Squid 2.7.STABLE2 release notes
Squid Developers$Id: release.html,v 188.8.131.52 2008/05/31 21:23:33 hno Exp $
This document contains the release notes for version 2.7 of Squid.
Squid is a WWW Cache application developed by the Web Caching community.
This section describes the main news since the 2.6 release
- Experimental support for HTTP/1.1, mainly targeted at reverse proxy
installations. Not yet HTTP/1.1 compliant hoewever.
- A number of performance improvements; including request/reply parser,
eliminating various redundant data copies and some completely rewritten
- Support for WAIS has been removed.
- "act-as-origin" option for http_port - Squid can now emulate an origin
server when acting as an accelerator.
- "min-size" option for cache_dir - the minimum object size to store in
a cache directory. Previously objects of any size up to a "max-size"
maximum size would be considered as candidated for storing in a store_dir;
this option allows the administrator to tune various stores for small
and large objects rather than trying to tune it for both.
- Support for Solaris /dev/poll for network IO - more efficient than
poll() or select() and backwards compatible to Solaris 7. This must
be manually enabled during configure by specifying "--enable-devpoll".
- Support for FreeBSD accept filters. Use "accept_filter httpready"
in squid.conf to enable this.
- A semi-modular logging framework has been introduced, which both
allows for more efficient non-blocking logging with the supplied logging
daemon, but also allows for third-party modules to intercept the squid
logs and process them. An example "UDP" logging helper, thanks to the
Wikimedia Foundation, is included.
- Support for rewriting URLs into canonical forms when storing and retrieving
objects. A common practice seen in Content Delivery Networks is to serve
the same content from a variety of different URLs or hosts; this makes
efficient caching difficult. The store URL rewriting framework allows the
administrator to rewrite a variety of URLs into one canonical form, so
matching content from a variety of sources can be stored and retrieved
as if they came from the same source, whilst still fetching the content
from the original destination.
See the "storeurl_rewrite_program" option for more information, and
http://wiki.squid-cache.org/Features/StoreURLRewrite for some examples.
- Object revalidation can now occur in the background. Cache validation can
now occur in the background without requiring an active client to drive it.
Stale content being revalidated can be served in situ whilst the object
is being refreshed. See the "max_stale" and "refresh_pattern" options for more
- introduce a new option, "zero_buffers", which controls whether Squid will
zero the memory used for buffers and other data structures before use.
This may or may not improve performance on specific workloads.
- Cache authentication based on source IP address. This reduces the pressure
on external authenticators which may not be able to keep up under high load -
NTLM/winbind is a good example of this. See the "authenticate_ip_shortcircuit_access"
and "authenticate_ip_shortcircuit_ttl" options for more information.
- Support for configuration file includes has been added. "include" can now be
used to include a configuration file or a glob of configuration files in a
- The default rules to not cache dynamic content from cgi-bin and query URLs
have been altered. Previously, the "cache" ACL was used to mark requests
as non-cachable - this is enforced even on dynamic content which returns
cachability information. This has changed in Squid-2.7 to use the default
refresh pattern. Dynamic content is now cached if it is marked as cachable.
You should remove the default configuration lines with QUERY (acl, and cache)
and replace them with the correct refresh_pattern entries.
- Accelerator mode support cleaned up to behave more consistent when
combining multiple accelerator mode options
- Zero Penalty Hit support, allowing cache misses to be marked by custom
TOS/priority values, useful when using packet shaping/prioritization
outside Squid and needing to separate cache hits from misses.
This release has a number of changes and additions to squid.conf
- acl myportname
new acl matching the incoming port name
controls the new IP based authentication cache.
controls the Zero Penalty Hit support
optimization to skip updating on-disk headers
new log file daemon support
sas hardcoded to the first cache_dir
controls the new store URL rewrite functionality
controls the new builtin URL rewrite functionality
Experimental HTTP/1.1 support knobs
new helper to allow custom cache validations in accelerator setups
new way of using delay pools based on response size
optimization mainly targeted for accelerator setups
can now be set runtime. Was previously hardcoded at build time and further limited by ulimit
optimization to avoid waking Squid up until a request has been received
new tuning knob for high traffic conditions
tuning knob to disable a new optimization
Suggested defaults modified
the "read-only" option has been renamed to "no-store" to better reflect the functionality
new multicast-siblings option, enabling multicast ICP sibling relations
new idle=n option to keep a minimum pool of idle connections
new http11 option to enable experimental HTTP/1.1 support
New %URI format tag
Suggested defaults cleaned up, defines a new "localnet" acl with RFC1918 addresses
new "myportname" acl type matching the http_port name
Suggested defaults cleaned up, now requires configuration to use ICP
Suggested defaults cleaned up, now requires configuration to use HTCP
Suggested defaults cleaned up, using a new "localnet" acl.
Accelerator mode options cleaned up (accel, defaultsite, vport, vhost and combinations thereof)
new "allow-direct" option
new "act-as-origin" option
new "http11" option (experimental)
new "name=" option
nee "keepalive=" option
New format codes: oa (Our outgoing IP address), rp (Request URL-Path), sn (Unique sequence number)
Several new options: stale-while-revalidate, ignore-stale-while-revalidate, max-stale, negative-ttl
Suggested defaults adjusted to match the changes in the cache directive.
Future protocol change adding key=value pairs after the requests
Has several new modes, allowing one to finetune how/if the requesting
client IP should be forwarded in X-Forwarded-For
the above tuning knobs no longer have any effect and has been removed.
There is a few known issues and limitations in this release of Squid
- Bug #2248 storeurl_rewrite mismatched when object stored on memory
- Bug #2112 Squid does not send If-None-Match tag for cache revalidation
- Bug #2160 Cache hits on objects with headers > 4KB
- Bug #1059: mime.conf and referenced icons must be within chroot
- Bug #692: tcp_outgoing_address using an ident ACL does not work
- Bug #581: acl max_user_ip and multiple authentication schemes
- Bug #528: miss_access fails on "slow" acl types such as dst.
- Bug #513: squid -F is starting server sockets to early
- Bug #457: does not handle swap.state corruption properly
- Bug #410: unstable if runs out of disk space
- Bug #355: diskd may appear slow on low loads
- Bug #219: delay_pools stops working on -k reconfigure
This Squid version can run on Windows as a system service using the Cygwin emulation environment,
or can be compiled in Windows native mode using the MinGW + MSYS development environment. Windows NT 4 SP4 and later are supported.
On Windows 2000 and later the service is configured to use the Windows Service Recovery option
restarting automatically after 60 seconds.
Some new command line options was added for the Windows service support:
The service installation is made with -i command line switch, it's possible to use -f switch at
the same time for specify a different config-file settings for the Squid Service that will be
stored on the Windows Registry.
A new -n switch specify the Windows Service Name, so multiple Squid instance are allowed.
"Squid" is the default when the switch is not used.
So, to install the service, the syntax is:
squid -i [-f file] [-n name]
Service uninstallation is made with -r command line switch with the appropriate -n switch.
The -k switch family must be used with the appropriate -f and -n switches, so the syntax is:
squid -k command [-f file] -n service-name
where service-name is the name specified with -n options at service install time.
To use the Squid original command line, the new -O switch must be used ONCE, the syntax is:
squid -O cmdline [-n service-name]
If multiple service command line options must be specified, use quote. The -n switch is
needed only when a non default service name is in use.
Don't use the "Start parameters" in the Windows 2000/XP/2003 Service applet: they are
specific to Windows services functionality and Squid is not designed for understand they.
In the following example the command line of the "squidsvc" Squid service is set to "-D -u 3130":
squid -O "-D -u 3130" -n squidsvc
The process status helper functions make it easier for you to obtain information about
processes and device drivers running on Microsoft® Windows NT®/Windows® 2000. These
functions are available in PSAPI.DLL, which is distributed in the Microsoft® Platform
Software Development Kit (SDK). The same information is generally available through the
performance data in the registry, but it is more difficult to get to it. PSAPI.DLL is
PSAPI.DLL is available only on Windows NT, 2000, XP and 2003. The implementation in Squid is
aware of this, and try to use it only on the right platform.
On Windows NT PSAPI.DLL can be found as component of many applications, if you need it,
you can find it on Windows NT Resource KIT. If you have problem, it can be
downloaded from here:
On Windows 2000 and later it is available installing the Windows Support Tools, located on the
Support\Tools folder of the installation Windows CD-ROM.
On Windows platforms, if no value is specified in the dns_nameservers option on
squid.conf or in the /etc/resolv.conf file, the list of DNS name servers are
taken from the Windows registry, both static and dynamic DHCP configurations
- DISKD: still needs to be ported
- WCCP: cannot work because user space GRE support on Windows is missing
- Transparent Proxy: missing Windows non commercial interception driver
- Some code sections can make blocking calls.
- Some external helpers may not work.
- File Descriptors number hard-limited to 2048 when building with MinGW.
On Windows, cache manager (cachemgr.cgi) can be used with Microsoft IIS or Apache.
Some specific configuration could be needed:
IIS 6 (Windows 2003)
On IIS 6.0 all CGI extensions are denied by default for security reason, so the following configuration is needed:
- Create a cgi-bin Directory
- Define the cgi-bin IIS Virtual Directory with read and CGI execute IIS
permissions, ASP scripts are not needed. This automatically defines a
cgi-bin IIS web application
- Copy cachemgr.cgi into cgi-bin directory and look to file permissions:
the IIS system account and SYSTEM must be able to read and execute the file
- In IIS manager go to Web Service extensions and add a new Web Service
Extension called "Squid Cachemgr", add the cachemgr.cgi file and set the
extension status to Allowed
On Windows, cachemgr.cgi needs to create a temporary file, so Apache must be instructed
to pass the TMP and TEMP Windows environment variables to CGI applications:
ScriptAlias /squid/cgi-bin/ "c:/squid/libexec/"
PassEnv TMP TEMP
Allow from workstation.example.com
- Compile error if --enable-delaypools used
- Bug #1893: Variant invalidation support removed again, caused a lot content to not get cached.
- Bug #2350: Linux Capabilities version mismatch causing startup crash on newer kernels
- See also the list of
squid-2.7.STABLE2 changes and the
ChangeLog file for details.