Source Format Enforcement (#532)

[thirdparty/squid.git] / doc / release-notes / release-3.2.sgml

<!doctype linuxdoc system>
<article>
<title>Squid 3.2.14 release notes</title>
<author>Squid Developers</author>

<abstract>
This document contains the release notes for version 3.2 of Squid.
Squid is a WWW Cache application developed by the National Laboratory
for Applied Network Research and members of the Web Caching community.
</abstract>

<toc>

<sect>Notice
<p>
The Squid Team are pleased to announce the release of Squid-3.2.14.

This new release is available for download from <url url="http://www.squid-cache.org/Versions/v3/3.2/"> or the
 <url url="http://www.squid-cache.org/Download/http-mirrors.html" name="mirrors">.

<p>A large number of the show-stopper bugs have been fixed along with general improvements to the IPv6 support.
While this release is not fully bug-free we believe it is ready for use in production on many systems.

<p>We welcome feedback and bug reports. If you find a bug, please see <url url="http://wiki.squid-cache.org/SquidFaq/BugReporting"> for how to submit a 
report with a stack trace.

<sect1>Known issues
<p>
Although this release is deemed good enough for use in many setups, please note the existence of 
<url url="http://bugs.squid-cache.org/buglist.cgi?query_format=advanced&amp;product=Squid&amp;bug_status=UNCONFIRMED&amp;bug_status=NEW&amp;bug_status=ASSIGNED&amp;bug_status=REOPENED&amp;version=3.2" name="open bugs against Squid-3.2">.

<p>Some issues to note as currently known in this release which are not able to be fixed in the 3.2 series are:

<itemize>
        <item>TCP logging of access.log does not recover from broken connections well.
        <item>SSL-Bump not re-wrapping decrypted traffic in CONNECT for peers.
        <item>Cache Manager reports in txt/plain format even when requested directly via browser.
</itemize>

<p>Currently known issues which only depends on available developer time and may still be resolved in a future 3.2 release are:

<itemize>
        <item>SMP Support still has a number of important bugs needing to be resolved. see the bugs list above for details.
        <item>Windows support is still incomplete.
        <item>The lack of some features available in Squid-2.x series. See the regression sections below for full details.
</itemize>


<sect1>Changes since earlier releases of Squid-3.2
<p>
The 3.2 change history can be <url url="http://www.squid-cache.org/Versions/v3/3.2/changesets/" name="viewed here">.

<sect>Major new features since Squid-3.1
<p>Squid 3.2 represents a new feature release above 3.1.

<p>The most important of these new features are:
<itemize>
        <item>CVE-2009-0801 : NAT interception vulnerability to malicious clients.
        <item>NCSA helper DES algorithm password limits
        <item>SMP scalability
        <item>Helper Multiplexer and On-Demand
        <item>Helper Name Changes
        <item>Multi-Lingual manuals
        <item>Solaris 10 pthreads Support
        <item>Surrogate/1.0 protocol extensions to HTTP
        <item>Logging Infrastructure Updated
        <item>Client Bandwidth Limits
        <item>Better eCAP support
        <item>Cache Manager access changes
</itemize>

Most user-facing changes are reflected in squid.conf (see below).


<sect1>CVE-2009-0801 : NAT interception vulnerability to malicious clients.
<p>Details in Advisory <url url="http://www.squid-cache.org/Advisories/SQUID-2011_1.txt" name="SQUID-2011:1">

<p>Squid locates the authority-URL details available in an HTTP request as
  defined by RFC 2616 and validates that all found representations are
  <em>textually</em> equivalent. In the case of intercepted traffic the
  client destination IP is also compared to the Host: authority domains
  DNS entries.

<p>When the Host: authority contradicts another authority source Squid will log
  "SECURITY ALERT: Host: header forgery detected". The response will then be determined
  by the <url url="http://www.squid-cache.org/Doc/config/host_verify_strict/" name="host_verify_strict">
  directive. Squid will respond with 409 Conflict error response when strict validation
  fails and handles the request normally when strict validation succeeds or is OFF (default).

<p>Relaying of messages which FAIL non-strict Host: validation are permitted through Squid but
  only to the original destination IP the client was requesting or to explicit peers. This means
  DNS lookups to locate alternative DIRECT destinations will not be done.

<p>Known Issue: When non-strict validation fails Squid will relay the request, but can only do
  so safely to the original destination IP the client was contacting. The client original
  destination IP is lost when relaying to peers in a hierarchy. This means the upstream peers
  are still at risk of causing same-origin bypass CVE-2009-0801 vulnerability.
  Developer time is required to implement safe transit of these requests.
  Please contact squid-dev if you are able to assist or sponsor the development.


<sect1>NCSA helper DES algorithm password limits
<p>Details in Advisory <url url="http://www.squid-cache.org/Advisories/SQUID-2011_2.txt" name="SQUID-2011:2">

<p>The DES algorithm used by the NCSA Basic authentication helper has an
   limit of 8 bytes but some implementations do not error when truncating
   longer passwords down to this unsafe level.

<p>This both significantly lowers the threshold of difficulty decrypting
   captured password files and hides from users the fact that the extra bits
   of their chosen long password is not being utilized.

<p>The NCSA helper bundled with Squid will prevent passwords longer than 8
   characters being sent to the DES algorithm. The MD5 hash algorithm which
   supports longer than 8 character passwords is also supported by this helper
   and should be used instead.


<sect1>SMP scalability
<p>The new "workers" squid.conf option can be used to launch multiple worker
   processes and utilize multiple CPU cores. The overall intent is to make
   multiple workers look like one to an outside observer, while providing
   knobs to customize each worker behavior if needed.

<p>By default, all worker processes are configured identically and do what a
   single Squid instance would have done. Squid.conf macro substitutions and
   conditionals (see below) can be used to customize individual worker
   configurations. In the paragraphs below, "can share" implies "will share by
   default".

<p>Workers can share HTTP, HTTPS, SNMP, ICP, and HTCP listening addresses.
   Configuration related to ICP and HTCP clients must be adjusted to avoid
   source address conflicts: Modify the IP address and/or the port used for
   the protocol. Workers do not share DNS addresses by default because the OS
   assigns each worker a unique DNS port.

<p>Workers can share logs.

