请看：我的squid代理服务器为什么无法正常工作？

Question

现象：
1)客户端无法通过squid代理上网，提示access denied.
2)iptables去掉“/sbin/iptables -t nat -A PREROUTING -i eth0 -p tcp --dport 80 -j REDIRECT --to-port 3128“，可以上网。

iptables配置
#delete any existing chains
/sbin/iptables -F
/sbin/iptables -F -t nat
/sbin/iptables -X
/sbin/iptables -Z
#squid proxy
/sbin/iptables -t nat -A PREROUTING -i eth0 -p tcp --dport 80 -j REDIRECT --to-port 3128
/sbin/iptables -t nat -A POSTROUTING -o eth1 -j SNAT --to xx.xx.xx.xx

squid.conf配置
#        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.
#

# 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, then 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, then 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.
#
#Default:
http_port 192.168.10.30:3128

#  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
#        The port number where Squid sends and receives HTCP queries to
#        and from neighbor caches.  To turn it on you want to set it 4827.
#        By default it is set to "0" (disabled).
#
#        To enable this option, you must use --enable-htcp with the
#        configure script.
#
#Default:
# htcp_port 0

#  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: tcp_outgoing_address
#  TAG: udp_incoming_address
#  TAG: udp_outgoing_address
#        Usage: tcp_incoming_address 10.20.30.40
#               udp_outgoing_address fully.qualified.domain.name
#
#        tcp_outgoing_address        is used for connections made to remote
#                                servers and other caches.
#        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 *_incoming_address value of 0.0.0.0 indicates that Squid should
#        listen on all available interfaces.
#
#        If udp_outgoing_address is set to 255.255.255.255 (the default)
#        then 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.
#
#        NOTE, tcp_incoming_address has been removed.  You can now
#        specify IP addresses on the 'http_port' line.
#
#Default:
# tcp_outgoing_address 255.255.255.255
# 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
#                     connect-timeout=nn
#                     digest-url=url
#                     allow-miss
#
#                     use 'proxy-only' to specify that 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 that 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.
#
#                     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)
#
#        NOTE: non-ICP 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 that 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 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 reply to
#        immediately removed from the cache.  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 PLACES A LIMIT ON ONE ASPECT OF SQUID'S MEMORY
#        USAGE.  SQUID USES MEMORY FOR OTHER THINGS AS WELL.
#        YOUR PROCESS WILL PROBABLY BECOME TWICE OR THREE TIMES
#        BIGGER THAN THE VALUE YOU PUT HERE
#
#        '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 64 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.  Most
#        everyone will want to use "ufs" as the type.  If you are using
#        Async I/O (--enable async-io) on Linux or Solaris, then you may
#        want to try "aufs" as the type.  Async IO support may be
#        buggy, however, so beware.
#
#        'Directory' is a top-level directory where cache swap
#        files will be stored.  If you want to use an entire disk
#        for caching, then 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.
#
#        '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 recevies some replies. Default is 72
#
#        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.
#
#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.
#
#Default:
cache_access_log /var/log/squid/access.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.log."  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 then 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
#        then 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, then these log files will NOT correspond to
#        the correct 'cache_dir' entry (unless you manually rename
#        them).  We recommend that 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 that 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 /usr/lib/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 that 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 that 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, then 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 then turn this off.
#
#Default:
# ftp_sanitycheck 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/

#  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 then all DNS servers for the queried domain
#        is assumed to be unavailable.
#
#Default:
# dns_timeout 5 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.
#
#        Example: dns_nameservers 10.0.0.1 192.172.0.4
#
#Default:
# none

#  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.
#        This is only useful if you configured Squid (during compilation)
#        with the '--enable-icmp' option.
#
#Default:
# pinger_program /usr/lib/squid/

#  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 Release-Notes 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 a accelerator then 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: authenticate_program
#        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.  If you use an authenticator,
#        make sure you have 1 acl of type proxy_auth.  By default, the
#        authenticator_program is not used.
#
#        If you want to use the traditional proxy authentication,
#        jump over to the ../auth_modules/NCSA directory and
#        type:
#                % make
#                % make install
#
#        Then, set this line to something like
#
#        authenticate_program /usr/bin/ncsa_auth /usr/etc/passwd
#
#Default:
# none

#  TAG: authenticate_children
#        The number of authenticator processes to spawn (default 5). 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.
#
#Default:
# authenticate_children 5

