MySQL Bugs: #48661: One table only becomes permanently locked after 2-3 successful write operations

Bug #48661	One table only becomes permanently locked after 2-3 successful write operations
Submitted:	10 Nov 2009 10:37	Modified:	11 Feb 2018 11:08
Reporter:	Steve Willett	Email Updates:
Status:	Can't repeat	Impact on me:	None
Category:	MySQL Server: DML	Severity:	S2 (Serious)
Version:	14.12 Distrib 5.0.26	OS:	Solaris (Sol 10 5/08)
Assigned to:		CPU Architecture:	Any

Description:
One table only becomes permanently locked after 2-3 successful write operations.

All other tables in the same database are unaffected, the user performing the operation is the same on all tables (ie. User grant permissions have been eliminated as the issue).

Once a write statement (We use update/insert/delete/replace into) causes the table to lock (This can be seen from 'show full processlist'), it cannot be unlocked, the memory usage of the MySQL daemon increases without control until the 'ulimit' has been reached (Normally memory usage is 400-500Mb), CPU usage increases to 100% of 1 CPU - the MySQL daemon has to be restarted to recover the situation.

Th table has 1.7 million records (Some of the other tables without the problem are larger).

The client program is a python script, however executing the same statements manually through the MySQL command line interface also shows the same problem - once the table becomes locked.

How to repeat:
Restart the daemon & send a few write requests.

Thank you for the problem report. Please, try to repeat with a newer version, 5.0.86 at least. In case of the same problem, please, send the results of SHOW CREATE TABLE and SHOW TABLE STATUS for the problmatic table.

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".

Posted by developer:
 
Closing since it has been suspended for 8 years.