[squid-users] Squid 2.5S1 dying unexpectedly

From: Lightfoot.Michael <Lightfoot.Michael@dont-contact.us>
Date: Mon, 3 Feb 2003 15:31:55 +1100

Several times in the last 8 days our main proxy has died 3 times with
the following logged in cache.log:

FATAL: Received Segment Violation...dying.
2003/02/03 14:49:30| storeDirWriteCleanLogs: Starting...
2003/02/03 14:49:31| 65536 entries written so far.
2003/02/03 14:49:32| 131072 entries written so far.
2003/02/03 14:49:32| 196608 entries written so far.
2003/02/03 14:49:33| 262144 entries written so far.
2003/02/03 14:49:33| 327680 entries written so far.
2003/02/03 14:49:34| 393216 entries written so far.
2003/02/03 14:49:34| 458752 entries written so far.
2003/02/03 14:49:34| WARNING: Closing open FD 27
2003/02/03 14:49:35| Finished. Wrote 462057 entries.
2003/02/03 14:49:35| Took 4.2 seconds (109876.8 entries/sec).
CPU Usage: 837.750 seconds = 529.470 user + 308.280 sys
Maximum Resident Size: 0 KB
Page faults with physical i/o: 110609
Memory usage for squid via mallinfo():
        total space in arena: 166322 KB
        Ordinary blocks: 165130 KB 12116 blks
        Small blocks: 0 KB 0 blks
        Holding blocks: 2736 KB 3 blks
        Free Small blocks: 0 KB
        Free Ordinary blocks: 1191 KB
        Total in use: 167866 KB 101%
        Total free: 1191 KB 1%

There is no core file being produced. The above crash happened during
working hours, but as some have occurred outside busy periods, I don't
think it is traffic-related.

Details of the build being used are:

Squid Cache: Version 2.5.STABLE1-20021118
configure options: --enable-dlmalloc --enable-gnuregex
--enable-async-io=160 --enable-useragent-log --enable-referer-log
--disable-wccp --enable-snmp --enable-poll
--enable-err-languages=English '--enable-removal-policies=lru heap'

Together with:
SmartFilter Info: SmartFilter Version: 3.1.1.02
SmartFilter Info: SmartFilter API version: 3.1.3

Via the SmarFilter patches supplied by Secure Computing.

It was all built using gcc 3.2 (with the Sun linker) on Solaris 7.

A similar build without the SF patches and running on Solaris 2.6 hasn't
had this problem (it is the parent of the above in our DMZ.) In
addition, this symptom did not occur prior to 8 days ago (we have been
running the current config since December 22 and the only changes to SF
have been the site.txt file and weekly control list updates.)

Has anyone seen this symptom? In case it is SF-related I am also asking
Secure Computing.

Michael Lightfoot
Unix Consultant
ISG Host Systems
Comcare
+61 2 62750680

------------------------------------------------------------------------
NOTICE: This e-mail message and attachments may contain confidential
information. If you are not the intended recipient you should not use or
disclose any information in the message or attachments. If received in
error, please notify the sender by return email immediately. Comcare
does not waive any confidentiality or privilege.
Received on Sun Feb 02 2003 - 21:32:05 MST

This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 17:13:13 MST