#  TAG: authenticate_ttl
#        The time a checked username/password combination remains cached.
#        If a wrong password is given for a cached user, the user gets
#        removed from the username/password cache forcing a revalidation.
#
#Default:
# authenticate_ttl 1 hour

#  TAG: authenticate_ip_ttl
#        With this option you control how long a proxy authentication
#        will be bound to a specific IP address. If a request using
#        the same user name is received during this time then access
#        will be denied and both users are required to reauthenticate
#        them selves.  The idea behind this is to make it annoying
#        for people to share their password to their friends, but
#        yet allow a dialup user to reconnect on a different dialup
#        port.
#
#        The default is 0 to disable the check. Recommended value
#        if you have dialup users are no more than 60 seconds to allow
#        the user to redial without hassle. If all your users are
#        stationary then higher values may be used.
#
#        See also authenticate_ip_ttl_is_strict
#
#Default:
# authenticate_ip_ttl 0 seconds

#  TAG: authenticate_ip_ttl_is_strict
#        This option makes authenticate_ip_ttl a bit stricted. With this
#        enabled authenticate_ip_ttl will deny all access from other IP
#        addresses until the TTL has expired, and the IP address "owning"
#        the userid will not be forced to reauthenticate.
#
#Default:
# authenticate_ip_ttl_is_strict on

# 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, there will be no limit
#        imposed.
#
#Default:
# request_body_max_size 1 MB

#  TAG: reply_body_max_size        (KB)
#        This option specifies the maximum size of a reply body.  It
#        can be used to prevent users from downloading very large files,
#        such as MP3's and movies.   The reply 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 this parameter, 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.
#
#        NOTE: 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

#  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 was 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.
#               
#        Please see the file doc/Release-Notes-1.1.txt for a full
#        description of Squid's refresh algorithm.  Basically a
#        cached object is: (the order is changed from 1.1.X)
#
#                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, then 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.
#
#Default:
# refresh_pattern ^ftp:                1440        20%        10080
# refresh_pattern ^gopher:        1440        0%        1440
# refresh_pattern .                0        20%        4320

#  TAG: reference_age
#        As a part of normal operation, Squid performs Least Recently
#        Used removal of cached objects.  The LRU age for removal is
#        computed dynamically, based on the amount of disk space in
#        use.  The dynamic value can be seen in the Cache Manager 'info'
#        output.
#
#        The 'reference_age' parameter defines the maximum LRU age.  For
#        example, setting reference_age to '1 week' will cause objects
#        to be removed if they have not been accessed for a week or
#        more.  The default value is one year.
#
#        Specify a number here, followed by units of time.  For example:
#                1 week
#                3.5 days
#                4 months
#                2.2 hours
#
#        NOTE: this parameter is not used when using the enhanced
#        replacement policies, GDSH or LFUDA.
#
#Default:
# reference_age 1 year

#  TAG: quick_abort_min        (KB)
#  TAG: quick_abort_max        (KB)
#  TAG: quick_abort_pct        (percent)
#        The cache can be configured to continue downloading aborted
#        requests.  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.  Setting 'quick_abort_min' to -1
#        will disable the quick_abort feature.
#
#        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.
#
#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
#        Time-to-Live (TTL) for positive caching of successful DNS lookups.
#        Default is 6 hours (360 minutes).  If you want to minimize the
#        use of Squid's ipcache, set this to 1, not 0.
#
#Default:
# positive_dns_ttl 6 hours

#  TAG: negative_dns_ttl        time-units
#        Time-to-Live (TTL) for negative caching of failed DNS lookups.
#
#Default:
# negative_dns_ttl 5 minutes

#  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 then 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 that 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: connect_timeout        time-units
#        Some systems (notably Linux) can not be relied upon to properly
#        time out connect(2) requests.  Therefore the Squid process
#        enforces its own timeout on server connections.  This parameter
#        specifies how long to wait for the connect to complete.  The
#        default is two minutes (120 seconds).
#
#Default:
# connect_timeout 2 minutes

#  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: siteselect_timeout        time-units
#        For URN to multiple URL's URL selection
#
#Default:
# siteselect_timeout 4 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 connection
#        establishment.  For persistent connections, wait this long
#        after the previous request completes.
#
#Default:
# request_timeout 30 seconds