<p>Workers can share caches. Memory cache is automatically shared when multiple
   workers are used. Cache_dir are shared when configured with the <em>rock</em>
   storage type. Cache_dir of other types must be adjusted to point each
   disk-caching worker to its own disk area. ICP and HTCP responses are based
   on the responding worker cache state.

<p>Cache manager statistics are reported from a worker point of view, for now.
   Though some reports are combined. SNMP statistics are combined across all
   workers.

<p>Startup, reconfiguration, shutdown, and log rotation are handled as for a
   monolithic Squid. Abnormally terminated workers are restarted while
   other workers continue serving traffic.

<sect2>Squid.conf macros and conditionals
<p>Added support for process_name and process_number macros as well as simple
   if-statement conditionals in squid.conf. These features allow individual
   worker customization in SMP mode. For details, search for "Conditional
   configuration" and "SMP-Related Macros" sections in squid.conf.documented.


<sect1>Helper Multiplexer
<p>The helper multiplexer's purpose is to relieve some of the burden
   Squid has when dealing with slow helpers. It does so by acting as a
   middleman between squid and the actual helpers, talking to Squid via
   the multiplexed concurrent variant of the helper protocol and to the
   helpers via the non-concurrent variant.

<p>Helpers are started on demand, and in theory the muxer can handle up to
   1k helpers per instance. It's up to squid to decide how many helpers
   to start.

<p>The muxer knows nothing about the actual messages being passed around,
   and as such can't really (yet?) compensate for broken helpers.
   It is not yet able to manage dying helpers, but it will.

<p>To configure the multiplexer add its binary name (usually /usr/share/libexec/helper-mux.pl)
   in front of the name of whichever helper is being multiplexed. It takes the helper binary
   path and parameters as its own command parameters. The <em>concurrency</em> setting already
   existing in Squid is used to configure how many child helpers it may run.

<p>For example, a traditional configuration is
   <verb>
        url_rewrite_program /your/redirector.sh
        url_rewrite_children 5
   </verb>
   the alternative multiplexer configuration is:
   <verb>
        url_rewrite_program /usr/share/libexec/helper-mux.pl /your/redirector.sh
        url_rewrite_children 1 concurrency=5
   </verb>

<p>Helpers which are already concurrent protocol enabled gain little benefit from the multiplexer
   on most systems. However on some systems where Squid spawning helpers causes excess memory usage
   the reduction in direct helper spawned by Squid can result in a great reduction in resource use.

<p>The helper can be controlled using various signals:
   <itemize>
        <item>SIGHUP: dump the state of all helpers to STDERR
   </itemize>


<sect1>Helpers On-Demand
<p>Traditionally Squid has been configured with a fixed number of helpers and started them during
  it's start and reconfigure phases. This forces the hard configuration problem of how many helpers
  will be needed to be solved before starting Squid in production use.

<p>The on-demand helpers feature allows greater flexibility and resolves this problem by allowing
  maximum, initial and idle thresholds to be configured. Squid will start the initial set during
  start and reconfigure phases. However over the operational use new helpers up to the maxium will
  be started as load demands. The idle threshold determines how many more helpers to start if the
  currently running set is not enough to handle current request loads.

<p>For example, a traditional configuration is
   <verb>
        auth_param ntlm /usr/libexec/squid/ntlm_auth
        auth_param ntlm children 200
   </verb>
   the alternative on-demand configuration could be:
   <verb>
        auth_param ntlm /usr/libexec/squid/ntlm_auth
        auth_param ntlm children 200 startup=10 idle=2
   </verb>

<p>The example still permits up to 200 helpers to be running at once under peak traffic loads.
  But only starts 10 when Squid is initialized resulting in a faster boot up.
  When client requests threaten to overload the running helpers an additional 2 will be started.

<p>NOTE: if no <em>startup</em> and <em>idle</em> values are specified the traditional behaviour
  of starting the maximum number of helpers will occur.


<sect1>Helper Name Changes
<p>To improve the understanding of what each helper does and where it should be used the helper binaries
   which are bundled with Squid have undergone a naming change in this release.

<p>Below is a list of the old helper names and what their names have changed to.
   For several helpers the directory name used in --enable-X-helpers configure option has also changed.

<sect2>Basic Authentication protocol helpers
<p><itemize>
        <item>squid_db_auth - basic_db_auth - Retrieve authentication details from a simple SQL database table.
        <item>getpwnam_auth - basic_getpwname_auth - Authenticate with local system user accounts.
        <item>squid_ldap_auth - basic_ldap_auth - Authenticate with LDAP user accounts.
        <item>MSNT-multi-domain - basic_msnt_multi_domain_auth - Authenticate with any one of multiple Windows Domain Controllers.
        <item>msnt_auth - basic_msnt_auth - Authenticate with Windows Domain Controllers selected by username.
        <item>ncsa_auth - basic_ncsa_auth - Authenticate with NCSA httpd-style password file.
        <item>yp_auth - basic_nis_auth - Authenticate with NIS security system.
        <item>pam_auth - basic_pam_auth - Authenticate with the system PAM infrastructure.
        <item>pop3.pl - basic_pop3_auth - Authenticate with a mail server POP3/SMTP credentials.
        <item>squid_radius_auth - basic_radius_auth - Authenticate with RADIUS.
        <item>squid_sasl_auth - basic_sasl_auth - Authenticate with SASL.
        <item>smb_auth - basic_smb_auth - Authenticate with Samba SMB.
        <item>mswin_sspi - basic_sspi_auth - Authenticate with a Windows Domain Controller using SSPI.
</itemize>

<sect2>Digest Authentication protocol helpers
<p><itemize>
        <item>digest_pw_auth - digest_file_auth - Authenticate against credentials stored in a simple text file.
</itemize>

<sect2>External ACL helpers
<p><itemize>
        <item>mswin_check_ad_group - ext_ad_group_acl - Check logged in users Group membership using Active Directory.
        <item>ip_user_check - ext_file_userip_acl - Restrict users to certain IP addresses, using a text file backend.
        <item>squid_kerb_ldap - ext_kerberos_ldap_group_acl - Check logged in Kerberos or NTLM users Group membership using LDAP.
        <item>squid_ldap_group - ext_ldap_group_acl - Check logged in users Group membership using LDAP.
        <item>mswin_check_lm_group - ext_lm_group_acl - Check logged in users Group membership using LanManager.
        <item>squid_session - ext_session_acl - Maintain a session cache of client identifiers (usually IP address).
                        This helper has also gone through a version update and now uses more current BerkeleyDB 4.1+ APIs.
        <item>squid_unix_group - ext_unix_group_acl - Check logged in users Group membership using local UNIX groups.
        <item>wbinfo_group.pl - ext_wbinfo_group_acl - Check logged in users Group membership using wbinfo.
</itemize>

