement_policy below for a discussion of this policy.

#

#Default:

# maximum_object_size 4096 KB

#  TAG: minimum_object_size   (bytes)

#     Objects smaller than this size will NOT be saved on disk.  The

#     value is specified in kilobytes, and the default is 0 KB, which

#     means there is no minimum.

#

#Default:

# minimum_object_size 0 KB

#  TAG: maximum_object_size_in_memory     (bytes)

#        Objects greater than this size will not be attempted to kept in

#        the memory cache. This should be set high enough to keep objects

#        accessed frequently in memory to improve performance whilst low

#        enough to keep larger objects from hoarding cache_mem .

#

#Default:

# maximum_object_size_in_memory 8 KB

#  TAG: ipcache_size    (number of entries)

#  TAG: ipcache_low     (percent)

#  TAG: ipcache_high    (percent)

#     The size, low-, and high-water marks for the IP cache.

#

#Default:

# ipcache_size 1024

# ipcache_low 90

# ipcache_high 95

#  TAG: fqdncache_size  (number of entries)

#     Maximum number of FQDN cache entries.

#

#Default:

# fqdncache_size 1024

#  TAG: cache_replacement_policy

#     The cache replacement policy parameter determines which

#     objects are evicted (replaced) when disk space is needed.

#

#         lru       : Squid's original list based LRU policy

#         heap GDSF : Greedy-Dual Size Frequency

#         heap LFUDA: Least Frequently Used with Dynamic Aging

#         heap LRU  : LRU policy implemented using a heap

#

#     Applies to any cache_dir lines listed below this.

#

#     The LRU policies keeps recently referenced objects.

#

#     The heap GDSF policy optimizes object hit rate by keeping smaller

#     popular objects in cache so it has a better chance of getting a

#     hit.  It achieves a lower byte hit rate than LFUDA though since

#     it evicts larger (possibly popular) objects.

#

#     The heap LFUDA policy keeps popular objects in cache regardless of

#     their size and thus optimizes byte hit rate at the expense of

#     hit rate since one large, popular object will prevent many

#     smaller, slightly less popular objects from being cached.

#

#     Both policies utilize a dynamic aging mechanism that prevents

#     cache pollution that can otherwise occur with frequency-based

#     replacement policies.

#

#     NOTE: if using the LFUDA replacement policy you should increase

#     the value of maximum_object_size above its default of 4096 KB to

#     to maximize the potential byte hit rate improvement of LFUDA.

#

#     For more information about the GDSF and LFUDA cache replacement

#     policies see http://www.hpl.hp.com/techreports/1999/HPL-1999-69.html

#     and http://fog.hpl.external.hp.com/techreports/98/HPL-98-173.html.

#

#Default:

# cache_replacement_policy lru

#  TAG: memory_replacement_policy

#     The memory replacement policy parameter determines which

#     objects are purged from memory when memory space is needed.

#

#     See cache_replacement_policy for details.

#

#Default:

# memory_replacement_policy lru

# LOGFILE PATHNAMES AND CACHE DIRECTORIES

# -----------------------------------------------------------------------------

#  TAG: cache_dir

#     Usage:

#

#     cache_dir Type Directory-Name Fs-specific-data [options]

#

#     You can specify multiple cache_dir lines to spread the

#     cache among different disk partitions.

#

#     Type specifies the kind of storage system to use. Only "ufs"

#     is built by default. To eanble any of the other storage systems

#     see the --enable-storeio configure option.

#

#     'Directory' is a top-level directory where cache swap

#     files will be stored.  If you want to use an entire disk

#     for caching, this can be the mount-point directory.

#     The directory must exist and be writable by the Squid

#     process.  Squid will NOT create this directory for you.

#

#     The ufs store type:

#

#     "ufs" is the old well-known Squid storage format that has always

#     been there.

#

#     cache_dir ufs Directory-Name Mbytes L1 L2 [options]

#

#     'Mbytes' is the amount of disk space (MB) to use under this

#     directory.  The default is 100 MB.  Change this to suit your

#     configuration.  Do NOT put the size of your disk drive here.

#     Instead, if you want Squid to use the entire disk drive,

#     subtract 20% and use that value.

#

#     'Level-1' is the number of first-level subdirectories which

#     will be created under the 'Directory'.  The default is 16.

#

#     'Level-2' is the number of second-level subdirectories which

#     will be created under each first-level directory.  The default

#     is 256.

#

#     The aufs store type:

#

#     "aufs" uses the same storage format as "ufs", utilizing

#     POSIX-threads to avoid blocking the main Squid process on

#     disk-I/O. This was formerly known in Squid as async-io.

#

#     cache_dir aufs Directory-Name Mbytes L1 L2 [options]

#

#     see argument descriptions under ufs above

#

#     The diskd store type:

#

#     "diskd" uses the same storage format as "ufs", utilizing a

#     separate process to avoid blocking the main Squid process on

#     disk-I/O.

#

#     cache_dir diskd Directory-Name Mbytes L1 L2 [options] [Q1=n] [Q2=n]

#

#     see argument descriptions under ufs above

#

#     Q1 specifies the number of unacknowledged I/O requests when Squid

#     stops opening new files. If this many messages are in the queues,

#     Squid won't open new files. Default is 64

#

#     Q2 specifies the number of unacknowledged messages when Squid

#     starts blocking.  If this many messages are in the queues,

#     Squid blocks until it receives some replies. Default is 72

#

#     When Q1 < Q2 (the default), the cache directory is optimized

#     for lower response time at the expense of a decrease in hit

#     ratio.  If Q1 > Q2, the cache directory is optimized for

#     higher hit ratio at the expense of an increase in response

#     time.

#

#     The coss store type:

#

#     block-size=n defines the "block size" for COSS cache_dir's.

#     Squid uses file numbers as block numbers.  Since file numbers

#     are limited to 24 bits, the block size determines the maximum

#     size of the COSS partition.  The default is 512 bytes, which

#     leads to a maximum cache_dir size of 512<<24, or 8 GB.  Note

#     you should not change the coss block size after Squid

#     has written some objects to the cache_dir.

#

#     Common options:

#

#     read-only, this cache_dir is read only.

#

#     max-size=n, refers to the max object size this storedir supports.

#     It is used to initially choose the storedir to dump the object.

#     Note: To make optimal use of the max-size limits you should order

#     the cache_dir lines with the smallest max-size value first and the

#     ones with no max-size specification last.

#

#     Note that for coss, max-size must be less than COSS_MEMBUF_SZ

#     (hard coded at 1 MB).

#

#Default:

# cache_dir ufs /var/spool/squid 100 16 256

#  TAG: cache_access_log

#     Logs the client request activity.  Contains an entry for

#     every HTTP and ICP queries received. To disable, enter "none".

#

#Default:

 cache_access_log /var/log/squid/access08022011.log

#  TAG: cache_log

#     Cache logging file. This is where general information about

#     your cache's behavior goes. You can increase the amount of data

#     logged to this file with the "debug_options" tag below.

#

#Default:

# cache_log /var/log/squid/cache.log

#  TAG: cache_store_log

#     Logs the activities of the storage manager.  Shows which

#     objects are ejected from the cache, and which objects are

#     saved and for how long.  To disable, enter "none". There are

#     not really utilities to analyze this data, so you can safely

#     disable it.

#

#Default:

# cache_store_log /var/log/squid/store.log

#  TAG: cache_swap_log

#     Location for the cache "swap.state" file. This log file holds

#     the metadata of objects saved on disk.  It is used to rebuild

#     the cache during startup.  Normally this file resides in each

#     'cache_dir' directory, but you may specify an alternate

#     pathname here.  Note you must give a full filename, not just

#     a directory. Since this is the index for the whole object

#     list you CANNOT periodically rotate it!

#

#     If %s can be used in the file name it will be replaced with a

#     a representation of the cache_dir name where each / is replaced

#     with '.'. This is needed to allow adding/removing cache_dir

#     lines when cache_swap_log is being used.

#

#     If have more than one 'cache_dir', and %s is not used in the name

#     these swap logs will have names such as:

#

#           cache_swap_log.00

#           cache_swap_log.01

#           cache_swap_log.02

#

#     The numbered extension (which is added automatically)

#     corresponds to the order of the 'cache_dir' lines in this

#     configuration file.  If you change the order of the 'cache_dir'

#     lines in this file, these log files will NOT correspond to

#     the correct 'cache_dir' entry (unless you manually rename

#     them).  We recommend you do NOT use this option.  It is

#     better to keep these log files in each 'cache_dir' directory.

#

#Default:

# none

#  TAG: emulate_httpd_log     on|off

#     The Cache can emulate the log file format which many 'httpd'

#     programs use.  To disable/enable this emulation, set

#     emulate_httpd_log to 'off' or 'on'.  The default

#     is to use the native log format since it includes useful

#     information Squid-specific log analyzers use.

#

#Default:

# emulate_httpd_log off

#  TAG: log_ip_on_direct      on|off

#     Log the destination IP address in the hierarchy log tag when going

#     direct. Earlier Squid versions logged the hostname here. If you

#     prefer the old way set this to off.

#

#Default:

# log_ip_on_direct on

#  TAG: mime_table

#     Pathname to Squid's MIME table. You shouldn't need to change

#     this, but the default file contains examples and formatting

#     information if you do.

#

#Default:

# mime_table /etc/squid/mime.conf

#  TAG: log_mime_hdrs   on|off

#     The Cache can record both the request and the response MIME

#     headers for each HTTP transaction.  The headers are encoded

#     safely and will appear as two bracketed fields at the end of

#     the access log (for either the native or httpd-emulated log

#     formats).  To enable this logging set log_mime_hdrs to 'on'.

#

#Default:

# log_mime_hdrs off

#  TAG: useragent_log

#     Squid will write the User-Agent field from HTTP requests

#     to the filename specified here.  By default useragent_log

#     is disabled.

#

#Default:

# none

#  TAG: referer_log

#     Squid will write the Referer field from HTTP requests to the

#     filename specified here.  By default referer_log is disabled.

#

#Default:

# none

#  TAG: pid_filename

#     A filename to write the process-id to.  To disable, enter "none".

#

#Default:

# pid_filename /var/run/squid.pid

#  TAG: debug_options

#     Logging options are set as section,level where each source file

#     is assigned a unique section.  Lower levels result in less

#     output,  Full debugging (level 9) can result in a very large

#     log file, so be careful.  The magic word "ALL" sets debugging

#     levels for all sections.  We recommend normally running with

#     "ALL,1".

#

#Default:

# debug_options ALL,1

#  TAG: log_fqdn  on|off

#     Turn this on if you wish to log fully qualified domain names

#     in the access.log. To do this Squid does a DNS lookup of all

#     IP's connecting to it. This can (in some situations) increase

#     latency, which makes your cache seem slower for interactive

#     browsing.

#

#Default:

# log_fqdn off

#  TAG: client_netmask

#     A netmask for client addresses in logfiles and cachemgr output.

#     Change this to protect the privacy of your cache clients.

#     A netmask of 255.255.255.0 will log all IP's in that range with

#     the last digit set to '0'.

#

#Default:

# client_netmask 255.255.255.255

# OPTIONS FOR EXTERNAL SUPPORT PROGRAMS

# -----------------------------------------------------------------------------

#  TAG: ftp_user

#     If you want the anonymous login password to be more informative

#     (and enable the use of picky ftp servers), set this to something

#     reasonable for your domain, like wwwuser@somewhere.net

#

#     The reason why this is domainless by default is the

#     request can be made on the behalf of a user in any domain,

#     depending on how the cache is used.

#     Some ftp server also validate the email address is valid

#     (for example perl.com).

#

#Default:

# ftp_user Squid@

#  TAG: ftp_list_width

#     Sets the width of ftp listings. This should be set to fit in

#     the width of a standard browser. Setting this too small

#     can cut off long filenames when browsing ftp sites.

#

#Default:

