>
> One of our squid listeners will not stay alive. Typically, it will run
> as follows:
>
> # /usr/sbin/squid -f /etc/squid/squid_floor.conf -D -N -d 9 -u 0
> 2005/03/09 08:40:22| Starting Squid Cache version 2.5.STABLE3 for
> i386-redhat-linux-gnu...
> 2005/03/09 08:40:22| Process ID 3476
> 2005/03/09 08:40:22| With 1024 file descriptors available
> 2005/03/09 08:40:22| DNS Socket created at 0.0.0.0, port 50256, FD 4
> 2005/03/09 08:40:22| Adding nameserver 127.0.0.1 from /etc/resolv.conf
> 2005/03/09 08:40:22| Adding nameserver 216.111.65.217 from
> /etc/resolv.conf
> 2005/03/09 08:40:22| Adding nameserver 205.171.16.250 from
> /etc/resolv.conf
> 2005/03/09 08:40:22| helperOpenServers: Starting 20
> 'wrapzap_floor' processes
> 2005/03/09 08:40:22| helperOpenServers: Starting 5 'smb_auth'
> processes
> 2005/03/09 08:40:22| Referer logging is disabled.
> 2005/03/09 08:40:22| Unlinkd pipe opened on FD 35
> 2005/03/09 08:40:22| Swap maxSize 10240000 KB, estimated
> 787692 objects
> 2005/03/09 08:40:22| Target number of buckets: 39384
> 2005/03/09 08:40:22| Using 65536 Store buckets
> 2005/03/09 08:40:22| Max Mem size: 131072 KB
> 2005/03/09 08:40:22| Max Swap size: 10240000 KB
> 2005/03/09 08:40:22| Store logging disabled
> 2005/03/09 08:40:22| Rebuilding storage in /u03/cache_floor (DIRTY)
> 2005/03/09 08:40:22| Using Least Load store dir selection
> 2005/03/09 08:40:22| Set Current Directory to /var/spool/squid_floor
> 2005/03/09 08:40:22| Loaded Icons.
> 2005/03/09 08:40:23| Accepting HTTP connections at 10.1.4.51,
> port 3129, FD 34.
> 2005/03/09 08:40:23| Accepting SNMP messages on port 3402, FD 36.
> 2005/03/09 08:40:23| Outgoing SNMP messages on port 3402, FD 37.
> 2005/03/09 08:40:23| WCCP Disabled.
> 2005/03/09 08:40:23| Ready to serve requests.
> 2005/03/09 08:40:23| Done scanning /u03/cache_floor swaplog
> (0 entries)
> 2005/03/09 08:40:23| Finished rebuilding storage from disk.
> 2005/03/09 08:40:23| 0 Entries scanned
> 2005/03/09 08:40:23| 0 Invalid entries.
> 2005/03/09 08:40:23| 0 With invalid flags.
> 2005/03/09 08:40:23| 0 Objects loaded.
> 2005/03/09 08:40:23| 0 Objects expired.
> 2005/03/09 08:40:23| 0 Objects cancelled.
> 2005/03/09 08:40:23| 0 Duplicate URLs purged.
> 2005/03/09 08:40:23| 0 Swapfile clashes avoided.
> 2005/03/09 08:40:23| Took 0.4 seconds ( 0.0 objects/sec).
> 2005/03/09 08:40:23| Beginning Validation Procedure
> 2005/03/09 08:40:23| Completed Validation Procedure
> 2005/03/09 08:40:23| Validated 0 Entries
> 2005/03/09 08:40:23| store_swap_size = 0k
> 2005/03/09 08:40:23| storeLateRelease: released 0 objects
> Terminated
- Check your OS (error) logs for info concerning
abnormal process termination.
(/var/log/messages)
Check :
% dmesg
too.
>
>
> One time, it also gave the following:
>
> 2005/03/09 07:55:28| storeLateRelease: released 0 objects
> 2005/03/09 07:55:30| Preparing for shutdown after 0 requests
> 2005/03/09 07:55:30| Waiting 30 seconds for active
> connections to finish
> 2005/03/09 07:55:30| FD 34 Closing HTTP connection
> 2005/03/09 07:55:30| FD 36 Closing SNMP socket
> 2005/03/09 07:55:31| Shutting down...
> 2005/03/09 07:55:31| FD 37 Closing SNMP socket
> 2005/03/09 07:55:31| Closing unlinkd pipe on FD 35
> 2005/03/09 07:55:31| storeDirWriteCleanLogs: Starting...
> 2005/03/09 07:55:31| Finished. Wrote 0 entries.
> 2005/03/09 07:55:31| Took 0.0 seconds ( 0.0 entries/sec).
>
> Does anyone have experience with this issue?
>
This one seems to be an initiated shutdown, (squid -k shutdown);
check whether :
- no unauthorized system users are active
- all cronjobs e.d. are fully 'understood' and managed
by you.
M.
