Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP
Fetching contributors…

Cannot retrieve contributors at this time

171 lines (144 sloc) 5.865 kb
<?xml version="1.0" encoding="utf-8"?>
<!DOCTYPE group [
<!ENTITY mdash "&#8212;">
]>
<group from="2.1.3" to="2.1.4">
<subsystem>
<name>varnishd</name>
<change type="bug">
<para>An embarrasing typo in the new binary heap layout caused
inflated obj/objcore/objhdr counts and could cause odd problems
when the LRU expunge mechanism was invoked. This has been
fixed.</para>
</change>
<change type="enh">
<para>We now have updated documentation in the reStructuredText
format. Manual pages and reference documentation are both built
from this.</para>
</change>
<change type="enh">
<para>We now include a DNS director which uses DNS for choosing
which backend to route requests to. Please see the
documentation for more details.</para>
</change>
<change type="bug" ref="758">
<para>If you restarted a request, the HTTP header
<code>X-Forwarded-For</code> would be updated multiple times.
This has been fixed.</para>
</change>
<change type="bug" ref="742">
<para>If a VCL contained a % sign, and the <code>vcl.show</code>
CLI command was used, varnishd would crash. This has been
fixed.</para>
</change>
<change type="bug" ref="730">
<para>When doing a pass operation, we would remove the
<code>Content-Length</code>, <code>Age</code> and
<code>Proxy-Auth</code> headers. We are no longer doing
this.</para>
</change>
<change type="enh">
<para><code>now</code> has a string representation, making it
easier to construct <code>Expires</code> headers in VCL.</para>
</change>
<change type="bug" ref="739">
<para>In a high traffic environment, we would sometimes reuse a
file descriptor before flushing the logs from a worker thread to
the shared log buffer. This would cause confusion in some of
the tools. This has been fixed by explicitly flushing the log
when a backend connection is closed.</para>
</change>
<change type="bug" ref="744">
<para>If the communication between the management and the child
process gets out of sync, we have no way to recover.
Previously, <code>varnishd</code> would be confused, but we now
just kill the child and restart it.</para>
</change>
<change type="bug" ref="479">
<para>If the backend closes the connection on us just as we sent
a request to it, we retry the request. This should solve some
interoperability problems with Apache and the mpm-itk multi
processing module.</para>
</change>
<change type="enh">
<para><code>varnishd</code> now only provides help output the
current CLI session is authenticated for.</para>
</change>
<change type="bug" ref="733">
<para>If the backend does not tell us which length indication it
is using, we now assume the resource ends EOF at.</para>
</change>
<change type="enh">
<para>The client director now has a variable
<code>client.identity</code> which is used to choose which
backend should receive a given request.</para>
</change>
<change type="enh">
<para>The Solaris <code>port</code> waiter has been updated, and
other portability fixes for Solaris.</para>
</change>
<change type="bug" ref="746">
<para>There was a corner case in the close-down processing of pipes, this has now been fixed.</para>
</change>
<change type="bug">
<para>Previously, if we stopped polling a backend which was
sick, it never got marked as healthy. This has now been
changed.</para>
</change>
<change type="enh">
<para>It is now possible to specify ports as part of the <code>.host</code> field in VCL.</para>
</change>
<change type="bug" ref="768">
<para>The synthetic counters were not locked properly, and so
the <code>sms_</code> counters could underflow. This has now
been fixed.</para>
</change>
<change type="bug" ref="769">
<para>The value of <code>obj.status</code> as a string in vcl_error would not be correct in all cases. This has been fixed.</para>
</change>
<change type="bug" ref="776">
<para>Varnish would try to trim storage segments completely
filled when using the malloc stevedore and the object was
received chunked encoding. This has been fixed.</para>
</change>
<change type="bug" ref="763">
<para>If a buggy backend sends us a <code>Vary</code> header
with two colons, we would previously abort. We now rather fix
this up and ignore the extra colon.</para>
</change>
<change type="enh">
<para><code>req.hash_always_miss</code> and
<code>req.hash_ignore_busy</code> has been added, to make
preloading or periodically refreshing content work better.</para>
</change>
</subsystem>
<subsystem>
<name>varnishncsa</name>
<change type="bug">
<para><code>varnishncsa</code> would in some cases be confused
by ESI requests and output invalid lines. This has now been
fixed.</para>
</change>
</subsystem>
<subsystem>
<name>varnishlog</name>
<change type="enh">
<para><code>varnishlog</code> now allows <code>-o</code> and <code>-u</code> together.</para>
</change>
</subsystem>
<subsystem>
<name>varnishtop</name>
<change type="bug">
<para><code>varnishtop</code> would crash on 32 bit
architectures. This has been fixed.</para>
</change>
</subsystem>
<subsystem>
<name>libvarnishapi</name>
<change type="bug">
<para>Regex inclusion and exclusion had problems with matching
particular parts of the string being matched. This has been
fixed.</para>
</change>
</subsystem>
</group>
Jump to Line
Something went wrong with that request. Please try again.