# ftp_list_width 32

#  TAG: ftp_passive

#     If your firewall does not allow Squid to use passive

#     connections, turn off this option.

#

#Default:

# ftp_passive on

#  TAG: ftp_sanitycheck

#     For security and data integrity reasons Squid by default performs

#     sanity checks of the addresses of FTP data connections ensure the

#     data connection is to the requested server. If you need to allow

#     FTP connections to servers using another IP address for the data

#     connection turn this off.

#

#Default:

# ftp_sanitycheck on

#  TAG: ftp_telnet_protocol

#     The FTP protocol is officially defined to use the telnet protocol

#     as transport channel for the control connection. However, many

#     implemenations are broken and does not respect this aspect of

#     the FTP protocol.

#

#     If you have trouble accessing files with ASCII code 255 in the

#     path or similar problems involving this ASCII code you can

#     try setting this directive to off. If that helps, report to the

#     operator of the FTP server in question that their FTP server

#     is broken and does not follow the FTP standard.

#

#Default:

# ftp_telnet_protocol on

#  TAG: cache_dns_program

# Note: This option is only available if Squid is rebuilt with the

#       --disable-internal-dns option

#

#     Specify the location of the executable for dnslookup process.

#

#Default:

# cache_dns_program /usr/lib/squid/dnsserver

#  TAG: dns_children

# Note: This option is only available if Squid is rebuilt with the

#       --disable-internal-dns option

#

#     The number of processes spawn to service DNS name lookups.

#     For heavily loaded caches on large servers, you should

#     probably increase this value to at least 10.  The maximum

#     is 32.  The default is 5.

#

#     You must have at least one dnsserver process.

#

#Default:

# dns_children 5

#  TAG: dns_retransmit_interval

#     Initial retransmit interval for DNS queries. The interval is

#     doubled each time all configured DNS servers have been tried.

#

#

#Default:

# dns_retransmit_interval 5 seconds

#  TAG: dns_timeout

#     DNS Query timeout. If no response is received to a DNS query

#     within this time all DNS servers for the queried domain

#     are assumed to be unavailable.

#

#Default:

# dns_timeout 2 minutes

#  TAG: dns_defnames    on|off

# Note: This option is only available if Squid is rebuilt with the

#       --disable-internal-dns option

#

#     Normally the 'dnsserver' disables the RES_DEFNAMES resolver

#     option (see res_init(3)).  This prevents caches in a hierarchy

#     from interpreting single-component hostnames locally.  To allow

#     dnsserver to handle single-component names, enable this

#     option.

#

#Default:

# dns_defnames off

#  TAG: dns_nameservers

#     Use this if you want to specify a list of DNS name servers

#     (IP addresses) to use instead of those given in your

#     /etc/resolv.conf file.

#     On Windows platforms, if no value is specified here or in

#     the /etc/resolv.conf file, the list of DNS name servers are

#     taken from the Windows registry, both static and dynamic DHCP

#     configurations are supported.

#

#     Example: dns_nameservers 10.0.0.1 192.172.0.4

#

#Default:

# none

#  TAG: hosts_file

#     Location of the host-local IP name-address associations

#     database.  Most Operating Systems have such a file: under

#     Un*X it's by default in /etc/hosts.  MS-Windows NT/2000 places

#     it in %SystemRoot%(by default

#     c:\winnt)\system32\drivers\etc\hosts, while Windows 9x/ME

#     places it in %windir%(usually c:\windows)\hosts

#

#     The file contains newline-separated definitions, in the

#     form ip_address_in_dotted_form name [name ...] names are

#     whitespace-separated.  lines beginnng with an hash (#)

#     character are comments.

#

#     The file is checked at startup and upon configuration.  If

#     set to 'none', it won't be checked.  If append_domain is

#     used, that domain will be added to domain-local (i.e. not

#     containing any dot character) host definitions.

#

#Default:

# hosts_file /etc/hosts

#  TAG: diskd_program

#     Specify the location of the diskd executable.

#     Note that this is only useful if you have compiled in

#     diskd as one of the store io modules.

#

#Default:

# diskd_program /usr/lib/squid/diskd

#  TAG: unlinkd_program

#     Specify the location of the executable for file deletion process.

#

#Default:

# unlinkd_program /usr/lib/squid/unlinkd

#  TAG: pinger_program

# Note: This option is only available if Squid is rebuilt with the

#       --enable-icmp option

#

#     Specify the location of the executable for the pinger process.

#

#Default:

# pinger_program /usr/lib/squid/pinger

#  TAG: redirect_program

#     Specify the location of the executable for the URL redirector.

#     Since they can perform almost any function there isn't one included.

#     See the FAQ (section 15) for information on how to write one.

#     By default, a redirector is not used.

#

#Default:

# none

#  TAG: redirect_children

#     The number of redirector processes to spawn. If you start

#     too few Squid will have to wait for them to process a backlog of

#     URLs, slowing it down. If you start too many they will use RAM

#     and other system resources.

#

#Default:

# redirect_children 5

#  TAG: redirect_rewrites_host_header

#     By default Squid rewrites any Host: header in redirected

#     requests.  If you are running an accelerator this may

#     not be a wanted effect of a redirector.

#

#Default:

# redirect_rewrites_host_header on

#  TAG: redirector_access

#     If defined, this access list specifies which requests are

#     sent to the redirector processes.  By default all requests

#     are sent.

#

#Default:

