MySQL Bugs: #15857: MYSQL crash

Bug #15857	MYSQL crash
Submitted:	19 Dec 2005 11:42	Modified:	20 Jan 2006 11:14
Reporter:	Isabel Garcia Lorenzo	Email Updates:
Status:	No Feedback	Impact on me:	None
Category:	MySQL Server	Severity:	S3 (Non-critical)
Version:	4.1.11	OS:	Linux (Linux 2.4.21-27.0.4)
Assigned to:		CPU Architecture:	Any

Description:
The problem is that mysqld crashes without any trace. We have the red hat linux operating system with clumanager-1.2.22-2 cluster and one mysql instance for each service. We installed the mysql with the rpm binaries.This cluster detects that mysql crash and immediately restart the mysql. After that the mysql is OK.
The last crash was at 17:24. It seems like mysqld shutdown normally... I attach the logs 

051216 06:00:02  mysqld started
051216  6:00:02 [Warning] Asked for 196608 thread stack, but got 126976
051216  6:00:02  InnoDB: Started; log sequence number 0 2665248
/usr/sbin/mysqld: ready for connections.
Version: '4.1.11-standard'  socket: '/tmp/mysqlFIXBridge.sock'  port: 3310  MySQL Community Edition - Standard (GPL)
051216 17:24:07 [Note] /usr/sbin/mysqld: Normal shutdown

051216 17:24:07  InnoDB: Starting shutdown...
051216 17:24:09  InnoDB: Shutdown completed; log sequence number 0 2665248
051216 17:24:09 [Note] /usr/sbin/mysqld: Shutdown complete

051216 17:24:09  mysqld ended

051216 17:24:13  mysqld started
051216 17:24:13 [Warning] Asked for 196608 thread stack, but got 126976
051216 17:24:13  InnoDB: Started; log sequence number 0 2665248
/usr/sbin/mysqld: ready for connections.
Version: '4.1.11-standard'  socket: '/tmp/mysqlFIXBridge.sock'  port: 3310  MySQL Community Edition - Standard (GPL)
051216 18:45:04 [Note] /usr/sbin/mysqld: Normal shutdown

051216 18:45:04  InnoDB: Starting shutdown...
051216 18:45:07  InnoDB: Shutdown completed; log sequence number 0 2665248
051216 18:45:07 [Note] /usr/sbin/mysqld: Shutdown complete

051216 18:45:07  mysqld ended

051216 19:10:01  mysqld started
051216 19:10:01 [Warning] Asked for 196608 thread stack, but got 126976
051216 19:10:01  InnoDB: Started; log sequence number 0 2665248
/usr/sbin/mysqld: ready for connections.
Version: '4.1.11-standard'  socket: '/tmp/mysqlFIXBridge.sock'  port: 3310  MySQL Community Edition - Standard (GPL)
051216 19:10:09 [Note] /usr/sbin/mysqld: Normal shutdown

051216 19:10:09  InnoDB: Starting shutdown...
051216 19:10:11  InnoDB: Shutdown completed; log sequence number 0 2665248
051216 19:10:11 [Note] /usr/sbin/mysqld: Shutdown complete

051216 19:10:11  mysqld ended

How to repeat:
This bug is not exactly repeatable, although it has happened more than once.

Thank you for a problem report. First of all, can you try to upgrade to 4.1.16? You 4.1.11 is really old. You can upload the (resolved) MySQL stack trace either in a different comment or using the File tab.

Please, sewnd your my.cnf setting and describe your hardware also.

Hardware information: Blade Center HS40 (two nodes in cluster) with IBM DS4300 storage.

The mysql crash was for instance 1.

my.cnf file:

[mysqld_multi]
mysqld=/usr/bin/mysqld_safe
mysqladmin=/usr/bin/mysqladmin

[mysqld1]
socket=/tmp/mysqlFIXBridge.sock
port=3310
pid_file=/tmp/mysqlFIXBridge.pid
datadir=/data/programs/fix/db
max_connections=200

[mysqld2]
socket=/tmp/mysqlNSCMxOFeedHandler.sock
port=3311
pid_file=/tmp/mysqlNSCMxOFeedHandler.pid
datadir=/data/programs/ticker-plant/NSCMxO/db
max_heap_table_size=96000000
max_connections=200
[mysqld3]
socket=/tmp/mysqlNSCMxP.sock
port=3312
pid_file=/tmp/mysqlNSCMxP.pid
datadir=/data/programs/ticker-plant/NSCMxP/db
max_heap_table_size=192000000
[mysqld4]
socket=/tmp/mysqlMEFFDriver.sock
port=3313
pid_file=/tmp/mysqlMEFFDriver.pid
datadir=/data/programs/drivers/MEFFDriver/db

I've tried to update with 4.1.16 version but the rpms has the BUG 15223

Thanks in advanced

In fact, your error log snippet contains no evidence of any MySQL crashes. Looks like it was simply shutdown externally, possibly - by your clustering software, for unknown reason:

051216 17:24:07 [Note] /usr/sbin/mysqld: Normal shutdown

051216 17:24:07  InnoDB: Starting shutdown...
051216 17:24:09  InnoDB: Shutdown completed; log sequence number 0 2665248
051216 17:24:09 [Note] /usr/sbin/mysqld: Shutdown complete

You will not see "Normal shutdown" in case of crash. So, please, check your clustering software - may be, it has some logs that describe this situation in more details. Can you try to run MySQL withouth that clustering solution, for some time?

What I really need to know about your hardware is the amount of RAM and df -k command results.

No feedback was provided for this bug for over a month, so it is
being suspended automatically. If you are able to provide the
information that was originally requested, please do so and change
the status of the bug back to "Open".