Potential Coredump: message

From: Apiset Tananchai <aet@dont-contact.us>
Date: Fri, 22 May 1998 19:06:10 +0700 (ICT)

Hi,

I found a lot of this message (70-80/day) in my cache.log of my
squid1.2b20:

1998/05/22 01:21:04| pumpRestart: YES!
1998/05/22 01:21:04| Potential Coredump: httpRestart POST http://www.download.com/PC/Result/TitleList/1,2,0-a-0-0-b-1,00.html

and sometime right after above message, squid die with Segment Violation:

1998/05/21 21:02:18| pumpRestart: YES!
1998/05/21 21:02:18| Potential Coredump: httpRestart POST http://www.gay-world.com/chester/showdir.cgi
FATAL: Received Segment Violation...dying.
1998/05/21 21:02:42| storeDirWriteCleanLogs: Starting...
1998/05/21 21:02:42| WARNING: Closing open FD 51
1998/05/21 21:02:43| 65536 entries written so far.

but sometime squid die with Segment Violation without getting "Potential
Coredump" message first.

Is this a known problem?

Linux 2.0.34pre12 + 3000fd patch
PentiumII-300
284MB Ram (to be 384 soon)
1 x AIC-7880 onboard SCSI Controller
2 x AHA-2940UW SCSI Controller
4 x 4.5GB UW SCSI Harddisk (barrarcuda) for swap
1 x 2.5GB UW SCSI Harddisk (barrarcuda) for os/log files
squid-1.2b20 + asyncio.patch + asyncio_cache.patch

Squid is running in transparent proxy mode.

Thanks for you help,

--
aet
Received on Fri May 22 1998 - 05:00:11 MDT

This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:40:19 MST