# none

#  TAG: auth_param

#     This is used to define parameters for the various authentication

#     schemes supported by Squid.

#

#     format: auth_param scheme parameter [setting]

#

#     The order in which authentication schemes are presented to the client is

#     dependant on the order the scheme first appears in config file. IE

#     has a bug (it's not rfc 2617 compliant) in that it will use the basic

#     scheme if basic is the first entry presented, even if more secure

#     schemes are presented. For now use the order in the recommended

#     settings section below. If other browsers have difficulties (don't

#     recognise the schemes offered even if you are using basic) either

#     put basic first, or disable the other schemes (by commenting out their

#     program entry).

#

#     Once an authentication scheme is fully configured, it can only be

#     shutdown by shutting squid down and restarting. Changes can be made on

#     the fly and activated with a reconfigure. I.E. You can change to a

#     different helper, but not unconfigure the helper completely.

#

#     Please note that while this directive defines how Squid processes

#     authentication it does not automatically activate authentication.

#     To use authenticaiton you must in addition make use of acls based

#     on login name in http_access (proxy_auth, proxy_auth_regex or

#     external with %LOGIN used in the format tag). The browser will be

#     challenged for authentication on the first such acl encountered

#     in http_access processing and will also be rechallenged for new

#     login credentials if the request is being denied by a proxy_auth

#     type acl.

#

#     === Parameters for the basic scheme follow. ===

#

#     "program" cmdline

#     Specify the command for the external authenticator.  Such a program

#     reads a line containing "username password" and replies "OK" or

#     "ERR" in an endless loop.

#

#     By default, the basic authentication sheme is not used unless a

#     program is specified.

#

#     If you want to use the traditional proxy authentication, jump over to

#     the helpers/basic_auth/NCSA directory and type:

#           % make

#           % make install

#

#     Then, set this line to something like

#

#     auth_param basic program /usr/libexec/ncsa_auth /usr/etc/passwd

#

#     "children" numberofchildren

#     The number of authenticator processes to spawn.

#     If you start too few Squid will have to wait for them to process a

#     backlog of usercode/password verifications, slowing it down. When

#     password verifications are done via a (slow) network you are likely to

#     need lots of authenticator processes.

#     auth_param basic children 5

#

#     "realm" realmstring

#     Specifies the realm name which is to be reported to the client for

#     the basic proxy authentication scheme (part of the text the user

#     will see when prompted their username and password).

#     auth_param basic realm Squid proxy-caching web server

#

#     "credentialsttl" timetolive

#     Specifies how long squid assumes an externally validated

#     username:password pair is valid for - in other words how often the

#     helper program is called for that user. Set this low to force

#     revalidation with short lived passwords.  Note that setting this high

#     does not impact your susceptability to replay attacks unless you are

#     using an one-time password system (such as SecureID). If you are using

#     such a system, you will be vulnerable to replay attacks unless you

#     also use the max_user_ip ACL in an http_access rule.

#     auth_param basic credentialsttl 2 hours

#

#     "casesensitive" on|off

#     Specifies if usernames are case sensitive. Most user databases are

#     case insensitive allowing the same username to be spelled using both

#     lower and upper case letters, but some are case sensitive. This

#     makes a big difference for user_max_ip ACL processing and similar.

#     auth_param basic casesensitive off

#

#     === Parameters for the digest scheme follow ===

#

#     "program" cmdline

#     Specify the command for the external authenticator.  Such a program

#     reads a line containing "username":"realm" and replies with the

#     appropriate H(A1) value base64 encoded or ERR if the user (or his H(A1)

#     hash) does not exists.  See rfc 2616 for the definition of H(A1).

#

#     By default, the digest authentication scheme is not used unless a

#     program is specified.

#

#     If you want to use a digest authenticator, jump over to the

#     helpers/digest_auth/ directory and choose the authenticator to use.

#     It it's directory type

#           % make

#             % make install

#

#     Then, set this line to something like

#

#     auth_param digest program /usr/libexec/digest_auth_pw /usr/etc/digpass

#

#

#     "children" numberofchildren

#     The number of authenticator processes to spawn (no default). If you

#     start too few Squid will have to wait for them to process a backlog of

#     H(A1) calculations, slowing it down.  When the H(A1) calculations are

#     done via a (slow) network you are likely to need lots of authenticator

#     processes.

#     auth_param digest children 5

#

#     "realm" realmstring

#     Specifies the realm name which is to be reported to the client for the

#     digest proxy authentication scheme (part of the text the user will see

#     when prompted their username and password).

#     auth_param digest realm Squid proxy-caching web server

#

#     "nonce_garbage_interval" timeinterval

#     Specifies the interval that nonces that have been issued to clients are

#     checked for validity.

#     auth_param digest nonce_garbage_interval 5 minutes

#

#     "nonce_max_duration" timeinterval

#     Specifies the maximum length of time a given nonce will be valid for.

#     auth_param digest nonce_max_duration 30 minutes

#

#     "nonce_max_count" number

#     Specifies the maximum number of times a given nonce can be used.

#     auth_param digest nonce_max_count 50

#

#     "nonce_strictness" on|off

#     Determines if squid requires strict increment-by-1 behaviour for nonce

#     counts, or just incrementing (off - for use when useragents generate

#     nonce counts that occasionally miss 1 (ie, 1,2,4,6)).

#     auth_param digest nonce_strictness off

#

#     "check_nonce_count" on|off

#     This directive if set to off can disable the nonce count check

#     completely to work around buggy digest qop implementations in certain

#     mainstream browser versions. Default on to check the nonce count to

#     protect from authentication replay attacks.

#     auth_param digest check_nonce_count on

#

#     "post_workaround" on|off

#     This is a workaround to certain buggy browsers who sends an incorrect

#     request digest in POST requests when reusing the same nonce as aquired

#           earlier in response to a GET request.

#     auth_param digest post_workaround off

#

#     === NTLM scheme options follow ===

#

#     "program" cmdline

#     Specify the command for the external ntlm authenticator. Such a

#     program participates in the NTLMSSP exchanges between Squid and the

#     client and reads commands according to the Squid ntlmssp helper

#     protocol. See helpers/ntlm_auth/ for details. Recommended ntlm

#     authenticator is ntlm_auth from Samba-3.X, but a number of other

#     ntlm authenticators is available.

#

#     By default, the ntlm authentication scheme is not used unless a

#     program is specified.

#

#     auth_param ntlm program /path/to/samba/bin/ntlm_auth --helper-protocol=squid-2.5-ntlmssp

#

#     "children" numberofchildren

#     The number of authenticator processes to spawn (no default). If you

#     start too few Squid will have to wait for them to process a backlog

#     of credential verifications, slowing it down. When crendential

#     verifications are done via a (slow) network you are likely to need

#     lots of authenticator processes.

#     auth_param ntlm children 5

#

#     "max_challenge_reuses" number

#     The maximum number of times a challenge given by a ntlm authentication

#     helper can be reused. Increasing this number increases your exposure

#     to replay attacks on your network. 0 (the default) means use the

#     challenge is used only once. See also the max_ntlm_challenge_lifetime

#     directive if enabling challenge reuses.

#     auth_param ntlm max_challenge_reuses 0

#

#     "max_challenge_lifetime" timespan

#     The maximum time period a ntlm challenge is reused over. The

#     actual period will be the minimum of this time AND the number of

#     reused challenges.

#     auth_param ntlm max_challenge_lifetime 2 minutes

#

#     "use_ntlm_negotiate" on|off

#     Enables support for NTLM NEGOTIATE packet exchanges with the helper.

#     The configured ntlm authenticator must be able to handle NTLM

#     NEGOTIATE packet. See the authenticator programs documentation if

#     unsure. ntlm_auth from Samba-3.0.2 or later supports the use of this

#     option.

#     The NEGOTIATE packet is required to support NTLMv2 and a

#     number of other negotiable NTLMSSP options, and also makes it

#     more likely the negotiation is successful. Enabling this parameter

#     will also solve problems encountered when NT domain policies

#     restrict users to access only certain workstations. When this is off,

#     all users must be allowed to log on the proxy servers too, or they'll

#     get "invalid workstation" errors - and access denied - when trying to

#     use Squid's services.

#     Use of ntlm NEGOTIATE is incompatible with challenge reuse, so

#     enabling this parameter will OVERRIDE the max_challenge_reuses and

#     max_challenge_lifetime parameters and set them to 0.

#     auth_param ntlm use_ntlm_negotiate off

#

#Recommended minimum configuration:

#auth_param digest program <uncomment and complete this line>

#auth_param digest children 5

#auth_param digest realm Squid proxy-caching web server

#auth_param digest nonce_garbage_interval 5 minutes

#auth_param digest nonce_max_duration 30 minutes

#auth_param digest nonce_max_count 50

#auth_param ntlm program <uncomment and complete this line to activate>

#auth_param ntlm children 5

#auth_param ntlm max_challenge_reuses 0

#auth_param ntlm max_challenge_lifetime 2 minutes

#auth_param ntlm use_ntlm_negotiate off

#auth_param basic program <uncomment and complete this line>

auth_param basic children 5

auth_param basic realm Squid proxy-caching web server

auth_param basic credentialsttl 2 hours

auth_param basic casesensitive off

auth_param basic program /usr/lib/squid/ncsa_auth /etc/squid/squid_passwd

#  TAG: authenticate_cache_garbage_interval

#     The time period between garbage collection across the username cache.

#     This is a tradeoff between memory utilisation (long intervals - say

#     2 days) and CPU (short intervals - say 1 minute). Only change if you

#     have good reason to.

#

#Default:

# authenticate_cache_garbage_interval 1 hour

#  TAG: authenticate_ttl

#     The time a user & their credentials stay in the logged in user cache

#     since their last request. When the garbage interval passes, all user

#     credentials that have passed their TTL are removed from memory.

#

#Default:

# authenticate_ttl 1 hour

#  TAG: authenticate_ip_ttl

#     If you use proxy authentication and the 'max_user_ip' ACL, this

#     directive controls how long Squid remembers the IP addresses

#     associated with each user.  Use a small value (e.g., 60 seconds) if

#     your users might change addresses quickly, as is the case with

#     dialups. You might be safe using a larger value (e.g., 2 hours) in a

#     corporate LAN environment with relatively static address assignments.

#

#Default:

# authenticate_ip_ttl 0 seconds

#  TAG: external_acl_type

#     This option defines external acl classes using a helper program to

#     look up the status

#

#       external_acl_type name [options] FORMAT.. /path/to/helper [helper arguments..]

#

#     Options:

#

#       ttl=n           TTL in seconds for cached results (defaults to 3600

#                 for 1 hour)

#       negative_ttl=n

#                 TTL for cached negative lookups (default same

#                 as ttl)

#       children=n      Concurrency level / number of processes spawn

#                 to service external acl lookups of this type.

#                 Note: see compatibility note below

#       cache=n   result cache size, 0 is unbounded (default)

#       protocol=3.0    Use URL-escaped strings instead of quoting

#

#     FORMAT specifications

#

#       %LOGIN    Authenticated user login name

#       %IDENT    Ident user name

#       %SRC            Client IP

#       %DST            Requested host

#       %PROTO    Requested protocol

#       %PORT           Requested port

#       %METHOD   Request method

#       %{Header} HTTP request header

#       %{Hdr:member}   HTTP request header list member

#       %{Hdr:;member}

#                 HTTP request header list member using ; as

#                 list separator. ; can be any non-alphanumeric

#                 character.

#

#     In addition, any string specified in the referencing acl will

#     also be included in the helper request line, after the specified

#     formats (see the "acl external" directive)

#

#     The helper receives lines per the above format specification,

#     and returns lines starting with OK or ERR indicating the validity

#     of the request and optionally followed by additional keywords with

#     more details.

#

#     General result syntax:

#

#       OK/ERR keyword=value ...

#

#     Defined keywords:

#

#       user=           The users name (login)

#       error=    Error description (only defined for ERR results)

#

#     Keyword values need to be enclosed in quotes if they may contain

#     whitespace, or the whitespace escaped using \. Any quotes or \

#     characters within the keyword value must be \ escaped.

#

#     If protocol=3.0 then URL escaping of the strings is used instead

#     of the above described quoting format.

#

#     Compatibility Note: The children= option was named concurrency= in

#     Squid-2.5.STABLE3 and earlier and such syntax is still accepted to

#     keep compatibility within the Squid-2.5 release. However, the meaning

#     of concurrency= option has changed in Squid-3 and the old syntax of

#     the directive is therefore depreated from Squid-2.5.STABLE4 and later.

#     If you want to be able to easily downgrade to earlier Squid-2.5

#     releases you may want to continue using the old name, if not

#     please use the new name.

#

#Default:

# none

# OPTIONS FOR TUNING THE CACHE

# -----------------------------------------------------------------------------

#  TAG: wais_relay_host

#  TAG: wais_relay_port

#     Relay WAIS request to host (1st arg) at port (2 arg).

#

#Default:

# wais_relay_port 0

#  TAG: request_header_max_size     (KB)

#     This specifies the maximum size for HTTP headers in a request.

#     Request headers are usually relatively small (about 512 bytes).

#     Placing a limit on the request header size will catch certain

#     bugs (for example with persistent connections) and possibly

#     buffer-overflow or denial-of-service attacks.

#

#Default:

# request_header_max_size 10 KB

#  TAG: request_body_max_size (KB)

#     This specifies the maximum size for an HTTP request body.

#     In other words, the maximum size of a PUT/POST request.

#     A user who attempts to send a request with a body larger

#     than this limit receives an "Invalid Request" error message.

#     If you set this parameter to a zero (the default), there will

#     be no limit imposed.

#

#Default:

# request_body_max_size 0 KB

#  TAG: refresh_pattern

#     usage: refresh_pattern [-i] regex min percent max [options]

#

#     By default, regular expressions are CASE-SENSITIVE.  To make

#     them case-insensitive, use the -i option.

#

#     'Min' is the time (in minutes) an object without an explicit

#     expiry time should be considered fresh. The recommended

#     value is 0, any higher values may cause dynamic applications

#     to be erroneously cached unless the application designer

#     has taken the appropriate actions.

#

#     'Percent' is a percentage of the objects age (time since last

#     modification age) an object without explicit expiry time

#     will be considered fresh.

#

#     'Max' is an upper limit on how long objects without an explicit

#     expiry time will be considered fresh.

#

#     options: override-expire

#           override-lastmod

#           reload-into-ims

#           ignore-reload

#

#           override-expire enforces min age even if the server

#           sent a Expires: header. Doing this VIOLATES the HTTP

#           standard.  Enabling this feature could make you liable

#           for problems which it causes.

#

#           override-lastmod enforces min age even on objects

#           that were modified recently.

#

#           reload-into-ims changes client no-cache or ``reload''

#           to If-Modified-Since requests. Doing this VIOLATES the

#           HTTP standard. Enabling this feature could make you

#           liable for problems which it causes.

#

#           ignore-reload ignores a client no-cache or ``reload''

#           header. Doing this VIOLATES the HTTP standard. Enabling

#           this feature could make you liable for problems which

#           it causes.

#

#     Basically a cached object is:

#

#           FRESH if expires < now, else STALE

#           STALE if age > max

#           FRESH if lm-factor < percent, else STALE

#           FRESH if age < min

#           else STALE

#

#     The refresh_pattern lines are checked in the order listed here.

#     The first entry which matches is used.  If none of the entries

#     match the default will be used.

#

#     Note, you must uncomment all the default lines if you want

#     to change one. The default setting is only active if none is

#     used.

#

#Suggested default:

refresh_pattern ^ftp:         1440  20%   10080

refresh_pattern ^gopher:      1440  0%    1440

refresh_pattern .       0     20%   4320

#  TAG: quick_abort_min (KB)

#  TAG: quick_abort_max (KB)

#  TAG: quick_abort_pct (percent)

#     The cache by default continues downloading aborted requests

#     which are almost completed (less than 16 KB remaining). This

#     may be undesirable on slow (e.g. SLIP) links and/or very busy

#     caches.  Impatient users may tie up file descriptors and

#     bandwidth by repeatedly requesting and immediately aborting

#     downloads.

#

#     When the user aborts a request, Squid will check the

#     quick_abort values to the amount of data transfered until

#     then.

#

#     If the transfer has less than 'quick_abort_min' KB remaining,

#     it will finish the retrieval.

#

#     If the transfer has more than 'quick_abort_max' KB remaining,

#     it will abort the retrieval.

#

#     If more than 'quick_abort_pct' of the transfer has completed,

#     it will finish the retrieval.

#

#     If you do not want any retrieval to continue after the client

#     has aborted, set both 'quick_abort_min' and 'quick_abort_max'

#     to '0 KB'.

#

#     If you want retrievals to always continue if they are being

#     cached set 'quick_abort_min' to '-1 KB'.

#

#Default:

# quick_abort_min 16 KB

# quick_abort_max 16 KB

# quick_abort_pct 95

#  TAG: negative_ttl    time-units

#     Time-to-Live (TTL) for failed requests.  Certain types of

#     failures (such as "connection refused" and "404 Not Found") are

#     negatively-cached for a configurable amount of time.  The

#     default is 5 minutes.  Note that this is different from

#     negative caching of DNS lookups.

#

#Default:

# negative_ttl 5 minutes

#  TAG: positive_dns_ttl      time-units

#     Upper limit on how long Squid will cache positive DNS responses.

#     Default is 6 hours (360 minutes). This directive must be set

#     larger than negative_dns_ttl.

#

#Default:

# positive_dns_ttl 6 hours

#  TAG: negative_dns_ttl      time-units

#     Time-to-Live (TTL) for negative caching of failed DNS lookups.

#     This also makes sets the lower cache limit on positive lookups.

#     Minimum value is 1 second, and it is not recommendable to go

#     much below 10 seconds.

#

#Default:

# negative_dns_ttl 1 minute

#  TAG: range_offset_limit    (bytes)

#     Sets a upper limit on how far into the the file a Range request

#     may be to cause Squid to prefetch the whole file. If beyond this

#     limit Squid forwards the Range request as it is and the result

#     is NOT cached.

#

#     This is to stop a far ahead range request (lets say start at 17MB)

#     from making Squid fetch the whole object up to that point before

#     sending anything to the client.

#

#     A value of -1 causes Squid to always fetch the object from the

#     beginning so it may cache the result. (2.0 style)

#

#     A value of 0 causes Squid to never fetch more than the

#     client requested. (default)

#

#Default:

# range_offset_limit 0 KB

# TIMEOUTS

# -----------------------------------------------------------------------------

#  TAG: forward_timeout time-units

#     This parameter specifies how long Squid should at most attempt in

#     finding a forwarding path for the request before giving up.

#

#Default:

# forward_timeout 4 minutes

#  TAG: connect_timeout time-units

#     This parameter specifies how long to wait for the TCP connect to

#     the requested server or peer to complete before Squid should

#     attempt to find another path where to forward the request.

#

#Default:

# connect_timeout 1 minute

#  TAG: peer_connect_timeout  time-units

#     This parameter specifies how long to wait for a pending TCP

#     connection to a peer cache.  The default is 30 seconds.   You

#     may also set different timeout values for individual neighbors

#     with the 'connect-timeout' option on a 'cache_peer' line.

#

#Default:

# peer_connect_timeout 30 seconds

#  TAG: read_timeout    time-units

#     The read_timeout is applied on server-side connections.  After

#     each successful read(), the timeout will be extended by this

#     amount.  If no data is read again after this amount of time,

#     the request is aborted and logged with ERR_READ_TIMEOUT.  The

#     default is 15 minutes.

#

#Default:

# read_timeout 15 minutes

#  TAG: request_timeout

#     How long to wait for an HTTP request after initial

#     connection establishment.

#

#Default:

# request_timeout 5 minutes

#  TAG: persistent_request_timeout

#     How long to wait for the next HTTP request on a persistent

#     connection after the previous request completes.

#

#Default:

# persistent_request_timeout 1 minute

#  TAG: client_lifetime time-units

#     The maximum amount of time a client (browser) is allowed to

#     remain connected to the cache process.  This protects the Cache

#     from having a lot of sockets (and hence file descriptors) tied up

#     in a CLOSE_WAIT state from remote clients that go away without

#     properly shutting down (either because of a network failure or

#     because of a poor client implementation).  The default is one

#     day, 1440 minutes.

#

#     NOTE:  The default value is intended to be much larger than any

#     client would ever need to be connected to your cache.  You

#     should probably change client_lifetime only as a last resort.

#     If you seem to have many client connections tying up

#     filedescriptors, we recommend first tuning the read_timeout,

#     request_timeout, persistent_request_timeout and quick_abort values.

#

#Default:

# client_lifetime 1 day

#  TAG: half_closed_clients

#     Some clients may shutdown the sending side of their TCP

#     connections, while leaving their receiving sides open.      Sometimes,

#     Squid can not tell the difference between a half-closed and a

#     fully-closed TCP connection.  By default, half-closed client

#     connections are kept open until a read(2) or write(2) on the

#     socket returns an error.  Change this option to 'off' and Squid

#     will immediately close client connections when read(2) returns

#     "no more data to read."

#

#Default:

# half_closed_clients on

#  TAG: pconn_timeout

#     Timeout for idle persistent connections to servers and other

#     proxies.

#

#Default:

# pconn_timeout 120 seconds

#  TAG: ident_timeout

#     Maximum time to wait for IDENT lookups to complete.

#

#     If this is too high, and you enabled IDENT lookups from untrusted

#     users, you might be susceptible to denial-of-service by having

#     many ident requests going at once.

#

#Default:

# ident_timeout 10 seconds

#  TAG: shutdown_lifetime     time-units

#     When SIGTERM or SIGHUP is received, the cache is put into

#     "shutdown pending" mode until all active sockets are closed.

#     This value is the lifetime to set for all open descriptors

#     during shutdown mode.  Any active clients after this many

#     seconds will receive a 'timeout' message.

#

#Default:

# shutdown_lifetime 30 seconds

# ACCESS CONTROLS

# -----------------------------------------------------------------------------

#  TAG: acl

#     Defining an Access List

#

#     acl aclname acltype string1 ...

#     acl aclname acltype "file" ...

#

#     when using "file", the file should contain one item per line

#

#     acltype is one of the types described below

#

#     By default, regular expressions are CASE-SENSITIVE.  To make

#     them case-insensitive, use the -i option.

#

#     acl aclname src      ip-address/netmask ... (clients IP address)

#     acl aclname src      addr1-addr2/netmask ... (range of addresses)

#     acl aclname dst      ip-address/netmask ... (URL host's IP address)

#     acl aclname myip     ip-address/netmask ... (local socket IP address)

#

#     acl aclname srcdomain   .foo.com ...    # reverse lookup, client IP

#     acl aclname dstdomain   .foo.com ...    # Destination server from URL

#     acl aclname srcdom_regex [-i] xxx ...   # regex matching client name

#     acl aclname dstdom_regex [-i] xxx ...   # regex matching server

#       # For dstdomain and dstdom_regex  a reverse lookup is tried if a IP

#       # based URL is used. The name "none" is used if the reverse lookup

#       # fails.

#

#     acl aclname time     [day-abbrevs]  [h1:m1-h2:m2]

#         day-abbrevs:

#           S - Sunday

#           M - Monday

#           T - Tuesday

#           W - Wednesday

#           H - Thursday

#           F - Friday

#           A - Saturday

#         h1:m1 must be less than h2:m2

#     acl aclname url_regex [-i] ^http:// ...   # regex matching on whole URL

#     acl aclname urlpath_regex [-i] \.gif$ ... # regex matching on URL path

#     acl aclname urllogin [-i] [^a-zA-Z0-9] ...      # regex matching on URL login field

#     acl aclname port     80 70 21 ...

#     acl aclname port     0-1024 ...           # ranges allowed

#     acl aclname myport   3128 ...       # (local socket TCP port)

#     acl aclname proto    HTTP FTP ...

#     acl aclname method   GET POST ...

#     acl aclname browser  [-i] regexp ...

#       # pattern match on User-Agent header (see also req_header below)

#        acl aclname referer_regex  [-i] regexp ...

#          # pattern match on Referer header

#          # Referer is highly unreliable, so use with care

#     acl aclname ident    username ...

#     acl aclname ident_regex [-i] pattern ...

#       # string match on ident output.

#       # use REQUIRED to accept any non-null ident.

#     acl aclname src_as   number ...

#     acl aclname dst_as   number ...

#       # Except for access control, AS numbers can be used for

#       # routing of requests to specific caches. Here's an

#       # example for routing all requests for AS#1241 and only

#       # those to mycache.mydomain.net:

#       # acl asexample dst_as 1241

#       # cache_peer_access mycache.mydomain.net allow asexample

#       # cache_peer_access mycache_mydomain.net deny all

#

#     acl aclname proxy_auth username ...

#     acl aclname proxy_auth_regex [-i] pattern ...

#       # list of valid usernames

#       # use REQUIRED to accept any valid username.

#       #

#       # NOTE: when a Proxy-Authentication header is sent but it is not

#       # needed during ACL checking the username is NOT logged

#       # in access.log.

#       #

#       # NOTE: proxy_auth requires a EXTERNAL authentication program

#       # to check username/password combinations (see

#       # auth_param directive).

#       #

#       # WARNING: proxy_auth can't be used in a transparent proxy. It

#       # collides with any authentication done by origin servers. It may

#       # seem like it works at first, but it doesn't.

#

#     acl aclname snmp_community string ...

#       # A community string to limit access to your SNMP Agent

#       # Example:

#       #

#       #   acl snmppublic snmp_community public

#

#     acl aclname maxconn number

#       # This will be matched when the client's IP address has

#       # more than <number> HTTP connections established.

#

#     acl aclname max_user_ip [-s] number

#       # This will be matched when the user attempts to log in from more

#       # than <number> different ip addresses. The authenticate_ip_ttl

#       # parameter controls the timeout on the ip entries.

#       # If -s is specified the limit is strict, denying browsing

#       # from any further IP addresses until the ttl has expired. Without

#      # -s Squid will just annoy the user by "randomly" denying requests.

#       # (the counter is reset each time the limit is reached and a

#       # request is denied)

#       # NOTE: in acceleration mode or where there is mesh of child proxies,

#       # clients may appear to come from multiple addresses if they are

#       # going through proxy farms, so a limit of 1 may cause user problems.

#

#     acl aclname req_mime_type mime-type1 ...

#       # regex match against the mime type of the request generated

#       # by the client. Can be used to detect file upload or some

#       # types HTTP tunelling requests.

#       # NOTE: This does NOT match the reply. You cannot use this

#       # to match the returned file type.

#

#     acl aclname req_header header-name [-i] any\.regex\.here

#       # regex match against any of the known request headers.  May be

#       # thought of as a superset of "browser", "referer" and "mime-type"

#       # acls.

#

#     acl aclname rep_mime_type mime-type1 ...

#       # regex match against the mime type of the reply recieved by

#       # squid. Can be used to detect file download or some

#       # types HTTP tunelling requests.

#       # NOTE: This has no effect in http_access rules. It only has

#       # effect in rules that affect the reply data stream such as

#       # http_reply_access.

#

#     acl aclname rep_header header-name [-i] any\.regex\.here

#       # regex match against any of the known response headers.

#       # Example:

#       #

#       # acl many_spaces rep_header Content-Disposition -i [[:space:]]{3,}

#

#     acl acl_name external class_name [arguments...]

#       # external ACL lookup via a helper class defined by the

#       # external_acl_type directive.

#

#Examples:

#acl myexample dst_as 1241

#acl password proxy_auth REQUIRED

#acl fileupload req_mime_type -i ^multipart/form-data$

#acl javascript rep_mime_type -i ^application/x-javascript$

#

#Recommended minimum configuration:

acl all src 0.0.0.0/0.0.0.0

acl manager proto cache_object

acl localhost src 127.0.0.0/255.255.255.255

acl to_localhost dst 127.0.0.0/8

acl SSL_ports port 443 563

acl Safe_ports port 80        # http

acl Safe_ports port 21        # ftp

acl Safe_ports port 443 563   # https, snews

acl Safe_ports port 70        # gopher

acl Safe_ports port 210       # wais

acl Safe_ports port 1025-65535      # unregistered ports

acl Safe_ports port 280       # http-mgmt

acl Safe_ports port 488       # gss-http

acl Safe_ports port 591       # filemaker

acl Safe_ports port 777       # multiling http

acl CONNECT method CONNECT

acl local src 192.168.0.2/24

acl GoodSites dstdomain "/usr/local/etc/allowed-sites.squid

acl BadSites  dstdomain "/usr/local/etc/restricted-sites.squid

acl ncsa_users proxy_auth REQUIRED

#http_access allow ncsa_users

#http_access deny ncsa_users BadSites

#  TAG: http_access

#     Allowing or Denying access based on defined access lists

#

#     Access to the HTTP port:

#     http_access allow|deny [!]aclname ...

#

#     NOTE on default values:

#

#     If there are no "access" lines present, the default is to deny

#     the request.

#

#     If none of the "access" lines cause a match, the default is the

#     opposite of the last line in the list.  If the last line was

#     deny, the default is allow.  Conversely, if the last line

#     is allow, the default will be deny.  For these reasons, it is a

#     good idea to have an "deny all" or "allow all" entry at the end

#     of your access lists to avoid potential confusion.

#

#Default:

# http_access deny all

#

#Recommended minimum configuration:

#

# Only allow cachemgr access from localhost

http_access allow  GoodSites

http_access deny   BadSites

http_access allow manager localhost

http_access allow local

http_access deny manager

#http_access allow ncsa_users GoodSites

#http_access deny ncsa_users BadSites

# Deny requests to unknown ports

http_access deny !Safe_ports

# Deny CONNECT to other than SSL ports

http_access deny CONNECT !SSL_ports

#

# We strongly recommend the following be uncommented to protect innocent

# web applications running on the proxy server who think the only

# one who can access services on "localhost" is a local user

#http_access deny to_localhost

#

# INSERT YOUR OWN RULE(S) HERE TO ALLOW ACCESS FROM YOUR CLIENTS

# Example rule allowing access from your local networks. Adapt

# to list your (internal) IP networks from where browsing should

# be allowed

#acl our_networks src 192.168.0.3/24 192.168.0.254/24

#http_access allow our_networks

# And finally deny all other access to this proxy

http_access allow localhost

http_access deny all

#  TAG: http_reply_access

#        Allow replies to client requests. This is complementary to http_access.

#

#        http_reply_access allow|deny [!] aclname ...

#

#        NOTE: if there are no access lines present, the default is to allow

#     all replies

#

#        If none of the access lines cause a match the opposite of the

#        last line will apply. Thus it is good practice to end the rules

#        with an "allow all" or "deny all" entry.

#

#Default:

# http_reply_access allow all

#

#Recommended minimum configuration:

#

# Insert your own rules here.

#

#

# and finally allow by default

http_reply_access allow all

#  TAG: icp_access

#     Allowing or Denying access to the ICP port based on defined

#     access lists

#

#     icp_access  allow|deny [!]aclname ...

#

#     See http_access for details

#

#Default:

# icp_access deny all

#

#Allow ICP queries from everyone

icp_access allow all

#  TAG: miss_access

#     Use to force your neighbors to use you as a sibling instead of

#     a parent.  For example:

#

#           acl localclients src 172.16.0.0/16

#           miss_access allow localclients

#           miss_access deny  !localclients

#

#     This means only your local clients are allowed to fetch

#     MISSES and all other clients can only fetch HITS.

#

#     By default, allow all clients who passed the http_access rules

#     to fetch MISSES from us.

#

#Default setting:

# miss_access allow all

#  TAG: cache_peer_access

#     Similar to 'cache_peer_domain' but provides more flexibility by

#     using ACL elements.

#

#     cache_peer_access cache-host allow|deny [!]aclname ...

#

#     The syntax is identical to 'http_access' and the other lists of

#     ACL elements.  See the comments for 'http_access' below, or

#     the Squid FAQ (http://www.squid-cache.org/FAQ/FAQ-10.html).

#

#Default:

# none

#  TAG: ident_lookup_access

#     A list of ACL elements which, if matched, cause an ident

#     (RFC 931) lookup to be performed for this request.  For

#Default:

# none

#  TAG: ident_lookup_access

#     A list of ACL elements which, if matched, cause an ident

#     (RFC 931) lookup to be performed for this request.  For

#     example, you might choose to always perform ident lookups

#     for your main multi-user Unix boxes, but not for your Macs

#     and PCs.  By default, ident lookups are not performed for

#     any requests.

#

#     To enable ident lookups for specific client addresses, you

#     can follow this example:

#

#     acl ident_aware_hosts src 198.168.1.0/255.255.255.0

#     ident_lookup_access allow ident_aware_hosts

#     ident_lookup_access deny all

#

#     Only src type ACL checks are fully supported.  A src_domain

#     ACL might work at times, but it will not always provide

#     the correct result.

#

#Default:

# ident_lookup_access deny all

#  TAG: tcp_outgoing_tos

#     Allows you to select a TOS/Diffserv value to mark outgoing

#     connections with, based on the username or source address

#     making the request.

#

#     tcp_outgoing_tos ds-field [!]aclname ...

#

#     Example where normal_service_net uses the TOS value 0x00

#     and normal_service_net uses 0x20

#

#     acl normal_service_net src 10.0.0.0/255.255.255.0

#     acl good_service_net src 10.0.1.0/255.255.255.0

#     tcp_outgoing_tos 0x00 normal_service_net 0x00

#     tcp_outgoing_tos 0x20 good_service_net

#

#     TOS/DSCP values really only have local significance - so you should

#     know what you're specifying. For more, see RFC 2474

#

#     The TOS/DSCP byte must be exactly that - a byte, value  0 - 255, or

#     "default" to use whatever default your host has.

#

#     Processing proceeds in the order specified, and stops at first fully

#     matching line.

#

#Default:

# none

#  TAG: tcp_outgoing_address

#     Allows you to map requests to different outgoing IP addresses

#     based on the username or sourceaddress of the user making

#     the request.

#

#     tcp_outgoing_address ipaddr [[!]aclname] ...

#

#     Example where requests from 10.0.0.0/24 will be forwareded

#     with source address 10.1.0.1, 10.0.2.0/24 forwarded with

#     source address 10.1.0.2 and the rest will be forwarded with

#     source address 10.1.0.3.

#

#     acl normal_service_net src 10.0.0.0/255.255.255.0

#     acl good_service_net src 10.0.1.0/255.255.255.0

#     tcp_outgoing_address 10.0.0.1 normal_service_net

#     tcp_outgoing_address 10.0.0.2 good_service_net

#     tcp_outgoing_address 10.0.0.3

#

#     Processing proceeds in the order specified, and stops at first fully

#     matching line.

#

#Default:

# none

#  TAG: reply_header_max_size (KB)

#     This specifies the maximum size for HTTP headers in a reply.

#     Reply headers are usually relatively small (about 512 bytes).

#     Placing a limit on the reply header size will catch certain

#     bugs (for example with persistent connections) and possibly

#     buffer-overflow or denial-of-service attacks.

#

#Default:

# reply_header_max_size 20 KB

#  TAG: reply_body_max_size   bytes allow|deny acl acl...

#        This option specifies the maximum size of a reply body in bytes.

#     It can be used to prevent users from downloading very large files,

#     such as MP3's and movies. When the reply headers are recieved,

#     the reply_body_max_size lines are processed, and the first line with

#     a result of "allow" is used as the maximum body size for this reply.

#     This size is checked twice. First when we get the reply headers,

#     we check the content-length value.  If the content length value exists

#     and is larger than the allowed size, the request is denied and the

#     user receives an error message that says "the request or reply

#     is too large." If there is no content-length, and the reply

#     size exceeds this limit, the client's connection is just closed

#     and they will receive a partial reply.

#

#     WARNING: downstream caches probably can not detect a partial reply

#     if there is no content-length header, so they will cache

#     partial responses and give them out as hits.  You should NOT

#     use this option if you have downstream caches.

#

#     If you set this parameter to zero (the default), there will be

#     no limit imposed.

#

#Default:

# reply_body_max_size 0 allow all

# ADMINISTRATIVE PARAMETERS

# -----------------------------------------------------------------------------

#  TAG: cache_mgr

#     Email-address of local cache manager who will receive

#     mail if the cache dies.  The default is "root".

#cache_mgr root

#

#Default:

# cache_mgr root

#  TAG: cache_effective_user

#     If you start Squid as root, it will change its effective/real

#     UID/GID to the user specified below.  The default is to change

#     to UID to "squid".  If you define cache_effective_user, but not

#     cache_effective_group, Squid sets the GID to the effective

#     user's default group ID (taken from the password file) and

#     supplementary group list from the from groups membership of

#     cache_effective_user.

#cache_effective_user squid

#

#Default:

# cache_effective_user squid

#  TAG: cache_effective_group

#     If you want Squid to run with a specific GID regardless of

#     the group memberships of the effective user then set this

#     to the group (or GID) you want Squid to run as. When set

#     all other group privileges of the effective user is ignored

#     and only this GID is effective. If Squid is not started as

#     root the user starting Squid must be member of the specified

#     group.

#cache_effective_group squid

#

#Default:

# cache_effective_group squid

#  TAG: visible_hostname

#     If you want to present a special hostname in error messages, etc,

#     define this.  Otherwise, the return value of gethostname()

#     will be used. If you have multiple caches in a cluster and

#     get errors about IP-forwarding you must set them to have individual

#     names with this setting.

#

#Default:

# none

#  TAG: unique_hostname

#     If you want to have multiple machines with the same

#     'visible_hostname' you must give each machine a different

#     'unique_hostname' so forwarding loops can be detected.

#

#Default:

# none

#  TAG: hostname_aliases

#     A list of other DNS names your cache has.

#

#Default:

# none

# OPTIONS FOR THE CACHE REGISTRATION SERVICE

# -----------------------------------------------------------------------------

#

#     This section contains parameters for the (optional) cache

#     announcement service.  This service is provided to help

#     cache administrators locate one another in order to join or

#     create cache hierarchies.

#

#     An 'announcement' message is sent (via UDP) to the registration

#     service by Squid.  By default, the announcement message is NOT

#     SENT unless you enable it with 'announce_period' below.

#

#     The announcement message includes your hostname, plus the

#     following information from this configuration file:

#

#           http_port

#           icp_port

#           cache_mgr

#

#     All current information is processed regularly and made

#     available on the Web at http://www.ircache.net/Cache/Tracker/.

#  TAG: announce_period

#     This is how frequently to send cache announcements.  The

#     default is `0' which disables sending the announcement

#     messages.

#

#     To enable announcing your cache, just uncomment the line

#     below.

#

#Default:

# announce_period 0

#

#To enable announcing your cache, just uncomment the line below.

#announce_period 1 day

#  TAG: announce_host

#  TAG: announce_file

#  TAG: announce_port

#     announce_host and announce_port set the hostname and port

#     number where the registration message will be sent.

#

#     Hostname will default to 'tracker.ircache.net' and port will

#     default default to 3131.  If the 'filename' argument is given,

#     the contents of that file will be included in the announce

#     message.

#

#Default:

# announce_host tracker.ircache.net

# announce_port 3131

# HTTPD-ACCELERATOR OPTIONS

# -----------------------------------------------------------------------------

#  TAG: httpd_accel_host

#  TAG: httpd_accel_port

#     If you want to run Squid as an httpd accelerator, define the

#     host name and port number where the real HTTP server is.

#

#     If you want IP based virtual host support specify the

#     hostname as "virtual". This will make Squid use the IP address

#     where it accepted the request as hostname in the URL.

#

#     If you want virtual port support specify the port as "0".

#

#     NOTE: enabling httpd_accel_host disables proxy-caching and

#     ICP.  If you want these features enabled also, set

#     the 'httpd_accel_with_proxy' option.

#

#Default:

# httpd_accel_port 80

#  TAG: httpd_accel_single_host     on|off

#     If you are running Squid as an accelerator and have a single backend

#     server set this to on. This causes Squid to forward the request

#     to this server, regardles of what any redirectors or Host headers

#     say.

#

#     Leave this at off if you have multiple backend servers, and use a

#     redirector (or host table or private DNS) to map the requests to the

#     appropriate backend servers. Note that the mapping needs to be a

#     1-1 mapping between requested and backend (from redirector) domain

#     names or caching will fail, as cacing is performed using the

#     URL returned from the redirector.

#

#     See also redirect_rewrites_host_header.

#

#Default:

# httpd_accel_single_host off

#  TAG: httpd_accel_with_proxy      on|off

#     If you want to use Squid as both a local httpd accelerator

#     and as a proxy, change this to 'on'. Note however your

#     proxy users may have trouble to reach the accelerated domains

#     unless their browsers are configured not to use this proxy for

#     those domains (for example via the no_proxy browser configuration

#     setting)

#

#Default:

# httpd_accel_with_proxy off

#  TAG: httpd_accel_uses_host_header      on|off

#     HTTP/1.1 requests include a Host: header which is basically the

#     hostname from the URL.  The Host: header is used for domain based

#     virutal hosts. If your accelerator needs to provide domain based

#     virtual hosts on the same IP address you will need to turn this

#     on.

#

#     Note Squid does NOT check the value of the Host header matches

#     any of your accelerated server, so it may open a big security hole

#     unless you take care to set up access controls proper.  We recommend

#     this option remain disabled unless you are sure of what you

#     are doing.

#

#     However, you will need to enable this option if you run Squid

#     as a transparent proxy.  Otherwise, virtual servers which

#     require the Host: header will not be properly cached.

#

#Default:

# httpd_accel_uses_host_header off

# MISCELLANEOUS

# -----------------------------------------------------------------------------

#  TAG: dns_testnames

#     The DNS tests exit as soon as the first site is successfully looked up

#

#     This test can be disabled with the -D command line option.

#

#Default:

# dns_testnames netscape.com internic.net nlanr.net microsoft.com

#  TAG: logfile_rotate

#     Specifies the number of logfile rotations to make when you

#     type 'squid -k rotate'.  The default is 10, which will rotate

#     with extensions 0 through 9.  Setting logfile_rotate to 0 will

#     disable the rotation, but the logfiles are still closed and

#     re-opened.  This will enable you to rename the logfiles

#     yourself just before sending the rotate signal.

#

#     Note, the 'squid -k rotate' command normally sends a USR1

#     signal to the running squid process.  In certain situations

#     (e.g. on Linux with Async I/O), USR1 is used for other

#     purposes, so -k rotate uses another signal.  It is best to get

#     in the habit of using 'squid -k rotate' instead of 'kill -USR1

#     <pid>'.

#    

#logfile_rotate 0

#

#Default:

# logfile_rotate 0

#  TAG: append_domain

#     Appends local domain name to hostnames without any dots in

#     them.  append_domain must begin with a period.

#

#     Be warned there are now Internet names with no dots in

#     them using only top-domain names, so setting this may

#     cause some Internet sites to become unavailable.

#

#Example:

# append_domain .yourdomain.com

#

#Default:

# none

#  TAG: tcp_recv_bufsize      (bytes)

#     Size of receive buffer to set for TCP sockets.  Probably just

#     as easy to change your kernel's default.  Set to zero to use

#     the default buffer size.

#

#Default:

# tcp_recv_bufsize 0 bytes

#  TAG: err_html_text

#     HTML text to include in error messages.  Make this a "mailto"

#     URL to your admin address, or maybe just a link to your

#     organizations Web page.

#

#     To include this in your error messages, you must rewrite

#     the error template files (found in the "errors" directory).

#     Wherever you want the 'err_html_text' line to appear,

#     insert a %L tag in the error template file.

#

#Default:

# none

#  TAG: deny_info

#     Usage:   deny_info err_page_name acl

#     or       deny_info http://... acl

#     Example: deny_info ERR_CUSTOM_ACCESS_DENIED bad_guys

#

#     This can be used to return a ERR_ page for requests which

#     do not pass the 'http_access' rules.  A single ACL will cause

#     the http_access check to fail.  If a 'deny_info' line exists

#     for that ACL Squid returns a corresponding error page.

#

#     You may use ERR_ pages that come with Squid or create your own pages

#     and put them into the configured errors/ directory.

#

#     Alternatively you can specify an error URL. The browsers will

#     get redirected (302) to the specified URL. %s in the redirection

#     URL will be replaced by the requested URL.

#

#     Alternatively you can tell Squid to reset the TCP connection

#     by specifying TCP_RESET.

#

#Default:

# none

#  TAG: memory_pools    on|off

#     If set, Squid will keep pools of allocated (but unused) memory

#     available for future use.  If memory is a premium on your

#     system and you believe your malloc library outperforms Squid

#     routines, disable this.

#

#Default:

# memory_pools on

#  TAG: memory_pools_limit    (bytes)

#     Used only with memory_pools on:

#     memory_pools_limit 50 MB

#

#     If set to a non-zero value, Squid will keep at most the specified

#     limit of allocated (but unused) memory in memory pools. All free()

#     requests that exceed this limit will be handled by your malloc

#     library. Squid does not pre-allocate any memory, just safe-keeps

#     objects that otherwise would be free()d. Thus, it is safe to set

#     memory_pools_limit to a reasonably high value even if your

#     configuration will use less memory.

#

#     If not set (default) or set to zero, Squid will keep all memory it

#     can. That is, there will be no limit on the total amount of memory

#     used for safe-keeping.

#

#     To disable memory allocation optimization, do not set

#     memory_pools_limit to 0. Set memory_pools to "off" instead.

#

#     An overhead for maintaining memory pools is not taken into account

#     when the limit is checked. This overhead is close to four bytes per

#     object kept. However, pools may actually _save_ memory because of

#     reduced memory thrashing in your malloc library.

#

#Default:

# none

#  TAG: forwarded_for   on|off

#     If set, Squid will include your system's IP address or name

#     in the HTTP requests it forwards.  By default it looks like

#     this:

#

#           X-Forwarded-For: 192.1.2.3

#

#     If you disable this, it will appear as

#

#           X-Forwarded-For: unknown

#

#Default:

# forwarded_for on

#  TAG: log_icp_queries on|off

#     If set, ICP queries are logged to access.log. You may wish

#     do disable this if your ICP load is VERY high to speed things

#     up or to simplify log analysis.

#

#Default:

# log_icp_queries on

#  TAG: icp_hit_stale   on|off

#     If you want to return ICP_HIT for stale cache objects, set this

#     option to 'on'.  If you have sibling relationships with caches

#     in other administrative domains, this should be 'off'.  If you only

#     have sibling relationships with caches under your control,

#     it is probably okay to set this to 'on'.

#     If set to 'on', your siblings should use the option "allow-miss"

#     on their cache_peer lines for connecting to you.

#

#Default:

# icp_hit_stale off

#  TAG: minimum_direct_hops

#     If using the ICMP pinging stuff, do direct fetches for sites

#     which are no more than this many hops away.

#

#Default:

# minimum_direct_hops 4

#  TAG: minimum_direct_rtt

#     If using the ICMP pinging stuff, do direct fetches for sites

#     which are no more than this many rtt milliseconds away.

#

#Default:

# minimum_direct_rtt 400

#  TAG: cachemgr_passwd

#     Specify passwords for cachemgr operations.

#

#     Usage: cachemgr_passwd password action action ...

#

#     Some valid actions are (see cache manager menu for a full list):

#           5min

#           60min

#           asndb

#           authenticator

#           cbdata

#           client_list

#           comm_incoming

#           config *

#           counters

#           delay

#           digest_stats

#           dns

#           events

#           filedescriptors

#           fqdncache

#           histograms

#           http_headers

#           info

#           io

#           ipcache

#           mem

#           menu

#           netdb

#           non_peers

#           objects

#           offline_toggle *

#           pconn

#           peer_select

#           redirector

#           refresh

#           server_list

#           shutdown *

#           store_digest

#           storedir

#           utilization

#           via_headers

#           vm_objects

#

#     * Indicates actions which will not be performed without a

#       valid password, others can be performed if not listed here.

#

#     To disable an action, set the password to "disable".

#     To allow performing an action without a password, set the

#     password to "none".

#

#     Use the keyword "all" to set the same password for all actions.

#

#Example:

# cachemgr_passwd secret shutdown

# cachemgr_passwd lesssssssecret info stats/objects

# cachemgr_passwd disable all

#

#Default:

# none

#  TAG: store_avg_object_size (kbytes)

#     Average object size, used to estimate number of objects your

#     cache can hold.  See doc/Release-Notes-1.1.txt.  The default is

#     13 KB.

#

#Default:

# store_avg_object_size 13 KB

#  TAG: store_objects_per_bucket

#     Target number of objects per bucket in the store hash table.

#     Lowering this value increases the total number of buckets and

#     also the storage maintenance rate.  The default is 50.

#

#Default:

# store_objects_per_bucket 20

#  TAG: client_db on|off

#     If you want to disable collecting per-client statistics,

#     turn off client_db here.

#

#Default:

# client_db on

#  TAG: netdb_low

#  TAG: netdb_high

#     The low and high water marks for the ICMP measurement

#     database.  These are counts, not percents.  The defaults are

#     900 and 1000.  When the high water mark is reached, database

#     entries will be deleted until the low mark is reached.

#

#Default:

# netdb_low 900

# netdb_high 1000

#  TAG: netdb_ping_period

#     The minimum period for measuring a site.  There will be at

#     least this much delay between successive pings to the same

#     network.  The default is five minutes.

#

#Default:

# netdb_ping_period 5 minutes

#  TAG: query_icmp      on|off

#     If you want to ask your peers to include ICMP data in their ICP

#     replies, enable this option.

#

#     If your peer has configured Squid (during compilation) with

#     '--enable-icmp' that peer will send ICMP pings to origin server

#     sites of the URLs it receives.  If you enable this option the

#     ICP replies from that peer will include the ICMP data (if available).

#     Then, when choosing a parent cache, Squid will choose the parent with

#     the minimal RTT to the origin server.  When this happens, the

#     hierarchy field of the access.log will be

#     "CLOSEST_PARENT_MISS".  This option is off by default.

#

#Default:

# query_icmp off

#  TAG: test_reachability     on|off

#     When this is 'on', ICP MISS replies will be ICP_MISS_NOFETCH

#     instead of ICP_MISS if the target host is NOT in the ICMP

#     database, or has a zero RTT.

#

#Default:

# test_reachability off

#  TAG: buffered_logs   on|off

#     cache.log log file is written with stdio functions, and as such

#     it can be buffered or unbuffered. By default it will be unbuffered.

#     Buffering it can speed up the writing slightly (though you are

#     unlikely to need to worry unless you run with tons of debugging

#     enabled in which case performance will suffer badly anyway..).

#

#Default:

# buffered_logs off

#  TAG: reload_into_ims on|off

#     When you enable this option, client no-cache or ``reload''

#     requests will be changed to If-Modified-Since requests.

#     Doing this VIOLATES the HTTP standard.  Enabling this

#     feature could make you liable for problems which it

#     causes.

#

#     see also refresh_pattern for a more selective approach.

#

#Default:

# reload_into_ims off

#  TAG: always_direct

#     Usage: always_direct allow|deny [!]aclname ...

#

#     Here you can use ACL elements to specify requests which should

#     ALWAYS be forwarded directly to origin servers.  For example,

#     to always directly forward requests for local servers use

#     something like:

#

#           acl local-servers dstdomain my.domain.net

#           always_direct allow local-servers

#

#     To always forward FTP requests directly, use

#

#           acl FTP proto FTP

#           always_direct allow FTP

#

#     NOTE: There is a similar, but opposite option named

#     'never_direct'.  You need to be aware that "always_direct deny

#     foo" is NOT the same thing as "never_direct allow foo".  You

#     may need to use a deny rule to exclude a more-specific case of

#     some other rule.  Example:

#

#           acl local-external dstdomain external.foo.net

#           acl local-servers dstdomain  .foo.net

#           always_direct deny local-external

#           always_direct allow local-servers

#

#     This option replaces some v1.1 options such as local_domain

#     and local_ip.

#

#Default:

# none

#  TAG: never_direct

#     Usage: never_direct allow|deny [!]aclname ...

#

#     never_direct is the opposite of always_direct.  Please read

#     the description for always_direct if you have not already.

#

#     With 'never_direct' you can use ACL elements to specify

#     requests which should NEVER be forwarded directly to origin

#     servers.  For example, to force the use of a proxy for all

#     requests, except those in your local domain use something like:

#

#           acl local-servers dstdomain .foo.net

#           acl all src 0.0.0.0/0.0.0.0

#           never_direct deny local-servers

#           never_direct allow all

#

#     or if Squid is inside a firewall and there are local intranet

#     servers inside the firewall use something like:

#

#           acl local-intranet dstdomain .foo.net

#           acl local-external dstdomain external.foo.net

#           always_direct deny local-external

#           always_direct allow local-intranet

#           never_direct allow all

#

#     This option replaces some v1.1 options such as inside_firewall

#     and firewall_ip.

#

#Default:

# none

#  TAG: header_access

#     Usage: header_access header_name allow|deny [!]aclname ...

#

#     WARNING: Doing this VIOLATES the HTTP standard.  Enabling

#     this feature could make you liable for problems which it

#     causes.

#

#     This option replaces the old 'anonymize_headers' and the

#     older 'http_anonymizer' option with something that is much

#     more configurable. This new method creates a list of ACLs

#     for each header, allowing you very fine-tuned header

#     mangling.

#

#     You can only specify known headers for the header name.

#     Other headers are reclassified as 'Other'. You can also

#     refer to all the headers with 'All'.

#

#     For example, to achieve the same behaviour as the old

#     'http_anonymizer standard' option, you should use:

#

#           header_access From deny all

#           header_access Referer deny all

#           header_access Server deny all

#           header_access User-Agent deny all

#           header_access WWW-Authenticate deny all

#           header_access Link deny all

#

#     Or, to reproduce the old 'http_anonymizer paranoid' feature

#     you should use:

#

#           header_access Allow allow all

#           header_access Authorization allow all

#           header_access WWW-Authenticate allow all

#           header_access Cache-Control allow all

#           header_access Content-Encoding allow all

#           header_access Content-Length allow all

#           header_access Content-Type allow all

#           header_access Date allow all

#           header_access Expires allow all

#           header_access Host allow all

#           header_access If-Modified-Since allow all

#           header_access Last-Modified allow all

#           header_access Location allow all

#           header_access Pragma allow all

#           header_access Accept allow all

#           header_access Accept-Charset allow all

#           header_access Accept-Encoding allow all

#           header_access Accept-Language allow all

#           header_access Content-Language allow all

#           header_access Mime-Version allow all

#           header_access Retry-After allow all

#           header_access Title allow all

#           header_access Connection allow all

#           header_access Proxy-Connection allow all

#           header_access All deny all

#

#     By default, all headers are allowed (no anonymizing is

#     performed).

#

#Default:

# none

#  TAG: header_replace

#     Usage:   header_replace header_name message

#     Example: header_replace User-Agent Nutscrape/1.0 (CP/M; 8-bit)

#

#     This option allows you to change the contents of headers

#     denied with header_access above, by replacing them with

#     some fixed string. This replaces the old fake_user_agent

#     option.

#

#     By default, headers are removed if denied.

#

#Default:

# none

#  TAG: icon_directory

#     Where the icons are stored. These are normally kept in

#     /usr/share/squid/icons

#

#Default:

# icon_directory /usr/share/squid/icons

#  TAG: short_icon_urls

#     If this is enabled Squid will use short URLs for icons.

#

#     If off the URLs for icons will always be absolute URLs

#     including the proxy name and port.

#

#Default:

# short_icon_urls off

#  TAG: error_directory

#     Directory where the error files are read from.

#     /usr/lib/squid/errors contains sets of error files

#     in different languages. The default error directory

#     is /etc/squid/errors, which is a link to one of these

#     error sets.

#

#     If you wish to create your own versions of the error files,

#     either to customize them to suit your language or company,

#     copy the template English files to another

#     directory and point this tag at them.

#    

#error_directory /etc/squid/errors

#

#Default:

# error_directory /etc/squid/errors

#  TAG: maximum_single_addr_tries

#     This sets the maximum number of connection attempts for a

#     host that only has one address (for multiple-address hosts,

#     each address is tried once).

#

#     The default value is one attempt, the (not recommended)

#     maximum is 255 tries.  A warning message will be generated

#     if it is set to a value greater than ten.

#

#     Note: This is in addition to the request reforwarding which

#     takes place if Squid fails to get a satisfying response.

#

#Default:

# maximum_single_addr_tries 1

#  TAG: snmp_port

#     Squid can now serve statistics and status information via SNMP.

#     A value of "0" disables SNMP support. If you wish to use SNMP,

#     set this to "3401" to use the normal SNMP support.

#

#Default:

# snmp_port 0

#  TAG: snmp_access

#     Allowing or denying access to the SNMP port.

#

#     All access to the agent is denied by default.

#     usage:

#

#     snmp_access allow|deny [!]aclname ...

#

#Example:

# snmp_access allow snmppublic localhost

# snmp_access deny all

#

#Default:

# snmp_access deny all

#  TAG: snmp_incoming_address

#  TAG: snmp_outgoing_address

#     Just like 'udp_incoming_address' above, but for the SNMP port.

#

#     snmp_incoming_address   is used for the SNMP socket receiving

#                       messages from SNMP agents.

#     snmp_outgoing_address   is used for SNMP packets returned to SNMP

#                       agents.

#

#     The default snmp_incoming_address (0.0.0.0) is to listen on all

#     available network interfaces.

#

#     If snmp_outgoing_address is set to 255.255.255.255 (the default)

#     it will use the same socket as snmp_incoming_address. Only

#     change this if you want to have SNMP replies sent using another

#     address than where this Squid listens for SNMP queries.

#

#     NOTE, snmp_incoming_address and snmp_outgoing_address can not have

#     the same value since they both use port 3401.

#

#Default:

# snmp_incoming_address 0.0.0.0

# snmp_outgoing_address 255.255.255.255

#  TAG: as_whois_server

#     WHOIS server to query for AS numbers.  NOTE: AS numbers are

#     queried only when Squid starts up, not for every request.

#

#Default:

# as_whois_server whois.ra.net

# as_whois_server whois.ra.net

#  TAG: wccp_router

#     Use this option to define your WCCP ``home'' router for

#     Squid.   Setting the 'wccp_router' to 0.0.0.0 (the default)

#     disables WCCP.

#

#Default:

# wccp_router 0.0.0.0

#  TAG: wccp_version

#     According to some users, Cisco IOS 11.2 only supports WCCP

#     version 3.  If you're using that version of IOS, change

#     this value to 3.

#

#Default:

# wccp_version 4

#  TAG: wccp_incoming_address

#  TAG: wccp_outgoing_address

#        wccp_incoming_address   Use this option if you require WCCP

#                       messages to be received on only one

#                       interface.  Do NOT use this option if

#                       you're unsure how many interfaces you

#                       have, or if you know you have only one

#                       interface.

#

#     wccp_outgoing_address   Use this option if you require WCCP

#                       messages to be sent out on only one

#                       interface.  Do NOT use this option if

#                       you're unsure how many interfaces you

#                       have, or if you know you have only one

#                       interface.

#

#        The default behavior is to not bind to any specific address.

#

#        NOTE, wccp_incoming_address and wccp_outgoing_address can not have

#        the same value since they both use port 2048.

#

#Default:

# wccp_incoming_address 0.0.0.0

# wccp_outgoing_address 255.255.255.255

# DELAY POOL PARAMETERS (all require DELAY_POOLS compilation option)

# -----------------------------------------------------------------------------

#  TAG: delay_pools

#     This represents the number of delay pools to be used.  For example,

#     if you have one class 2 delay pool and one class 3 delays pool, you

#     have a total of 2 delay pools.

#

#Default:

# delay_pools 0

#  TAG: delay_class

#     This defines the class of each delay pool.  There must be exactly one

#     delay_class line for each delay pool.  For example, to define two

#     delay pools, one of class 2 and one of class 3, the settings above

#     and here would be:

#

#Example:

# delay_pools 2      # 2 delay pools

# delay_class 1 2    # pool 1 is a class 2 pool

# delay_class 2 3    # pool 2 is a class 3 pool

#

#     The delay pool classes are:

#

#           class 1           Everything is limited by a single aggregate

#                       bucket.

#

#           class 2     Everything is limited by a single aggregate

#                       bucket as well as an "individual" bucket chosen

#                       from bits 25 through 32 of the IP address.

#

#           class 3           Everything is limited by a single aggregate

#                       bucket as well as a "network" bucket chosen

#                       from bits 17 through 24 of the IP address and a

#                       "individual" bucket chosen from bits 17 through

#                       32 of the IP address.

#

#     NOTE: If an IP address is a.b.c.d

#           -> bits 25 through 32 are "d"

#           -> bits 17 through 24 are "c"

#           -> bits 17 through 32 are "c * 256 + d"

#

#Default:

# none

#  TAG: delay_access

#     This is used to determine which delay pool a request falls into.

#     The first matched delay pool is always used, i.e., if a request falls

#     into delay pool number one, no more delay are checked, otherwise the

#     rest are checked in order of their delay pool number until they have

#     all been checked.  For example, if you want some_big_clients in delay

#     pool 1 and lotsa_little_clients in delay pool 2:

#

#Example:

# delay_access 1 allow some_big_clients

# delay_access 1 deny all

# delay_access 2 allow lotsa_little_clients

# delay_access 2 deny all

#

#Default:

# none

#  TAG: delay_parameters

#     This defines the parameters for a delay pool.  Each delay pool has

#     a number of "buckets" associated with it, as explained in the

#     description of delay_class.  For a class 1 delay pool, the syntax is:

#

#delay_parameters pool aggregate

#

#     For a class 2 delay pool:

#

#delay_parameters pool aggregate individual

#

#     For a class 3 delay pool:

#

#delay_parameters pool aggregate network individual

#

#     The variables here are:

#

#           pool        a pool number - ie, a number between 1 and the

#                       number specified in delay_pools as used in

#                       delay_class lines.

#

#           aggregate   the "delay parameters" for the aggregate bucket

#                       (class 1, 2, 3).

#

#           individual  the "delay parameters" for the individual

#                       buckets (class 2, 3).

#

#           network           the "delay parameters" for the network buckets

#                       (class 3).

#

#     A pair of delay parameters is written restore/maximum, where restore is

#     the number of bytes (not bits - modem and network speeds are usually

#     quoted in bits) per second placed into the bucket, and maximum is the

#     maximum number of bytes which can be in the bucket at any time.

#

#     For example, if delay pool number 1 is a class 2 delay pool as in the

#     above example, and is being used to strictly limit each host to 64kbps

#     (plus overheads), with no overall limit, the line is:

#

#delay_parameters 1 -1/-1 8000/8000

#

#     Note that the figure -1 is used to represent "unlimited".

#

#     And, if delay pool number 2 is a class 3 delay pool as in the above

#     example, and you want to limit it to a total of 256kbps (strict limit)

#     with each 8-bit network permitted 64kbps (strict limit) and each

#     individual host permitted 4800bps with a bucket maximum size of 64kb

#     to permit a decent web page to be downloaded at a decent speed

#     (if the network is not being limited due to overuse) but slow down

#     large downloads more significantly:

#

#delay_parameters 2 32000/32000 8000/8000 600/8000

#

#     There must be one delay_parameters line for each delay pool.

#

#Default:

# none

#  TAG: delay_initial_bucket_level  (percent, 0-100)

#     The initial bucket percentage is used to determine how much is put

#     in each bucket when squid starts, is reconfigured, or first notices

#     a host accessing it (in class 2 and class 3, individual hosts and

#     networks only have buckets associated with them once they have been

#     "seen" by squid).

#

#Default:

# delay_initial_bucket_level 50

#  TAG: incoming_icp_average

#  TAG: incoming_http_average

#  TAG: incoming_dns_average

#  TAG: min_icp_poll_cnt

#  TAG: min_dns_poll_cnt

#  TAG: min_http_poll_cnt

#     Heavy voodoo here.  I can't even believe you are reading this.

#     Are you crazy?  Don't even think about adjusting these unless

#     you understand the algorithms in comm_select.c first!

#

#Default:

# incoming_icp_average 6

# incoming_http_average 4

# incoming_dns_average 4

# min_icp_poll_cnt 8

# min_dns_poll_cnt 8

# min_http_poll_cnt 8

#  TAG: max_open_disk_fds

#     To avoid having disk as the I/O bottleneck Squid can optionally

#     bypass the on-disk cache if more than this amount of disk file

#     descriptors are open.

#

#     A value of 0 indicates no limit.

#

#Default:

# max_open_disk_fds 0

#  TAG: offline_mode

#     Enable this option and Squid will never try to validate cached

#     objects.

#

#Default:

# offline_mode off

#  TAG: uri_whitespace

#     What to do with requests that have whitespace characters in the

#     URI.  Options:

#

#     strip:  The whitespace characters are stripped out of the URL.

#           This is the behavior recommended by RFC2396.

#     deny:   The request is denied.  The user receives an "Invalid

#           Request" message.

#     allow:  The request is allowed and the URI is not changed.  The

#           whitespace characters remain in the URI.  Note the

#           whitespace is passed to redirector processes if they

#           are in use.

#     encode:     The request is allowed and the whitespace characters are

#           encoded according to RFC1738.  This could be considered

#           a violation of the HTTP/1.1

#           RFC because proxies are not allowed to rewrite URI's.

#     chop: The request is allowed and the URI is chopped at the

#           first whitespace.  This might also be considered a

#           violation.

#

#Default:

# uri_whitespace strip

#  TAG: broken_posts

#     A list of ACL elements which, if matched, causes Squid to send

#     an extra CRLF pair after the body of a PUT/POST request.

#

#     Some HTTP servers has broken implementations of PUT/POST,

#     and rely on an extra CRLF pair sent by some WWW clients.

#

#     Quote from RFC 2068 section 4.1 on this matter:

#

#       Note: certain buggy HTTP/1.0 client implementations generate an

#       extra CRLF's after a POST request. To restate what is explicitly

#       forbidden by the BNF, an HTTP/1.1 client must not preface or follow

#       a request with an extra CRLF.

#

#Example:

# acl buggy_server url_regex ^http://....

# broken_posts allow buggy_server

#

#Default:

# none

#  TAG: mcast_miss_addr

# Note: This option is only available if Squid is rebuilt with the

#       -DMULTICAST_MISS_STREAM option

#

#     If you enable this option, every "cache miss" URL will

#     be sent out on the specified multicast address.

#

#     Do not enable this option unless you are are absolutely

#     certain you understand what you are doing.

#

#Default:

# mcast_miss_addr 255.255.255.255

#  TAG: mcast_miss_ttl

# Note: This option is only available if Squid is rebuilt with the

#       -DMULTICAST_MISS_TTL option

#

#     This is the time-to-live value for packets multicasted

#     when multicasting off cache miss URLs is enabled.  By

#     default this is set to 'site scope', i.e. 16.

#

#Default:

# mcast_miss_ttl 16

#  TAG: mcast_miss_port

# Note: This option is only available if Squid is rebuilt with the

#       -DMULTICAST_MISS_STREAM option

#

#     This is the port number to be used in conjunction with

#     'mcast_miss_addr'.

#

#Default:

# mcast_miss_port 3135

#  TAG: mcast_miss_encode_key

# Note: This option is only available if Squid is rebuilt with the

#       -DMULTICAST_MISS_STREAM option

#

#     The URLs that are sent in the multicast miss stream are

#     encrypted.  This is the encryption key.

#

#Default:

# mcast_miss_encode_key XXXXXXXXXXXXXXXX

#  TAG: nonhierarchical_direct

#     By default, Squid will send any non-hierarchical requests

#     (matching hierarchy_stoplist or not cachable request type) direct

#     to origin servers.

#

#     If you set this to off, Squid will prefer to send these

#     requests to parents.

#

#     Note that in most configurations, by turning this off you will only

#     add latency to these request without any improvement in global hit

#     ratio.

#

#     If you are inside an firewall see never_direct instead of

#     this directive.

#

#Default:

# nonhierarchical_direct on

#  TAG: prefer_direct

#     Normally Squid tries to use parents for most requests. If you for some

#     reason like it to first try going direct and only use a parent if

#     going direct fails set this to on.

#

#     By combining nonhierarchical_direct off and prefer_direct on you

#     can set up Squid to use a parent as a backup path if going direct

#     fails.

#

#     Note: If you want Squid to use parents for all requests see

#     the never_direct directive. prefer_direct only modifies how Squid

#     acts on cachable requests.

#

#Default:

# prefer_direct off

#  TAG: strip_query_terms

#     By default, Squid strips query terms from requested URLs before

#     logging.  This protects your user's privacy.

#

#Default:

# strip_query_terms on

#  TAG: coredump_dir

#     By default Squid leaves core files in the directory from where

#     it was started. If you set 'coredump_dir' to a directory

#     that exists, Squid will chdir() to that directory at startup

#     and coredump files will be left there.

#

#Default:

# coredump_dir none

#

# Leave coredumps in the first cache dir

coredump_dir /var/spool/squid

#  TAG: redirector_bypass

#     When this is 'on', a request will not go through the

#     redirector if all redirectors are busy.  If this is 'off'

#     and the redirector queue grows too large, Squid will exit

#     with a FATAL error and ask you to increase the number of

#     redirectors.  You should only enable this if the redirectors

#     are not critical to your caching system.  If you use

#     redirectors for access control, and you enable this option,

#     users may have access to pages they should not

#     be allowed to request.

#

#Default:

# redirector_bypass off

#  TAG: ignore_unknown_nameservers

#     By default Squid checks that DNS responses are received

#     from the same IP addresses they are sent to.  If they

#     don't match, Squid ignores the response and writes a warning

#     message to cache.log.  You can allow responses from unknown

#     nameservers by setting this option to 'off'.

#

#Default:

# ignore_unknown_nameservers on

#  TAG: digest_generation

# Note: This option is only available if Squid is rebuilt with the

#       --enable-cache-digests option

#

#     This controls whether the server will generate a Cache Digest

#     of its contents.  By default, Cache Digest generation is

#     enabled if Squid is compiled with USE_CACHE_DIGESTS defined.

#

#Default:

# digest_generation on

#  TAG: digest_bits_per_entry

# Note: This option is only available if Squid is rebuilt with the

#       --enable-cache-digests option

#

#     This is the number of bits of the server's Cache Digest which

#     will be associated with the Digest entry for a given HTTP

#     Method and URL (public key) combination.  The default is 5.

#

#Default:

# digest_bits_per_entry 5

#  TAG: digest_rebuild_period (seconds)

# Note: This option is only available if Squid is rebuilt with the

#       --enable-cache-digests option

#

#     This is the number of seconds between Cache Digest rebuilds.

#

#Default:

# digest_rebuild_period 1 hour

#  TAG: digest_rewrite_period (seconds)

# Note: This option is only available if Squid is rebuilt with the

#       --enable-cache-digests option

#

#     This is the number of seconds between Cache Digest writes to

#     disk.

#

#Default:

# digest_rewrite_period 1 hour

#  TAG: digest_swapout_chunk_size   (bytes)

# Note: This option is only available if Squid is rebuilt with the

#       --enable-cache-digests option

#

#     This is the number of bytes of the Cache Digest to write to

#     disk at a time.  It defaults to 4096 bytes (4KB), the Squid

#     default swap page.

#

#Default:

# digest_swapout_chunk_size 4096 bytes

#  TAG: digest_rebuild_chunk_percentage   (percent, 0-100)

# Note: This option is only available if Squid is rebuilt with the

#       --enable-cache-digests option

#

#     This is the percentage of the Cache Digest to be scanned at a

#     time.  By default it is set to 10% of the Cache Digest.

#

#Default:

# digest_rebuild_chunk_percentage 10

#  TAG: chroot

#     Use this to have Squid do a chroot() while initializing.  This

#     also causes Squid to fully drop root privileges after

#     initializing.  This means, for example, that if you use a HTTP

#     port less than 1024 and try to reconfigure, you will get an

#     error.

#

#Default:

# none

#  TAG: client_persistent_connections

#  TAG: server_persistent_connections

#     Persistent connection support for clients and servers.  By

#     default, Squid uses persistent connections (when allowed)

#     with its clients and servers.  You can use these options to

#     disable persistent connections with clients and/or servers.

#

#Default:

# client_persistent_connections on

# server_persistent_connections on

#  TAG: detect_broken_pconn

#     Some servers have been found to incorrectly signal the use

#     of HTTP/1.0 persistent connections even on replies not

#     compatible, causing significant delays. This server problem

#     has mostly been seen on redirects.

#

#     By enabling this directive Squid attempts to detect such

#     broken replies and automatically assume the reply is finished

#     after 10 seconds timeout.

#

#Default:

# detect_broken_pconn off

#  TAG: balance_on_multiple_ip

#     Some load balancing servers based on round robin DNS have been

#     found not to preserve user session state across requests

#     to different IP addresses.

#

#     By default Squid rotates IP's per request. By disabling

#     this directive only connection failure trigers rotation.

#

#Default:

# balance_on_multiple_ip on

#  TAG: pipeline_prefetch

#     To boost the performance of pipelined requests to closer

#     match that of a non-proxied environment Squid can try to fetch

#     up to two requests in parallell from a pipeline.

#

#     Defaults to off for bandwidth management and access logging

#     reasons.

#

#Default:

# pipeline_prefetch off

#  TAG: extension_methods

#     Squid only knows about standardized HTTP request methods.

#     You can add up to 20 additional "extension" methods here.

#

#Default:

# none

#     on sending request entities in GET or HEAD requests.

#

#Default:

# request_entities off

#  TAG: high_response_time_warning  (msec)

#     If the one-minute median response time exceeds this value,

#     Squid prints a WARNING with debug level 0 to get the

#     administrators attention.  The value is in milliseconds.

#

#Default:

# high_response_time_warning 0

#  TAG: high_page_fault_warning

#     If the one-minute average page fault rate exceeds this

#     value, Squid prints a WARNING with debug level 0 to get

#     the administrators attention.  The value is in page faults

#     per second.

#

#Default:

# high_page_fault_warning 0

#  TAG: high_memory_warning

#     If the memory usage (as determined by mallinfo) exceeds

#     value, Squid prints a WARNING with debug level 0 to get

#     the administrators attention.

#

#Default:

# high_memory_warning 0

#  TAG: store_dir_select_algorithm

#     Set this to 'round-robin' as an alternative.

#

#Default:

# store_dir_select_algorithm least-load

#  TAG: forward_log

# Note: This option is only available if Squid is rebuilt with the

#       -DWIP_FWD_LOG option

#

#     Logs the server-side requests.

#

#     This is currently work in progress.

#

#Default:

# none

#  TAG: ie_refresh      on|off

#     Microsoft Internet Explorer up until version 5.5 Service

#     Pack 1 has an issue with transparent proxies, wherein it

#     is impossible to force a refresh.  Turning this on provides

#     a partial fix to the problem, by causing all IMS-REFRESH

#     requests from older IE versions to check the origin server

#     for fresh content.  This reduces hit ratio by some amount

#     (~10% in my experience), but allows users to actually get

#     fresh content when they want it.  Note that because Squid

#     cannot tell if the user is using 5.5 or 5.5SP1, the behavior

#     of 5.5 is unchanged from old versions of Squid (i.e. a

#     forced refresh is impossible).  Newer versions of IE will,

#     hopefully, continue to have the new behavior and will be

#     handled based on that assumption.  This option defaults to

#     the old Squid behavior, which is better for hit ratios but

#     worse for clients using IE, if they need to be able to

#     force fresh content.

#

#Default:

# ie_refresh off

#  TAG: vary_ignore_expire    on|off

#     Many HTTP servers supporting Vary gives such objects

#     immediate expiry time with no cache-control header

#     when requested by a HTTP/1.0 client. This option

#     enables Squid to ignore such expiry times until

#     HTTP/1.1 is fully implemented.

#     WARNING: This may eventually cause some varying

#     objects not intended for caching to get cached.

#

#Default:

# vary_ignore_expire off

#  TAG: sleep_after_fork      (microseconds)

#     When this is set to a non-zero value, the main Squid process

#     sleeps the specified number of microseconds after a fork()

#     system call. This sleep may help the situation where your

#     system reports fork() failures due to lack of (virtual)

#     memory. Note, however, that if you have a lot of child

#     processes, these sleep delays will add up and your

#     Squid will not service requests for some amount of time

#     until all the child processes have been started.

#

#Default:

# sleep_after_fork 0

#  TAG: relaxed_header_parser on|off

#     Set this to off if you want Squid to be strict about

#     the HTTP protocol syntax and reject non-compliant requests

#     or responses.

#

#     In the default "on" setting Squid accepts certain forms

#     of non-compliant HTTP messages where it is unambigous

#     what the sending application indended even if the message

#     is not correctly formatted. The messages is then normalized

#     to the correct form when forwarded by Squid.

#

#Default:

# relaxed_header_parser on