[squid-users] 3.5.0.2 core, constant under heavy load.

Tory M Blue tmblue at gmail.com
Thu Jun 11 19:51:35 UTC 2015


On Thu, Jun 11, 2015 at 12:25 PM, Eliezer Croitoru <eliezer at ngtech.co.il>
wrote:

> What is the issue??
> Did you tried the latest RPM's ??
> http://wiki.squid-cache.org/KnowledgeBase/CentOS
>
> Eliezer
>
>
I spun my own as there were no precompiled versions when I started with
3.5.x. I'll look at your rpms and test those, but in a nutshell after
putting a server in production I get the following errors, where squid
seems to restart

*FATAL: Received Segment Violation...dying.*

2015/06/11 11:25:00 kid1| Closing HTTP port [::]:80

2015/06/11 11:25:00 kid1| storeDirWriteCleanLogs: Starting...

2015/06/11 11:25:00 kid1|     65536 entries written so far.

2015/06/11 11:25:00 kid1|    131072 entries written so far.

2015/06/11 11:25:00 kid1|    196608 entries written so far.

2015/06/11 11:25:00 kid1|   Finished.  Wrote 246863 entries.

2015/06/11 11:25:00 kid1|   Took 0.06 seconds (4364312.99 entries/sec).

CPU Usage: 40.481 seconds = 24.353 user + 16.128 sys

Maximum Resident Size: 345808 KB
Page faults with physical i/o: 0

Which results in a core

Jun 11 11:21:55 cache03 squid[23658]: Squid Parent: will start 1 kids

Jun 11 11:21:55 cache03 squid[23658]: Squid Parent: (squid-1) process 23660
started

Jun 11 11:25:01 cache03 abrt[23710]: Saved core dump of pid 23660
(/usr/sbin/squid) to /var/spool/abrt/ccpp-2015-06-11-11:25:00-23660
(94121984 bytes)

Jun 11 11:25:01 cache03 abrtd: Directory 'ccpp-2015-06-11-11:25:00-23660'
creation detected

Jun 11 11:25:01 cache03 squid[23658]: Squid Parent: (squid-1) process 23660
exited due to signal 6 with status 0

Jun 11 11:25:02 cache03 kernel: Bridge firewalling registered

Jun 11 11:25:04 cache03 squid[23658]: Squid Parent: (squid-1) process 23918
started

Jun 11 11:25:18 cache03 abrtd: New problem directory
/var/spool/abrt/ccpp-2015-06-11-11:25:00-23660, processing


This runs fine with a single proc box in a low volume environment. This
does not seem to work in a multi core server in a very busy environment.
I've read some things and it appears others have had issues and your
suggestions and others was to try to run with -N, that doesn't seem to
affect my issue and I get the FATAL's and core dumps still


I'lll try your rpms first

Thanks

Tory
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.squid-cache.org/pipermail/squid-users/attachments/20150611/74c0ec79/attachment.html>


More information about the squid-users mailing list