<sect2>Negotiate Authentication protocol helpers
<p><itemize>
        <item>squid_kerb_auth - negotiate_kerberos_auth - Authenticate with Kerberos servers.
        <item>mswin_sspi - negotiate_sspi_auth - Authenticate with a Windows Domain Controller using SSPI.
        <item>negotiate_wrapper - negotiate_wrapper_auth - Split Negotiate traffic between Kerberos and NTLM helpers.
</itemize>

<sect2>NTLM Authentication protocol helpers
<p><itemize>
        <item>no_check.pl - Deprecated. - Use the faster and less easily decrypted ntlm_fake_auth instead.
        <item>fakeauth_auth - ntlm_fake_auth - Perform NTLMSSP to recover the username but don't verify the password.
        <item>ntlm_auth - ntlm_smb_lm_auth - Perform SMB LanManager domain-less authentication over NTLM protocol.
        <item>mswin_ntlm_auth - ntlm_sspi_auth - Perform NTLMSSP authentication using Windows native Security Support Provider Interface API.
</itemize>

<sect2>URL re-write helpers
<p>This group of helpers have been bundled to demonstrate how to code URL re-writers:
<itemize>
        <item>url_fake_rewrite - Accept various url_rewrite details and log the input.
</itemize>


<sect1>Multi-Lingual manuals
<p>The man(8) and man(1) pages bundled with Squid are now provided online for all
   versions and beginning with 3.2 they are available in languages other than English (where translated).

<p>Details in <url url="http://wiki.squid-cache.org/Translations" name="The Squid wiki">

<p>3.1 began the Internationalization of Squid with the public facing error pages.
   This move begins the Localization of the internal administrator facing manuals.


<sect1>Solaris 10 pthreads Support (Experimental)
<p>Automatic detection and use of the pthreads library available from Solaris 10

<p>The result of this addition means that faster more efficient AUFS cache storage mechanism
  is now available in Solaris 10.

<p>Support is experimental at this stage due to lack of feedback on the results of enabling it.
  We recommend giving AUFS a try for faster disk storage and encourage feedback.


<sect1>Surrogate/1.0 protocol extensions to HTTP
<p>The <em>Surrogate</em> extensions to HTTP protocol enable an origin web server to specify separate
  cache controls for a reverse proxy acting on its behalf. Previously this was closely tied with the ESI
  feature support in Squid. This release opens Surrogate support to all reverse proxies.

<p>Reverse proxy requests sent on to the web server include the HTTP header <em>Surrogate-Capabilities:</em>
  specifying the capabilities of the reverse proxy along with an ID which can be used to target responses with
  a <em>Surrogate-Control:</em> HTTP header used instead of the <em>Cache-Control:</em> header.

<p>The default surrogate ID is generated automatically from the Squid site-unique hostname as found by the
  automatic detection or manual configuration of <em>visible_hostname</em> although can be configured
  separately with the <em>httpd_accel_surrogate_id</em> option.

<p><em>Security Considerations:</em> Websites should be careful of accepting any surrogate ID.
  Older releases of Squid leak the Surrogate-Control headers to external servers.
  This 3.2 series of Squid will now prevent this leakage of its own ID destined responses, however it is possible
  and for some uses desirable to receive external reverse-proxies <em>Surrogate-Capabilities:</em> headers.

<p><em>NOTE:</em> Several operating system distributions historically package Squid with a forced value of
  <em>visible_hostname localhost</em>. If this is done on a Surrogate enabled install a manual re-configuration
  is required to prevent an unacceptable surrogate ID of 'localhost' being generated.


<sect1>Logging Infrastructure Updated
<p>The advanced logging modules introduced in Squid-2.7 are now available from Squid-3.2.

<p>This feature is documented at http://wiki.squid-cache.org/Features/LogModules

<p>The new infrastructure currently supports several different channels types (modules) ranging from
  direct filesystem logging (stdio, daemon) to network logging (syslog, UDP and TCP). The daemon logging
  interface allows for a custom helper to be written to process logs in real-time.

<p>Upgrading: the <em>access_log</em> and <em>cache_store_log</em> were previously logged via what is
  now called the <em>stdio</em> module.
  This is still supported and used by default if no module is named. For best performance particularly in SMP
  environments we recommend the <em>daemon</em> be used. The provided <em>log_file_daemon</em> helper
  performs the traditional logging to local filesystem.

<p>Additional to this the cache.log can now be limited to a smaller number of files stored.
  Traditionally cache.log.N has been fixed at the same number of rotated files as access.log.N through the
  <em>logfile_rotate</em> setting. The <em>debug_options</em> setting can now be used to configure the number
  of debug cache.log files to rotate through with a <em>rotate=N</em> option. This is particularly useful for
  logging a single cache.log at relatively high debug levels on a high-traffic system. Or one which is
  required to store a long period of access.log and needs to conserve disk space.

<p>The <em>referer_log</em> and <em>useragent_log</em> directives have been converted to built-in log formats.
  These logs are now created using an <em>access_log</em> line with the format "referrer" or "useragent".
  They also now log all client requests, if there was no Referer or User-Agent header a dash (-) is logged.

<p>Known Issue: The TCP logging module does not recover from broken connections well.
  At present it will restart the affected Squid instance if the TCP connection is broken.


<sect1> Client Bandwidth Limits
<p>In mobile environments, Squid may need to limit Squid-to-client bandwidth
   available to individual users, identified by their IP addresses. The IP
   address pool can be as large as a /10 IPv4 network (4 million unique IP
   addresses) and even larger in IPv6 environments. On the other hand, the code
   should support thousands of connections coming from a single IP (e.g.,
   a child proxy).

