[Varnish] #1036: Cannot allocate memory in 3.0.2-rc1

[Varnish]

#1036: Cannot allocate memory in 3.0.2-rc1
----------------------+-----------------------------------------------------
 Reporter:  nicholas  |        Type:  defect
   Status:  new       |    Priority:  normal
Milestone:            |   Component:  build 
  Version:  trunk     |    Severity:  normal
 Keywords:            |  
----------------------+-----------------------------------------------------

Comment(by nicholas):

 Meh. We are using the fallback director to use the same appservers for
 login stuff across dns round-robin. We didn't notice that the fallback
 director was first introduced in >3.0.0. This makes it tricky to test
 3.0.0 in production.

 I reproduced the error in a test environment on centos5, using -s
 malloc,4M and 4 wgets crawling different parts of the sites. 3.0.0-2
 behaves as expected, 3.0.2-rc1 crashes with the same error:
 {{{
 Last panic at: Mon, 24 Oct 2011 09:18:37 GMT
 Assert error in VGZ_Ibuf(), cache_gzip.c line 222:
   Condition((vg->vz.avail_in) == 0) not true.
 errno = 12 (Cannot allocate memory)
 thread = (cache-worker)
 ident =
 Linux,2.6.18-194.32.1.el5xen,x86_64,-smalloc,-smalloc,-hcritbit,epoll
 Backtrace:
   0x42c1b6: /usr/sbin/varnishd [0x42c1b6]
   0x4223aa: /usr/sbin/varnishd(VGZ_Ibuf+0x7a) [0x4223aa]
   0x4229e9: /usr/sbin/varnishd [0x4229e9]
   0x4211cd: /usr/sbin/varnishd(FetchBody+0x3fd) [0x4211cd]
   0x414fb8: /usr/sbin/varnishd [0x414fb8]
   0x417686: /usr/sbin/varnishd(CNT_Session+0x9f6) [0x417686]
   0x42e9c8: /usr/sbin/varnishd [0x42e9c8]
   0x42dbab: /usr/sbin/varnishd [0x42dbab]
   0x379e40673d: /lib64/libpthread.so.0 [0x379e40673d]
   0x33822d3f6d: /lib64/libc.so.6(clone+0x6d) [0x33822d3f6d]
 ...
 }}}

 Should be reproducible? Looks like a regression in 3.0.2-rc1?

 :-)
 Nicholas

-- 
Ticket URL: <https://www.varnish-cache.org/trac/ticket/1036#comment:1>
Varnish <https://varnish-cache.org/>
The Varnish HTTP Accelerator