[
https://issues.apache.org/jira/browse/DERBY-4773?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12898358#action_12898358
]
Soundararajan Velu commented on DERBY-4773:
-------------------------------------------
Thanks a lot Rick, this really helps.. I have checked the start-up scripts and
they don't carry this flag...
The other possibility that I suspect is when Hive is performing updates on its
metastore (Derby) and Derby goes down abruptly and is brought back online. at
this point of time (I guess in the clean up methods), derby spits out a lot of
ArrayIndexOutOfBounds exception at the initialization methods followed by NPE
at the exception blocks of the execute exception handlers. which causes the
leak of the generated classes.
I am unable to upload the stack trace (due to company policy). :(
I will reproduce this problem at home and upload the trace and exact scenario.
> Derby Network service mode creates a lot of generated classes (extends
> BaseActivation).
> ---------------------------------------------------------------------------------------
>
> Key: DERBY-4773
> URL: https://issues.apache.org/jira/browse/DERBY-4773
> Project: Derby
> Issue Type: Bug
> Components: JDBC
> Affects Versions: 10.6.1.0
> Environment: Suse Linux 11, Hive 5.0, Hadoop v.20.1
> Reporter: Soundararajan Velu
> Fix For: 10.6.1.0
>
>
> When I run Derby as a Network service, due to some reason it created about
> 1800 generated calss files (ac0a3c3082x012ax5c4fx64fcx00004650e8f2b.class,
> extends BaseActivation), I am unable to figure out the reason and now unable
> to reproduce the situation again. Has anyone faced this issue.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