<p>The implementation is based on storing bandwidth-related "bucket" information
   in the existing "client database" hash (client_db.cc). The old code already
   assigned each client IP a single ClientInfo object, which satisfies the
   client-side IP-based bandwidth pooling requirements. The old hash size is
   increased to support up to 32K concurrent clients if needed.

<p>Client-side pools are configured similarly to server-side ones, but there is
   only one pool class. See client_delay_pools,
   client_delay_initial_bucket_level, client_delay_parameters, and
   client_delay_access in squid.conf. The client_delay_access matches the client
   with delay parameters. It does not pool clients from different IP addresses
   together.

<p>Special care is taken to provide fair distribution of bandwidth among clients
   sharing the same bucket (i.e., clients coming from the same IP address).
   Multiple same-IP clients competing for bandwidth are queued using FIFO
   algorithm. If a bucket becomes empty, the first client among those sharing 
   the bucket is delayed by 1 second before it can attempt to receive more
   response data from Squid.  This delay may need to be lowered in 
   high-bandwidth environments.


<sect1>Better eCAP Suport
<p>Support for libecap version 0.2.0 has been added with this series of Squid. Bringing
   better support for body handling, and logging.

<p>Known Issue: Due to API changes in libecap this release of Squid will not build
  against any older libecap releases.


<sect1>Cache Manager access changes
<p>The Squid Cache Manager has previously only been accessible under the cache_object://
   URL scheme. Which has restricted its reporting to tools which can send arbitrary
   URI to the proxy.

<p>This version of Squid now provides access through the http:// and https:// URL schemes
   allowing web browsers access without having to use the cachemgr.cgi gateway and enabling
   the use of HTTPS security were desired.

<p>The cache manager is available under the path prefix /squid-internal-mgr/. For example
   the URL http://example/com/squid-internal-mgr/menu will bring up the manager menu. This
   means there are some configuration changes required to lock down manager access.
   The <em>manager</em> ACL needs changing. A built-in definition is now used, equivalent
   to the following regex pattern:
<verb>
        ^(cache_object://|https?://[^/]+/squid-internal-mgr/)
</verb>

<p>The manager prefix /squid-internal-mgr/ with no action attempts to load an optional
   template MGR_INDEX which may be installed amongst in the Squid error templates.
   This template is not supplied with Squid but intended to be supplied by separate
   cache manager applications as their front page embedding all scripts, accessors or
   redirects required for their initial GUI display.

<p>MGR_INDEX file
<itemize>
   <item>should contain a complete HTML page, with optional client-side scripting.
   <item>must not contain server-side scripting. 
   <item>will have macro substitution performed on it using the same macros as used by the error page templates.
</itemize>

<p>Version 3.2 of the CGI cache manager tool now presents XHR scripted probes to detect
   proxies presenting these manager index pagess and provides direct HTTP/HTTPS web links
   to those managers.


<sect>Changes to squid.conf since Squid-3.1
<p>
There have been changes to Squid's configuration file since Squid-3.1.

This section gives a thorough account of those changes in three categories:

<itemize>
        <item><ref id="newtags" name="New tags">
        <item><ref id="modifiedtags" name="Changes to existing tags">
        <item><ref id="removedtags" name="Removed tags">
</itemize>
<p>

<sect1>New tags<label id="newtags">
<p>
<descrip>
        <tag>adaptation_meta</tag>
        <p>This option allows Squid administrator to add custom ICAP request
           headers or eCAP options to Squid ICAP requests or eCAP transactions.

        <tag>adaptation_send_client_ip</tag>
        <p>Same as deprecated icap_send_client_ip
           but applies to both ICAP and eCAP.</p>

        <tag>adaptation_send_username</tag>
        <p>Same as deprecated icap_send_client_username
           but applies to both ICAP and eCAP.</p>

        <tag>adaptation_uses_indirect_client</tag>
        <p>Same as deprecated icap_uses_indirect_client
           but applies to both ICAP and eCAP.</p>

        <tag>client_delay_pools</tag>
        <p>New setting for client bandwidth limits to specifies the number 
          of client delay pools used.

        <tag>client_delay_initial_bucket_level</tag>
        <p>New setting for client bandwidth limits to determine the initial 
          bucket size as a percentage of  max_bucket_size from 
          client_delay_parameters.
            
        <tag>client_delay_parameters</tag>
        <p>New setting for client bandwidth limits to configures client-side 
           bandwidth limits.

        <tag>client_delay_access</tag>
        <p>New setting for client bandwidth limits to determines the 
          client-side delay pool for the request.

        <tag>client_dst_passthru</tag>
        <p>New setting to disable extra Host: header security on interception proxies.
           Impacts cache integrity/reliability and client browser security.
        <p><em>IMPORTANT:</em> disabling this directive only allows Squid to change the
           destination IP to another source indicated by Host: domain DNS or
           cache_peer configuration. It <em>does not</em> affect Host: validation.

        <tag>client_idle_pconn_timeout</tag>
        <p>Renamed from <em>persistent_request_timeout</em>.

        <tag>cpu_affinity_map</tag>
        <p>New setting for SMP support to map Squid processes onto specific CPU cores.

        <tag>connect_retries</tag>
        <p>Replacement for <em>maximum_single_addr_tries</em>, but instead of only applying to hosts with single addresses.
           This directive applies to all hosts, extending the number of connection attempts to each IP address.

        <tag>dns_packet_max</tag>
        <p>New setting to configure maximum number of bytes packet size to advertise via EDNS.
           Set to "none" (the initial default) to disable EDNS large packet support.

        <tag>else</tag>
        <p>Part of conditional SMP support syntax. see <em>if</em>

        <tag>endif</tag>
        <p>Part of conditional SMP support syntax. see <em>if</em>

        <tag>eui_lookup</tag>
        <p>Whether to lookup the EUI or MAC address of a connected client.

        <tag>host_verify_strict</tag>
        <p>New option to enable super-strict HTTP and DNS information match.
        Ensuring the HTTP URI details, DNS records, and TCP connection layers all match in a
        three-legged security verification. Preventing domain hijacking or malicious poisoning
        attacks by malicious scripts.
        <p>The default is to verify only intercepted traffic, to log all issues and let failed
        traffic through when doing so can be done safely.

        <tag>icap_206_enable</tag>
        <p>New option to toggle whether the ICAP 206 (Partial Content) responses extension.
           Default is on.