#  TAG: client_lifetime        time-units
#        The maximum amount of time that 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, pconn_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 requests.  If this is too high,
#        and you enabled 'ident_lookup', then you might be susceptible
#        to denial-of-service by having many ident requests going at
#        once.
#
#        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.
#
#        This option may be disabled by using --disable-ident with
#        the configure script.
#
#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 src dst srcdomain dstdomain url_pattern
#                urlpath_pattern time port proto method browser user
#
#        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]
#            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 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
#        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
#          # authenticate_program).
#          #
#          # 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 req_mime_type mime-type1 ...
#          # regex match agains 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.
#
#Examples:
#acl myexample dst_as 1241
#acl password proxy_auth REQUIRED
#acl fileupload req_mime_type -i ^multipart/form-data$
#
#Recommended minimum configuration:
acl tc src 192.168.10.0/255.255.255.255
acl class src 192.168.20.0/255.255.255.255
acl all src 0.0.0.0/0.0.0.0
acl manager proto cache_object
acl localhost src 127.0.0.1/255.255.255.255
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 Safe_ports port 901                # SWAT
acl purge method PURGE
acl CONNECT method CONNECT

#  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, then 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 manager localhost
http_access deny manager
# Only allow purge requests from localhost
http_access allow purge localhost
http_access deny purge
# Deny requests to unknown ports
http_access deny !Safe_ports
# Deny CONNECT to other than SSL ports
http_access deny CONNECT !SSL_ports
#
# INSERT YOUR OWN RULE(S) HERE TO ALLOW ACCESS FROM YOUR CLIENTS
#
http_access allow tc
http_access deny class
http_access allow localhost
# And finally deny all other access to this proxy
http_access deny 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 eveyone
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 that 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: proxy_auth_realm
#        Specifies the realm name which is to be reported to the client for
#        proxy authentication (part of the text the user will see when
#        prompted their username and password).
#
#Default:
# proxy_auth_realm Squid proxy-caching web server

#  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
#
#        This option may be disabled by using --disable-ident with
#        the configure script.
#
#Default:
# ident_lookup_access deny all

# ADMINISTRATIVE PARAMETERS
# -----------------------------------------------------------------------------

#  TAG: cache_mgr
#        Email-address of local cache manager who will receive
#        mail if the cache dies.  The default is "webmaster."
#
#Default:
cache_mgr webmaster@dyzx.cn

#  TAG: cache_effective_user
#  TAG: cache_effective_group
#
#        If the cache is run as root, it will change its effective/real
#        UID/GID to the UID/GID specified below.  The default is to
#        change to UID to proxy and GID to proxy.
#
#        If Squid is not started as root, the default is to keep the
#        current UID/GID.  Note that if Squid is not started as root then
#        you cannot set http_port to a value lower than 1024.
#
#Default:
# cache_effective_user proxy
# cache_effective_group proxy

#  TAG: visible_hostname
#        If you want to present a special hostname in error messages, etc,
#        then 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' then you must give each machine a different
#        'unique_hostname' so that forwarding loops can be detected.
#
#Default:
# none

#  TAG: hostname_aliases
#        A list of other DNS names that 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 virtual host support then specify the hostname
#        as "virtual".
#
#        If you want virtual port support then specify the port as "0".
#
#        NOTE: enabling httpd_accel_host disables proxy-caching and
#        ICP.  If you want these features enabled also, then set
#        the 'httpd_accel_with_proxy' option.
#
#Default:
httpd_accel_host virtual
httpd_accel_port 80

#  TAG: httpd_accel_single_host        on|off
#        If you are running Squid as a accelerator and have a single backend
#        server then set this to on. This causes Squid to forward the request
#        to this server irregardles of what any redirectors or Host headers
#        says.
#
#        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 that 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 on

#  TAG: httpd_accel_uses_host_header        on|off
#        HTTP/1.1 requests include a Host: header which is basically the
#        hostname from the URL.  Squid can be an accelerator for
#        different HTTP servers by looking at this header.  However,
#        Squid does NOT check the value of the Host header, so it opens
#        a big security hole.  We recommend that 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 on

# 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>;'.
#
#        Note2, for Debian/Linux the default of logfile_rotate is
#        zero, since it includes external logfile-rotation methods.
#
#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.
#
#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
#        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 then 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.
#
#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, then
#        it is probably okay to set this to 'on'.
#
#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 actio

Answer 1

补充：squid version 2.5.stable7