Bug #51544 select (no locking) fails with "lock wait timeout exceeded"

• Submitted: 26 Feb 2010 9:39
• Modified: 3 Nov 2010 12:36
• Reporter: Johan Andersson
• Status: Can't repeat
• Category: MySQL Cluster: Cluster (NDB) storage engine
• Severity: S3 (Non-critical)
• Version: 7.0.*
• OS: Any
• Assigned to: Jonas Oreland
• CPU Architecture: Any
• Tags: lock timeout, select timeout

[26 Feb 2010 9:39] Johan Andersson

Description:
I have data and query to reproduce this on:
MySQL Cluster 7.0.9b 
MySQL Cluster 7.0.11b 

ERROR 1205 (HY000): Lock wait timeout exceeded; try restarting transaction

You need to have:
4 data nodes
1 mysql server (i had ndb_cluster_connection_pool=4)

See private comments for query and config.ini

How to repeat:
see private comment

[26 Feb 2010 12:00] Jonas Oreland

Hi Johan,

Why have you changed BatchSizePerLocalScan (etc)
- was that to try to avoid the problem, or for some other reason
  if it was for "some other reason", can you try removing it, and
  see if it affects behavior

[31 Aug 2010 11:08] Jonas Oreland

Johan (reask)...

you changed default batchsize,
why did you do this ?
can you try with default value

/Jonas

[3 Nov 2010 12:36] Jonas Oreland

johan says that he can't repeat with latest version
(i never could)