Re: [squid-users] Squid exited due to signal 6 frequently

Axel.Mueller@t-systems.com wrote:

>Hi!
>
>We have an problem with our Squid Cache: Version 2.4.STABLE6. It was running over 2 years without problems, now we have much of reboots of squid, here's the log from the past 20 min:
>
>Jul 21 15:47:58 host1 squid[23828]: Squid Parent: child process 24727 started
>Jul 21 15:53:12 host1 squid[23828]: Squid Parent: child process 24727 exited due to signal 6
>Jul 21 15:53:15 host1 squid[23828]: Squid Parent: child process 24978 started
>Jul 21 15:54:39 host1 squid[23828]: Squid Parent: child process 24978 exited due to signal 6
>Jul 21 15:54:42 host1 squid[23828]: Squid Parent: child process 25041 started
>Jul 21 15:57:26 host1 syslog-ng[6589]: STATS: dropped 0
>Jul 21 16:00:36 host1 squid[23828]: Squid Parent: child process 25041 exited due to signal 6
>Jul 21 16:00:39 host1 squid[23828]: Squid Parent: child process 25166 started
>Jul 21 16:01:42 host1 squid[23828]: Squid Parent: child process 25166 exited due to signal 6
>Jul 21 16:01:45 host1 squid[23828]: Squid Parent: child process 25170 started
>
>
>He is starting again and again...
>
>This is what cache.log says for the last crash:
>
>2005/07/21 16:00:41| Completed Validation Procedure
>2005/07/21 16:00:41| Validated 213254 Entries
>2005/07/21 16:00:41| store_swap_size = 2831376k
>2005/07/21 16:00:42| storeLateRelease: released 3 objects
>2005/07/21 16:01:45| Starting Squid Cache version 2.4.STABLE6 for i386-debian-linux-gnu...
>2005/07/21 16:01:45| Process ID 25170
>2005/07/21 16:01:45| With 1024 file descriptors available
>2005/07/21 16:01:45| DNS Socket created on FD 5
>2005/07/21 16:01:45| Adding nameserver 127.0.0.1 from /etc/resolv.conf
>2005/07/21 16:01:45| User-Agent logging is disabled.
>2005/07/21 16:01:45| Referer logging is disabled.
>2005/07/21 16:01:45| Unlinkd pipe opened on FD 10
>2005/07/21 16:01:45| Swap maxSize 3145728 KB, estimated 241979 objects
>2005/07/21 16:01:45| Target number of buckets: 12098
>2005/07/21 16:01:45| Using 16384 Store buckets
>2005/07/21 16:01:45| Max Mem size: 16384 KB
>2005/07/21 16:01:45| Max Swap size: 3145728 KB
>2005/07/21 16:01:45| Local cache digest enabled; rebuild/rewrite every 3600/3600 sec
>2005/07/21 16:01:45| Store logging disabled
>2005/07/21 16:01:45| Rebuilding storage in /var/spool/squid (DIRTY)
>2005/07/21 16:01:45| Using Least Load store dir selection
>2005/07/21 16:01:45| Set Current Directory to /var/spool/squid
>2005/07/21 16:01:45| Loaded Icons.
>2005/07/21 16:01:46| Accepting HTTP connections at 0.0.0.0, port 3128, FD 11.
>2005/07/21 16:01:46| HTCP Disabled.
>2005/07/21 16:01:46| WCCP Disabled.
>2005/07/21 16:01:46| Ready to serve requests.
>2005/07/21 16:01:46| Store rebuilding is 1.8% complete
>2005/07/21 16:01:47| Done reading /var/spool/squid swaplog (231246 entries)
>2005/07/21 16:01:47| Finished rebuilding storage from disk.
>2005/07/21 16:01:47| 222277 Entries scanned
>2005/07/21 16:01:47| 0 Invalid entries.
>2005/07/21 16:01:47| 0 With invalid flags.
>2005/07/21 16:01:47| 213308 Objects loaded.
>2005/07/21 16:01:47| 0 Objects expired.
>2005/07/21 16:01:47| 8969 Objects cancelled.
>2005/07/21 16:01:47| 0 Duplicate URLs purged.
>2005/07/21 16:01:47| 0 Swapfile clashes avoided.
>2005/07/21 16:01:47| Took 1.9 seconds (114771.2 objects/sec).
>2005/07/21 16:01:47| Beginning Validation Procedure
>2005/07/21 16:01:47| Completed Validation Procedure
>2005/07/21 16:01:47| Validated 213308 Entries
>2005/07/21 16:01:47| store_swap_size = 2830772k
>2005/07/21 16:01:47| storeLateRelease: released 0 objects
>
>Any Ideas whats happen there?
>
>
>--
>Mit freundlichen Gruessen!
>Axel Mueller
>
> +------------------------------------------------------------+
> Axel Müller
> ICT - Services
> T-Systems GEI GmbH
> Service Line Systems Integration
> IBU Public and Healthcare
>
> Goslarer Ufer 35, 10589 Berlin, Germany
> Telefon: +49 30 3497-1859
> FAX: +49 30 3497-1177
> E-Mail: axel.mueller@T-Systems.com
> Internet: http://www.t-systems.com
>
>
> T-Systems ist eine Division der Deutschen Telekom
> +------------------------------------------------------------+
> The only problem with mornings is that they happen too early in the day
>
>
>
>
>-----Ursprüngliche Nachricht-----
>Von: Gaylord Van Brocklin [mailto:vanbrockling@saic.com]
>Gesendet: Donnerstag, 21. Juli 2005 01:12
>An: squid-users@squid-cache.org
>Betreff: [squid-users] Interception caching with Cisco routers?
>
>I have Interception caching working when my Squid servers are listening on port 80, but is there a way to do it on the Cisco router when my Squid boxes are listening on port 8080?
>
>-gvb
>
>
>
Check the log files about 2Gb or if the system go out of resources so
squid can not write to the log files.
Received on Sat Jul 23 2005 - 09:24:46 MDT