aries-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Christian Schneider (JIRA)" <>
Subject [jira] [Commented] (ARIES-1454) NPE in TxInterceptor
Date Thu, 19 Nov 2015 15:48:11 GMT


Christian Schneider commented on ARIES-1454:

I had a lot of problems with the GC when I started using the Coordinator. I just called begin
to put the coordination on the thread based stack. In this case the GC indeed cleans it up.
Now it is different though. The Coordination should be stored on the PrecallToken so the GC
should not happen. 

So I am not sure what is going on. Can you test with the code on trunk or do you need a backport?

> NPE in TxInterceptor
> --------------------
>                 Key: ARIES-1454
>                 URL:
>             Project: Aries
>          Issue Type: Bug
>          Components: Transaction
>    Affects Versions: transaction-blueprint-1.1.1
>         Environment: Karaf 4.0.2 (Felix)
> JPA (Hibernate)
>            Reporter: Mohammad Shamsi
>            Assignee: Christian Schneider
>             Fix For: transaction-3.0.0
> I am running some load tests agains an osgi app, which is running inside karaf 4.0.2,
using Aries Blueprint and JPA (Hibernate) .
> The test is using multiple threads (10) to issue  some random calls to the application.

> Sometimes, randomly, some of the call throwing NPE in TxInterceptor: 
> {code:java}
> java.lang.NullPointerException
>         at org.apache.aries.transaction.TxInterceptorImpl.postCallWithReturn([66:org.apache.aries.transaction.blueprint:1.1.1]
>         at org.apache.aries.blueprint.proxy.SingleInterceptorCollaborator.postInvoke([48:org.apache.aries.blueprint.core:1.4.4]
>         at Proxy65968de8_a064_4203_9a53_5f0c8e536279.lockAll(Unknown Source)[:]
> {code}
> The NPE is thrown from "postCallWithReturn" and "postCallWithException" in following
> {code:java}
>  token.getCoordination().end();
> {code}
> The "getCoordination()"  call in above line is returning null. 
> As I mentioned, it is happening from time to time. sometimes after 20 mins, sometimes
after hours of test execution and it is happening in different transactional methods in different
classes of the application. 

This message was sent by Atlassian JIRA

View raw message