[
https://issues.apache.org/jira/browse/ARIES-485?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12930637#action_12930637
]
Valentin Mahrwald commented on ARIES-485:
-----------------------------------------
On closer investigation it does not seem like it is actually the
TransactionTimer thread that is the problem. I have tried to terminate various
suspicious threads at the end of the JpaBlogSampleWithEbaTest as follows:
for (Map.Entry<Thread,StackTraceElement[]> e :
Thread.getAllStackTraces().entrySet()) {
boolean found = false;
for (StackTraceElement st : e.getValue()) {
if (st.getClassName().contains("java.util.Timer")) found = true;
}
if (found) e.getKey().stop();
}
Stopping all java.util.Timer threads solves the hang whereas removing the
TransactionTimer thread does not solve it. So maybe there are new timers
introduced in geronimo-transaction 2.2 or there is some subtle interaction that
is different ???
> On IBM J9 JVM Blog itests hang
> ------------------------------
>
> Key: ARIES-485
> URL: https://issues.apache.org/jira/browse/ARIES-485
> Project: Aries
> Issue Type: Bug
> Components: Transaction
> Affects Versions: 0.3
> Environment: IBM J9
> Reporter: Valentin Mahrwald
> Fix For: 0.3
>
> Attachments: javacore.paxexam.txt.zip
>
>
> With Geronimo transaction manager 2.2, on the J9 JVM we see consistent hangs
> in the JpaBlogSampleWithEbaTest, which is just the first itest to use the
> real transaction manager. Moving back to 2.1.3 fixes the problem, so almost
> certainly it is in some form related to the new level.
> Concretely this has been seen when building the trunk with:
> mvn -fae -e -B clean install
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