        <tag>if</tag>
        <p>New conditional syntax for SMP multiple-worker. 
        If-statements can be used to make configuration directives depend on conditions.
        <p>The else part is optional. The keywords <em>if</em>, <em>else</em> and <em>endif</em> 
        must be typed on their own lines, as if they were regular configuration directives.

        <tag>logfile_daemon</tag>
        <p>Ported from 2.7. Specify the file I/O daemon helper to run for logging.

        <tag>max_stale</tag>
        <p>Places an upper limit on how stale content Squid will serve from the cache if cache validation fails

        <tag>memory_cache_mode</tag>
        <p>Controls which objects to keep in the memory cache (cache_mem)
        <verb>
        'always'  Keep most recently fetched objects in memory (default)
        
        'disk'    Only disk cache hits are kept in memory, which means
                  an object must first be cached on disk and then hit
                  a second time before cached in memory.
        
        network   Only objects fetched from network is kept in memory
        </verb>

        <tag>memory_cache_shared</tag>
        <p>Controls whether the memory cache is shared among SMP workers.
        <p>Currently, entities exceeding 32KB in size cannot be shared.

        <tag>server_idle_pconn_timeout</tag>
        <p>Renamed from <em>pconn_timeout</em>.

        <tag>tproxy_uses_indirect_client</tag>
        <p>Controls whether the indirect client address found in the X-Forwarded-For
        header is used for spoofing instead of the directly connected client address.
        Requires both <em>--enable-follow-x-forwarded-for</em> and <em>--enable-linux-netfilter</em>

        <tag>workers</tag>
        <p>Number of main Squid processes or "workers" to fork and maintain.
        In SMP mode, each worker does nearly all what a single Squid daemon
        does (e.g., listen on http_port and forward HTTP requests).
        <verb>
        0: "no daemon" mode, like running "squid -N ..."
        1: "no SMP" mode, start one main Squid process daemon (default)
        N: start N main Squid process daemons (i.e., SMP mode)
        </verb>

        <tag>write_timeout</tag>
        <p>New setting to limit time spent waiting for data writes to be confirmed.
</descrip>

<sect1>Changes to existing tags<label id="modifiedtags">
<p>
<descrip>
        <tag>access_log</tag>
        <p>New <em>stdio</em> module to send log data directly from Squid to a disk file.
        This is the historic behaviour of Squid before logging modules were introduced, and
        remains the default used when no module is selected.
        It is recommended to upgrade logging to the faster <em>daemon:</em> module.
        <p>New <em>daemon</em> module to send each log line as text data to a file I/O daemon handling the slow disk I/O.
        New installs, or installs with no logs configured explicitly will use this module by default.
        <p>New <em>tcp</em> module to send each log line as text data to a TCP receiver.
        <p>New <em>udp</em> module to send each log line as text data to a UDP receiver.
        <p>New format <em>referrer</em> to log with the format previously used by referer_log directive.
        <p>New format <em>useragent</em> to log with the format previously used by useragent_log directive.

        <tag>acl : random, urllogin</tag>
        <p>New type <em>random</em>. Pseudo-randomly match requests based on a configured probability.
        <p>Ported <em>urllogin</em> option from Squid 2.7, to match a regex pattern on the URL login field (if any).
        <p>The <em>manager</em> ACL requires adjustment to cover new cache manager access. So it has now been
          built-in as a predefined ACL name matching URLs equivalent to the following regular expression:
        <verb>
                ^(cache_object://|https?://[^/]+/squid-internal-mgr/)
        </verb>
          squid.conf containing the old manager definition can expect to see ACL type collisions.

        <tag>auth_param</tag>
        <p>New options for Basic, Digest, NTLM, Negotiate <em>children</em> settings.
           <em>startup=N</em> determines minimum number of helper processes used.
           <em>idle=N</em> determines how many helper to retain as buffer against sudden traffic loads.
           <em>concurrency=N</em> previously called <em>auth_param ... concurrency</em> as a separate option.
        <p>Removed Basic, Digest, NTLM, Negotiate <em>auth_param ... concurrency</em> setting option.
        <p>Known Issue: NTLM and Negotiate protocols do not support concurrency. When set this option is ignored.

        <tag>cache_dir</tag>
        <p><em>min-size</em> option ported from Squid-2

        <tag>cache_peer</tag>
        <p><em>htcp-*</em> options collapsed into <em>htcp=</em> taking an optional comma-separated list of flags.
           The old form is deprecated but still accepted.

        <tag>cache_store_log</tag>
        <p>Now uses logging modules. Example: stdio:/file/path
           see <em>access_log</em> for a list of supported modules and their parameters.

        <tag>clientside_mark</tag>
        <p>New configuration parameter <em>clientside_mark</em>
        <p>Allows packets leaving Squid on the client side to be marked with a Netfilter mark value in the same way as the existing clientside_tos feature.
        <p>This feature is only available for Netfilter environments.

        <tag>deny_info</tag>
        <p>Support URL format tags. For dynamically generated URL in denial redirect.
        <p>Support the full range of 200-599 HTTP status codes.
           3xx status only available when redirecting to a URI.
           Other status only available when supplying an error template body.

        <tag>external_acl_type</tag>
        <p>New format tags and option parameters:
        <p><em>%SRCEUI48</em> EUI-48 / MAC address of client from ARP lookup.
        <p><em>%SRCEUI64</em> EUI-64 of clients with SLAAC address.
        <p><em>%EXT_LOG</em> log= message returned by previous external ACL calls. An updated version may be returned.
        <p><em>%EXT_TAG</em> tag= value returned by previous external ACL calls. Tag may not be altered once set.
        <p><em>children-max=N</em> determines maximum number of helper processes used.
        <p><em>children-startup=N</em> determines minimum number of helper processes used.
        <p><em>children-idle=N</em> determines how many helper to retain as buffer against sudden traffic loads.
        <p>Deprecated <em>children=N</em> in favor of <em>children-max=N</em>.

        <tag>http_port act-as-origin vhost no-vhost</tag>
        <p><em>act-as-origin</em> ported from 2.7.
           This option corrects several HTTP header issues when operating as a reverse proxy and cache.
           Notably the externally visible aging of objects stored in the server-side cache.
        <p><em>vhost</em> is deprecated. <em>accel</em> mode, reverse proxy, now defaults to always enable HTTP/1.1 virtual domain support.
        <p><em>no-vhost</em> option is added to disable the new reverse proxy behaviour.

