11:30 a.m.
[
https://issues.jboss.org/browse/WFLY-1313?page=com.atlassian.jira.plugin....
]
jaikiran pai closed WFLY-1313.
------------------------------
Resolution: Rejected
This can't really be handled with a proper graceful shutdown in place. Closing this
issue.
Here's a related conversation:
{quote}
[7:53:32 PM] Jaikiran Pai: i needed some input from stuart, but it's too late for him
now. jason, can you take a look at this commit that I added for
https://issues.jboss.org/browse/WFLY-1313
[7:53:42 PM] Jaikiran Pai:
https://github.com/jaikiran/jboss-eap/commit/b73c7bb197a08fa8962cdb4e9a52...
[7:53:46 PM] Jaikiran Pai: does it look fine to you?
[7:57:32 PM] Stuart Douglas: I am here
[7:57:38 PM] Jaikiran Pai: nice! :)
[7:57:58 PM] Jaikiran Pai: it looks like the components are misisng a dep on that CCM
service
[7:58:14 PM] Stuart Douglas: nah
[7:58:27 PM] Jaikiran Pai: something else?
[7:58:30 PM] Stuart Douglas: we have the same problem in the web with the CCM
[7:58:33 PM] Stuart Douglas: its not a dep issue
[7:58:56 PM] Stuart Douglas: but threads that are in flight when undelpoy starts cause a
problem once the CCM is shut down
[7:58:58 PM] Jaikiran Pai: from what i see, the CCM service is stopped even when the EJB
component is still alive
[7:59:12 PM] Jaikiran Pai: right
[7:59:45 PM] Jaikiran Pai: shouldn't we really stop the CCM (service) from being shut
down until the dependent component is undeployed fully?
[8:00:03 PM] Jaikiran Pai: a.k.a add a dep?
[8:00:07 PM] Stuart Douglas: This comes down to graceful shutdown
[8:01:06 PM] Stuart Douglas: Its not a dep problem, as unless we add a gate to the
components that dont all them to shutdown untill all threads are done processing you
can't stpo this
[8:01:40 PM] Stuart Douglas: If a remote invocation is in progress on an EJB the EJB can
still shutdown
[8:02:00 PM] Stuart Douglas: And for the most part the invocation just progresses as
normal
[8:02:21 PM] Stuart Douglas: but onces the CCM is in there you get that exception which
makes the non-graceful shutdown much more visible
[8:02:53 PM] Jaikiran Pai: yeah, that's actually what started showing up once i added
that test which does exactly this
[8:03:18 PM] Jaikiran Pai: i.e. trigger an undeploy which takes time to complete and at
the same time keep firing requests to that deployment remotely
[8:03:25 PM] Stuart Douglas: yea
{quote}
IllegalStateException in
CachedConnectionManagerSetupProcessor$CachedConnectionManagerSetupAction.teardown()
------------------------------------------------------------------------------------------------------------
Key: WFLY-1313
URL: https://issues.jboss.org/browse/WFLY-1313
Project: WildFly
Issue Type: Bug
Components: EE, JCA
Reporter: jaikiran pai
Assignee: jaikiran pai
The following exception was reported on lightning:
{code}
Caused by: java.lang.IllegalStateException
at org.jboss.msc.value.InjectedValue.getValue(InjectedValue.java:47)
at
org.jboss.as.connector.deployers.ra.processors.CachedConnectionManagerSetupProcessor$CachedConnectionManagerSetupAction.teardown(CachedConnectionManagerSetupProcessor.java:105)
at
org.jboss.as.ejb3.component.interceptors.AdditionalSetupInterceptor.processInvocation(AdditionalSetupInterceptor.java:62)
at org.jboss.invocation.InterceptorContext.proceed(InterceptorContext.java:288)
at org.jboss.as.ee.component.TCCLInterceptor.processInvocation(TCCLInterceptor.java:45)
at org.jboss.invocation.InterceptorContext.proceed(InterceptorContext.java:288)
at
org.jboss.invocation.ChainedInterceptor.processInvocation(ChainedInterceptor.java:61)
at org.jboss.as.ee.component.ViewService$View.invoke(ViewService.java:165)
at
org.jboss.as.ejb3.remote.protocol.versionone.MethodInvocationMessageHandler.invokeMethod(MethodInvocationMessageHandler.java:329)
at
org.jboss.as.ejb3.remote.protocol.versionone.MethodInvocationMessageHandler.access$100(MethodInvocationMessageHandler.java:70)
at
org.jboss.as.ejb3.remote.protocol.versionone.MethodInvocationMessageHandler$1.run(MethodInvocationMessageHandler.java:203)
at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:441)
at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:303)
at java.util.concurrent.FutureTask.run(FutureTask.java:138)
at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908)
at java.lang.Thread.run(Thread.java:662)
at org.jboss.threads.JBossThread.run(JBossThread.java:122)
{code}
The tests involved a EJB deployment and the EJBs were being invoked when the undeployment
was in progress (@see SlowUndeploymentRemoteFailoverTestCase).
It looks like the CachedConnectionManager service isn't being added as a dependency
on the components (like EJBs) which rely on it.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira