Joe wrote:
A brand new Dell server using Red Hat connect in production worked fine for a few months when suddenly start crashing 3 times per day.
Here are a few messages generated every hour:
2007/04/04 17:19:07| Unlinkd pipe opened on FD 21
2007/04/04 17:19:07| Swap maxSize 2048000 KB, estimated 157538 objects
2007/04/04 17:19:07| Target number of buckets: 7876
2007/04/04 17:19:07| Using 8192 Store buckets
2007/04/04 17:19:07| Max Mem size: 8192 KB
2007/04/04 17:19:07| Max Swap size: 2048000 KB
2007/04/04 17:19:07| Rebuilding storage in /usr/local/squid/var/cache (DIRTY)
2007/04/04 17:19:07| Using Least Load store dir selection
2007/04/04 17:19:07| Set Current Directory to /usr/local/squid/var/cache
2007/04/04 17:19:07| Loaded Icons.
2007/04/04 17:19:07| Accepting HTTP connections at 0.0.0.0, port 3128, FD 23.
2007/04/04 17:19:07| Accepting ICP messages at 0.0.0.0, port 3130, FD 24.
2007/04/04 17:19:07| WCCP Disabled.
2007/04/04 17:19:07| Ready to serve requests.
2007/04/04 17:19:08| WARNING: newer swaplog entry for dirno 0, fileno 00000022
2007/04/04 17:19:08| Store rebuilding is 16.4% complete
2007/04/04 17:19:10| WARNING: newer swaplog entry for dirno 0, fileno 0000007B
2007/04/04 17:19:10| WARNING: newer swaplog entry for dirno 0, fileno 00001CA1
2007/04/04 17:19:10| WARNING: newer swaplog entry for dirno 0, fileno 00004989
2007/04/04 17:19:10| Done reading /usr/local/squid/var/cache swaplog (25532 entri
007/04/04 17:20:16| client_side_request.cc(848) setting body_reader = NULL for request 0xa3502cc
2007/04/04 17:20:44| WARNING: 1 swapin MD5 mismatches
2007/04/04 17:23:10| client_side_request.cc(848) setting body_reader = NULL for request 0xa351960
2007/04/04 17:26:17| client_side_request.cc(848) setting body_reader = NULL for request 0xa3502cc
2007/04/04 17:27:11| client_side_request.cc(848) setting body_reader = NULL for request 0xa354688
2007/04/04 17:28:15| assertion failed: tunnel.cc:372: "amount == (size_t)len"
2007/04/04 17:28:18| Starting Squid Cache version 3.0.PRE5 for i686-pc-linux-gnu...
2007/04/04 17:28:18| Process ID 3483
2007/04/04 17:28:18| With 1024 file descriptors available
2007/04/04 17:28:23| Done reading /usr/local/squid/var/cache swaplog (26511 entries)
2007/04/04 17:28:23| Finished rebuilding storage from disk.
A script is running every night to rotate the logs, plus another one to delete and rebulid the cache.
Need a hint on how to find out why squid is down.
Thnaks
Joe.
Looks like you have run into bug #1898. Luckily there appears to be a
patch for it now.
see bugzilla for details.
http://www.squid-cache.org/bugs/show_bug.cgi?id=1898
If you still encounter the problem after using the patch please report
it in the bugzilla for tracking.
Amos