        <tag>icap_send_client_ip</tag>
        <p>Deprecated in favor of adaptation_send_client_ip
        which applies to both ICAP and eCAP.</p>

        <tag>icap_send_client_username</tag>
        <p>Deprecated in favor of adaptation_send_username
        which applies to both ICAP and eCAP.</p>

        <tag>icap_uses_indirect_client</tag>
        <p>Deprecated in favor of adaptation_uses_indirect_client 
        which applies to both ICAP and eCAP.</p>

        <tag>logformat</tag>
        <p><em>%&lt;a</em> Server or Peer IP address from the last server connection (next hop).
        <p><em>%&gt;bs</em> Number of HTTP-equivalent message body bytes received from the next hop.
        <p><em>icap::%&gt;bs</em> Number of message body bytes received from the ICAP server.
        <p><em>%sn</em> Unique sequence number per log line. Ported from 2.7
        <p><em>%&gt;eui</em> EUI logging (EUI-48 / MAC address for IPv4, EUI-64 for IPv6).
           Both EUI forms are logged in the same field. Type can be identified by length or byte delimiter.
        <p><em>%err_code</em> The ID of an error response served by Squid or a similar internal error identifier
        <p><em>%err_detail</em> Additional err_code-dependent error information.
        <p><em>%&gt;la</em> Rename of %la to indicate being a client connection detail.
        <p><em>%&gt;lp</em> Rename of %lp to indicate being a client connection detail.
        <p><em>%&lt;p</em> Server or Peer port number from the last server connection (next hop).

        <tag>memory_pools_limit</tag>
        <p>Memory limits have been revised and corrected from 3.1.4 onwards.
        <p>Please check and update your squid.conf to use the text <em>none</em> for no limit instead of the old 0 (zero).
        <p>All users upgrading need to be aware that from Squid-3.3 setting this option to 0 (zero) will mean zero bytes of memory get pooled.

        <tag>qos_flows</tag>
        <p>New options <em>mark</em> and <em>tos</em> and <em>miss</em>
        <p><em>tos</em> retains the original QOS functionality of the IP header TOS field.
        <p><em>mark</em> offers the same functionality, but with a netfilter mark value.
        <p>These options should be placed immediately after qos_flows.
        <p>The <em>tos</em> value is optional in order to maintain backwards compatability.
        <p>The preserve-miss functionality is available with the <em>mark</em> option and requires no kernel patching.
           It does, however, require libnetfilter_conntrack.
           This will be included by default if available (see the --without-netfilter-conntrack configure option for more details).
        <p><em>miss</em> sets a value for a cache miss. It is available for both the tos and mark options and takes precedence over the preserve-miss feature.

        <tag>range_offset_limit</tag>
        <p>Added ACL support for control over when the limit applies and when it is avoided.

        <tag>refresh_pattern</tag>
        <p>New option <em>max-stale=</em> to provide a maximum staleness factor. Squid won't
           serve objects more stale than this even if it failed to validate the object.
        <p>Removed option <em>ignore-no-cache</em>. Its commonly desired behaviour is obsoleted
           by correct HTTP/1.1 Cache-Control:no-cache handling.

        <tag>reply_header_access</tag>
        <p>Added support for custom response header names.</p>

        <tag>request_header_access</tag>
        <p>Added support for custom request header names.</p>

        <tag>reply_header_replace</tag>
        <p>Added support for custom response header names.</p>

        <tag>request_header_replace</tag>
        <p>Added support for custom request header names.</p>

        <tag>tcp_outgoing_address</tag>
        <p>This parameter is now compatible with persistent server connections.
           The IPv6 magic 'to_ipv6' hacks needed in 3.1 are now no longer necessary.

        <tag>tcp_outgoing_mark</tag>
        <p>New configuration parameter <em>tcp_outgoing_mark</em>
        <p>Allows packets leaving Squid on the server side to be marked with a Netfilter mark value in the same way as the existing tcp_outgoing_tos feature.
        <p>This feature is only available for Netfilter environments.

        <tag>tcp_outgoing_tos</tag>
        <p>This parameter is now compatible with persistent server connections.

        <tag>url_rewrite_children</tag>
        <p>New options <em>startup=N</em>, <em>idle=N</em>, <em>concurrency=N</em>
        <itemize>
                <item>startup=N allow finer tuning of how many helpers are started initially.
                <item>idle=N allow fine tuning of how many helper to retain as buffer against sudden traffic loads.
                <item>concurrency=N was previously called url_rewrite_concurrency as a distinct directive.
        </itemize>

        <tag>windows_ipaddrchangemonitor</tag>
        <p>Now only available to be set in Windows builds.

</descrip>


<sect1>Removed tags<label id="removedtags">
<p>
<descrip>
        <tag>balance_on_multiple_ip</tag>
        <p>Obsolete. The behaviour controlled by this directive is no longer possible
           with DNS parallel lookup and connection features (aka 'Happy Eyeballs').

        <tag>chunked_request_body_max_size</tag>
        <p>Obsolete. Squid is now HTTP/1.1 with support for streaming chunked encoded requests.

        <tag>dns_v4_fallback</tag>
        <p>Obsolete. Replaced by DNS parallel lookups.

        <tag>emulate_httpd_log</tag>
        <p>Replaced by <em>common</em> format option on an <em>access_log</em> directive.

        <tag>forward_log</tag>
        <p>Obsolete.

        <tag>ftp_list_width</tag>
        <p>Obsolete.

        <tag>ignore_expect_100</tag>
        <p>Obsolete.

        <tag>log_fqdn</tag>
        <p>Obsolete. Replaced by automatic detection of the %>A logformat tag.

        <tag>log_ip_on_direct</tag>
        <p>Obsolete. Use a custom log with <em>%&lt;A</em> format tag to receive server FQDN or peer name.

        <tag>maximum_single_addr_tries</tag>
        <p>The behaviour controlled by this directive is no longer possible.
           It has been replaced by <em>connect_retries</em> option which operates a little differently.

