[squid-users] Squid 4.3 assertion failed

Alex Rousskov rousskov at measurement-factory.com
Wed Oct 24 17:27:38 UTC 2018


On 10/24/18 7:21 AM, Julian Perconti wrote:

> Program received signal SIGHUP, Hangup.

This is normal. Ideally, you should configure your gdb to ignore SIGHUP
and pass that signal to Squid.

> The crash (assertion failed)
> 
> 2018/10/24 09:44:29| assertion failed: http.cc:1530: "!Comm::MonitorsRead(serverConnection->fd)"
> 
> Program received signal SIGABRT, Aborted.
> __GI_raise (sig=sig at entry=6) at ../sysdeps/unix/sysv/linux/raise.c:51
> 51	../sysdeps/unix/sysv/linux/raise.c: No such file or directory.
> #0  __GI_raise (sig=sig at entry=6) at ../sysdeps/unix/sysv/linux/raise.c:51
> #1  0x00007ffff40fb42a in __GI_abort () at abort.c:89
> #2  0x000055555573c57c in xassert (
>     msg=msg at entry=0x555555aec788 "!Comm::MonitorsRead(serverConnection->fd)", 
>     file=file at entry=0x555555aeb5c5 "http.cc", line=line at entry=1530)
>     at debug.cc:618
> #3  0x00005555557a1118 in HttpStateData::maybeReadVirginBody (
>     this=0x5555574c6888) at http.cc:1530


> The assertion seems to be similiar to bug
> https://bugs.squid-cache.org/show_bug.cgi?id=4896  as Alex said; but
> I am not sure. Should I post there anyway?

Thank you for posting your stack trace there. You did the right thing.


> By other hand, what can I do with those squid's crashes?

Please try the patch for bug 4864. That bug may be a duplicate of bug
4896, and it has an unpolished/unofficial fix:

  https://bugs.squid-cache.org/show_bug.cgi?id=4864

Please update bug 4896 to let us know whether bug 4864 fix worked for
you (including cases where the posted patch no longer applies).


Thank you,

Alex.


More information about the squid-users mailing list