MySQL Forums
Forum List  »  InnoDB

deadlock using lock tables with innodb on Athlon 64
Posted by: Jens Hennerich
Date: April 21, 2005 04:34AM

Hello all,

I'm running SuSE Linux 9.3 (Kernel 2.6.8-24-default) with MySQL 4.1.10a-3 on Athlon 64.

I have written an application which has to use both, MyISAM an InnoDB tables. To avoid deadlock the queries are surrounded with lock tables like described in the Manual (Section 7.5.9.7: How to cope with deadlocks?)
The query does something like: "replace process set updates=updates+1,pid='12345'..." and is called from different (up to 6) processes.

All works fine on MyISAM tables. The processes run without visiable waitstates.

If the tables are changed to InnoDB, the processes get stuck after a few seconds. The table remains locked, until all queries ended with an error result: "Lock wait timeout exceeded; try restarting transaction".

I tried different settings in my.cnf but the problem didn't change.

The same application didn't stuck or have any problem on a Athlon XP, MySQL 4.0.15-Max-log.
I suspect that this problem is somehow related to the 64 bit architecture.

Can anybody help me?
Thanks in advance!

Options: ReplyQuote


Subject
Views
Written By
Posted
deadlock using lock tables with innodb on Athlon 64
3290
April 21, 2005 04:34AM


Sorry, you can't reply to this topic. It has been closed.

Content reproduced on this site is the property of the respective copyright holders. It is not reviewed in advance by Oracle and does not necessarily represent the opinion of Oracle or any other party.