        <tag>pconn_timeout</tag>
        <p>Renamed to <em>server_idle_pconn_timeout</em>

        <tag>persistent_request_timeout</tag>
        <p>Renamed to <em>client_idle_pconn_timeout</em>

        <tag>referer_log</tag>
        <p>Replaced by the <em>referrer</em> format option on an <em>access_log</em> directive.

        <tag>url_rewrite_concurrency</tag>
        <p>Replaced by url_rewrite_children ... concurrency=N option.

        <tag>useragent_log</tag>
        <p>Replaced by the <em>useragent</em> format option on an <em>access_log</em> directive.
</descrip>


<sect>Changes to ./configure options since Squid-3.1
<p>
There have been some changes to Squid's build configuration since Squid-3.1.

This section gives an account of those changes in three categories:

<itemize>
        <item><ref id="newoptions" name="New options">
        <item><ref id="modifiedoptions" name="Changes to existing options">
        <item><ref id="removedoptions" name="Removed options">
</itemize>


<sect1>New options<label id="newoptions">
<p>
<descrip>
        <tag>--enable-auth-basic[=HELPERS]</tag>
        <p>Specified without any parameters all helpers will be auto-built.
        <p>With an explicit empty list <em>=""</em> protocol support will be built but no helpers.
        <p>With an explicit list protocol support and just those helpers will be built.

        <tag>--enable-auth-digest[=HELPERS]</tag>
        <p>Specified without any parameters all helpers will be auto-built.
        <p>With an explicit empty list <em>=""</em> protocol support will be built but no helpers.
        <p>With an explicit list protocol support and just those helpers will be built.

        <tag>--enable-auth-negotiate</tag>
        <p>Specified without any parameters all helpers will be auto-built.
        <p>With an explicit empty list <em>=""</em> protocol support will be built but no helpers.
        <p>With an explicit list protocol support and just those helpers will be built.

        <tag>--enable-auth-ntlm</tag>
        <p>Specified without any parameters all helpers will be auto-built.
        <p>With an explicit empty list <em>=""</em> protocol support will be built but no helpers.
        <p>With an explicit list protocol support and just those helpers will be built.

        <tag>--enable-build-info</tag>
        <p>Add an additional string in the output of "squid -v".

        <tag>--enable-eui</tag>
        <p>Enable Support for handling EUI operations.
           This includes ARP lookups for MAC (EUI-48) addresses and the ACL arp type tests.

        <tag>--enable-log-daemon-helpers</tag>
        <p>Build helpers for logging I/O.

        <tag>--enable-url-rewrite-helpers</tag>
        <p>Build helpers for some basic URL-rewrite actions. For use by url_rewrite_program.
           If omitted or set to =all then all bundled helpers that are able to build will be built.
           If set to a specific list of helpers then only those helpers will build.
           Currently one demo helper <em>fake</em> is provided in shell and C++ forms to demonstrate
           the helper protocol usage and provide exemplar code.

        <tag>--with-swapdir=PATH</tag>
        <p>Location to display in documentation for the default cache.
           Updated to indicate /var/cache/squid in accordance with the filesystem layout standards.
           Squid-3 no longer builds an implicit disk cache at this location, so the change is not expected
           to have any effect on existing builds other than fixing some mysterious lack of core dumps.
           The old /var/cache location was often non-writable which blocked core dumps creation.

        <tag>--without-netfiler-conntrack</tag>
        <p>Disables the libnetfilter_conntrack library being used for the new qos_flows option <em>mark</em>.
           default is to auto-detect the library and use where available.
</descrip>

<sect1>Changes to existing options<label id="modifiedoptions">
<p>
<descrip>
        <tag>--enable-auth</tag>
        <p>No longer takes a list of arguments. This option now is restricted to building Squid with or without authentication support.
        <p>The new <em>--enable-auth-X</em>/<em>--disable-auth-X</em> parameters determine which authentication protocols and helpers are built.

</descrip>
</p>

<sect1>Removed options<label id="removedoptions">
<p>
<descrip>
        <tag>--enable-arp-acl</tag>
        <p>Replaced by --enable-eui

        <tag>--enable-auth-basic-helpers</tag>
        <p>Replaced by <em>--enable-auth-basic</em>.

        <tag>--enable-auth-digest-helpers</tag>
        <p>Replaced by <em>--enable-auth-digest</em>.

        <tag>--enable-auth-negotiate-helpers</tag>
        <p>Replaced by <em>--enable-auth-negotiate</em>.

        <tag>--enable-auth-ntlm-helpers</tag>
        <p>Replaced by <em>--enable-auth-ntlm</em>.

        <tag>--enable-referer-log</tag>
        <p>Obsolete.

        <tag>--enable-useragent-log</tag>
        <p>Obsolete.

</descrip>


<sect>Options Removed since Squid-2

<p>Some squid.conf and ./configure options which were available in Squid-2.6 and Squid-2.7 are made obsolete in Squid-3.2.

<sect1>Removed squid.conf options since Squid-2.7
<p>
<descrip>
        <tag>auth_param</tag>
        <p><em>blankpassword</em> option for basic scheme removed.

        <tag>authenticate_ip_shortcircuit_access</tag>
        <p>Not safe for general use.
          An external_acl_type helper may be used to bypass authentication if that is suitable.

        <tag>authenticate_ip_shortcircuit_ttl</tag>
        <p>Not safe for general use.
          An external_acl_type helper may be used to bypass authentication if that is suitable.

        <tag>cache_peer</tag>
        <p>Option <em>http11</em> obsolete.

        <tag>external_acl_type</tag>
        <p>Format tag <em>%{Header}</em> replaced by <em>%>{Header}</em>
        <p>Format tag <em>%{Header:member}</em> replaced by <em>%>{Header:member}</em>

        <tag>header_access</tag>
        <p>Replaced by <em>request_header_access</em> and <em>reply_header_access</em>

        <tag>http_port</tag>
        <p>Option <em>no-connection-auth</em> replaced by <em>connection-auth=[on|off]</em>. Default is ON.
        <p>Option <em>transparent</em> option replaced by <em>intercept</em>
        <p>Option <em>http11</em> obsolete.

        <tag>http_access2</tag>
        <p>Replaced by <em>adapted_http_access</em>

        <tag>httpd_accel_no_pmtu_disc</tag>
        <p>Replaced by <em>http_port disable-pmtu-discovery=</em> option

        <tag>incoming_rate</tag>
        <p>Obsolete.

