[
https://issues.apache.org/jira/browse/DERBY-6506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13932147#comment-13932147
]
Mike Matrigali commented on DERBY-6506:
---------------------------------------
I agree with mynra, in most cases deadlocks involving user tables are not Derby
bugs. In order to help you will have to post more information about your
issue. A good start would be the complete deadlock error from derby.log. That
will likely lead to more questions about the ddl of the tables involved and the
application logic
of the other queries that might be in the transaction and isolation level of
the transaction.
See the following for how to debug locking problems:
http://wiki.apache.org/db-derby/LockDebugging
Applications should be coded to handle deadlocks (and in most cases just retry
the transaction). In
some cases changes to application ddl can avoid the deadlock, for instance
adding a useful index for the queries that are hitting the deadlock.
> Deadlock issues with Derby when using multithreading in Netcool Impact
> ----------------------------------------------------------------------
>
> Key: DERBY-6506
> URL: https://issues.apache.org/jira/browse/DERBY-6506
> Project: Derby
> Issue Type: Bug
> Components: SQL
> Affects Versions: 10.8.2.2
> Reporter: Shefali Sampat
> Fix For: 10.8.2.2
>
>
> Customer is using Derby embedded in Netcool Impact 6.1.1. Customer has
> created custom tables in the DB. When using multithreading and sending a lot
> of events, customer is noticing deadlock issues. The deadlock clears up after
> some time. Customer enabled row locking instead of table locking but still
> the same issue occurs. Please contact Shefali Sampat (Impact L3) at
> [email protected] for further info or if u need to look at the derby logs.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