        <tag>redirector_bypass</tag>
        <p>Replaced by <em>url_rewrite_bypass</em>

        <tag>server_http11</tag>
        <p>Obsolete.

        <tag>update_headers</tag>
        <p>Obsolete. The experimental actions enabled in 2.7 by this option have been integrated as default
           actions for the <em>rock</em> storage type and memory caches.
           The configuration option is no longer necessary and has been dropped.
           NOTE: It is not yet supported by <em>ufs</em>, <em>aufs</em>, or <em>diskd</em> storage.

        <tag>upgrade_http0.9</tag>
        <p>Obsolete.

        <tag>zph_local</tag>
        <p>Replaced by <em>qos_flows local-hit=</em>

        <tag>zph_mode</tag>
        <p>Obsolete.

        <tag>zph_option</tag>
        <p>Obsolete.

        <tag>zph_parent</tag>
        <p>Replaced by <em>qos_flows parent-hit=</em>

        <tag>zph_sibling</tag>
        <p>Replaced by <em>qos_flows sibling-hit=</em>

</descrip>

<sect1>Removed squid.conf options since Squid-2.6
<p>
<descrip>
        <tag>acl</tag>
        <p><em>urlgroup</em> type removed. Use <em>myportname</em> type instead.

        <tag>cache_dir</tag>
        <p><em>read-only</em> option replaced by <em>no-store</em>.

        <tag>http_port</tag>
        <p><em>urlgroup=</em> removed. Use <em>name=</em> feature instead.

        <tag>zero_buffers</tag>
        <p>Replaced by native support.

</descrip>

<sect1>Removed ./configure options since Squid-2.7
<p>
<descrip>
        <tag>--enable-coss-aio-ops</tag>
        <p>Obsolete.

        <tag>--enable-devpoll</tag>
        <p>Replaced by automatic detection.

        <tag>--enable-dlmalloc=LIB</tag>
        <p>Obsolete.

        <tag>--enable-epoll</tag>
        <p>Replaced by automatic detection.

        <tag>--enable-forward-log</tag>
        <p>Obsolete.

        <tag>--enable-heap-replacement</tag>
        <p>Obsolete.

        <tag>--enable-htcp</tag>
        <p>Obsolete. Enabled by default.

        <tag>--enable-large-cache-files</tag>
        <p>Obsolete.

        <tag>--enable-mempool-debug</tag>
        <p>Obsolete.

        <tag>--enable-multicast-miss</tag>
        <p>Obsolete.

        <tag>--enable-poll</tag>
        <p>Replaced by automatic detection.

        <tag>--enable-select</tag>
        <p>Replaced by automatic detection.

        <tag>--enable-select-simple</tag>
        <p>Replaced by automatic detection.

        <tag>--enable-snmp</tag>
        <p>Obsolete. Enabled by default.

        <tag>--enable-truncate</tag>
        <p>Obsolete.

        <tag>--disable-kqueue</tag>
        <p>Obsolete. Disabled by default.

        <tag>--without-system-md5</tag>
        <p>Obsolete. Disabled by default.

</descrip>


<sect>Regressions since Squid-2.7

<p>Some squid.conf and ./configure options which were available in Squid-2.7 are not yet available in Squid-3.2

<p>If you need something to do then porting one of these from Squid-2 to Squid-3 is most welcome.

<sect1>Missing squid.conf options available in Squid-2.7
<p>
<descrip>
        <tag>broken_vary_encoding</tag>
        <p>Not yet ported from 2.6

        <tag>cache_dir</tag>
        <p><em>COSS</em> storage type is lacking stability fixes from 2.6
        <p>COSS <em>overwrite-percent=</em> option not yet ported from 2.6
        <p>COSS <em>max-stripe-waste=</em> option not yet ported from 2.6
        <p>COSS <em>membufs=</em> option not yet ported from 2.6
        <p>COSS <em>maxfullbufs=</em> option not yet ported from 2.6

        <tag>cache_peer</tag>
        <p><em>idle=</em> not yet ported from 2.7
        <p><em>monitorinterval=</em> not yet ported from 2.6
        <p><em>monitorsize=</em> not yet ported from 2.6
        <p><em>monitortimeout=</em> not yet ported from 2.6
        <p><em>monitorurl=</em> not yet ported from 2.6

        <tag>cache_vary</tag>
        <p>Not yet ported from 2.6

        <tag>collapsed_forwarding</tag>
        <p>Not yet ported from 2.6

        <tag>error_map</tag>
        <p>Not yet ported from 2.6

        <tag>external_acl_type</tag>
        <p><em>%ACL</em> format tag not yet ported from 2.6
        <p><em>%DATA</em> format tag not yet ported from 2.6

        <tag>external_refresh_check</tag>
        <p>Not yet ported from 2.7

        <tag>ignore_ims_on_miss</tag>
        <p>Not yet ported from 2.7

        <tag>location_rewrite_access</tag>
        <p>Not yet ported from 2.6

        <tag>location_rewrite_children</tag>
        <p>Not yet ported from 2.6

        <tag>location_rewrite_concurrency</tag>
        <p>Not yet ported from 2.6

        <tag>location_rewrite_program</tag>
        <p>Not yet ported from 2.6

        <tag>refresh_pattern</tag>
        <p><em>stale-while-revalidate=</em> not yet ported from 2.7
        <p><em>ignore-stale-while-revalidate=</em> not yet ported from 2.7
        <p><em>negative-ttl=</em> not yet ported from 2.7

        <tag>refresh_stale_hit</tag>
        <p>Not yet ported from 2.7

        <tag>storeurl_access</tag>
        <p>Not yet ported from 2.7

        <tag>storeurl_rewrite_children</tag>
        <p>Not yet ported from 2.7

        <tag>storeurl_rewrite_concurrency</tag>
        <p>Not yet ported from 2.7

        <tag>storeurl_rewrite_program</tag>
        <p>Not yet ported from 2.7
        
</descrip>

<sect>Copyright
<p>
Copyright (C) 1996-2020 The Squid Software Foundation and contributors
<p>
Squid software is distributed under GPLv2+ license and includes
contributions from numerous individuals and organizations.
Please see the COPYING and CONTRIBUTORS files for details.

</article>