You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@openjpa.apache.org by "Hiroki Tateno (JIRA)" <ji...@apache.org> on 2009/02/09 07:52:59 UTC
[jira] Created: (OPENJPA-913) A deadlock issue happens when
DirtyListener is used
A deadlock issue happens when DirtyListener is used
---------------------------------------------------
Key: OPENJPA-913
URL: https://issues.apache.org/jira/browse/OPENJPA-913
Project: OpenJPA
Issue Type: Bug
Components: kernel
Affects Versions: 1.2.0
Environment: Linux x86
Reporter: Hiroki Tateno
A deadlock issue happens when OpenJPA entity manager is
concurrently called and DirtyListener is used and
DirtyListener.beforeDirty calls an entity manager related
operation. To call OpenJPA entity manager concurrently,
we need to define openjpa.MulthThreaded option as true.
Following is test scenario.
1. Thread A calls entityMangaer.refresh() repeatedly.
In refresh() method, entityManager acquires BrokerImpl
lock. And then, entityManager acquires LifecycleEventManager
lock to call lifecycle callback.
2. Thread B calls persistedObject.getAItems() (getting
collection items).
3. In enhanced getItems() method, entityManager tries
to mark it as "dirty". Before marking, callback listener
DirtyListener.beforeDirty is called. In this point,
LifecycleEventManager lock is acquired without acquiring
BrokerImpl lock.
4. In the testcase, beforeDirty calls persistedObject.getAItems().
And then, entity manager tries to acquire BrokerImpl lock.
But, sometimes BrokerImpl lock is already acquired by Thread B.
So, a deadlock issue happens. deadlock stack is as follows.
====
[java] "Thread-1" prio=1 tid=0x09e98b28 nid=0x7fcc waiting on condition [0xb15f3000..0xb15f4130]
[java] at sun.misc.Unsafe.park(Native Method)
[java] at java.util.concurrent.locks.LockSupport.park(LockSupport.java:118)
[java] at java.util.concurrent.locks.AbstractQueuedSynchronizer.parkAndCheckInterrupt(AbstractQueuedSynchronizer.java:716)
[java] at java.util.concurrent.locks.AbstractQueuedSynchronizer.acquireQueued(AbstractQueuedSynchronizer.java:746)
[java] at java.util.concurrent.locks.AbstractQueuedSynchronizer.acquire(AbstractQueuedSynchronizer.java:1076)
[java] at java.util.concurrent.locks.ReentrantLock$NonfairSync.lock(ReentrantLock.java:184)
[java] at java.util.concurrent.locks.ReentrantLock.lock(ReentrantLock.java:256)
[java] at org.apache.openjpa.kernel.BrokerImpl.lock(BrokerImpl.java:4168)
[java] at org.apache.openjpa.kernel.BrokerImpl.beginOperation(BrokerImpl.java:1770)
[java] at org.apache.openjpa.kernel.BrokerImpl.isActive(BrokerImpl.java:1742)
[java] at org.apache.openjpa.kernel.StateManagerImpl.beforeRead(StateManagerImpl.java:964)
[java] at org.apache.openjpa.kernel.StateManagerImpl.accessingField(StateManagerImpl.java:1501)
[java] at model.A.getAItems(A.java)
[java] at model.ADirtyListener.beforeDirty(ADirtyListener.java:24)
[java] at org.apache.openjpa.event.LifecycleEventManager.fireEvent(LifecycleEventManager.java:423)
[java] at org.apache.openjpa.event.LifecycleEventManager.fireEvent(LifecycleEventManager.java:289)
[java] - locked <0x51b4e4b0> (a org.apache.openjpa.event.LifecycleEventManager)
[java] at org.apache.openjpa.kernel.BrokerImpl.fireLifecycleEvent(BrokerImpl.java:693)
[java] at org.apache.openjpa.kernel.StateManagerImpl.fireLifecycleEvent(StateManagerImpl.java:364)
[java] at org.apache.openjpa.kernel.StateManagerImpl.dirty(StateManagerImpl.java:1596)
[java] at org.apache.openjpa.kernel.StateManagerImpl.dirty(StateManagerImpl.java:1539)
[java] at org.apache.openjpa.util.Proxies.dirty(Proxies.java:66)
[java] at org.apache.openjpa.util.ProxyCollections.beforeAdd(ProxyCollections.java:57)
[java] at org.apache.openjpa.util.java$util$HashSet$proxy.add(Unknown Source)
[java] at business.Test$2.run(Test.java:80)
[java] at java.lang.Thread.run(Thread.java:595)
[java] "Thread-0" prio=1 tid=0x09e9d010 nid=0x7fcb waiting for monitor entry [0xb1674000..0xb1674db0]
[java] at org.apache.openjpa.event.LifecycleEventManager.fireEvent(LifecycleEventManager.java:272)
[java] - waiting to lock <0x51b4e4b0> (a org.apache.openjpa.event.LifecycleEventManager)
[java] at org.apache.openjpa.kernel.BrokerImpl.fireLifecycleEvent(BrokerImpl.java:693)
[java] at org.apache.openjpa.kernel.StateManagerImpl.fireLifecycleEvent(StateManagerImpl.java:364)
[java] at org.apache.openjpa.kernel.StateManagerImpl.clearFields(StateManagerImpl.java:2647)
[java] at org.apache.openjpa.kernel.StateManagerImpl.beforeRefresh(StateManagerImpl.java:1239)
[java] at org.apache.openjpa.kernel.BrokerImpl.refreshInternal(BrokerImpl.java:2835)
[java] at org.apache.openjpa.kernel.BrokerImpl.refresh(BrokerImpl.java:2781)
[java] at org.apache.openjpa.kernel.DelegatingBroker.refresh(DelegatingBroker.java:1078)
[java] at org.apache.openjpa.persistence.EntityManagerImpl.refresh(EntityManagerImpl.java:694)
[java] at business.Test$1.run(Test.java:64)
[java] at java.lang.Thread.run(Thread.java:595)
====
Intially the problem is reproduced on OpenJPA 1.x. But,
I verified the problem could be reproduced with latest
OpenJPA head.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (OPENJPA-913) A deadlock issue happens when
DirtyListener is used
Posted by "Hiroki Tateno (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/OPENJPA-913?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Hiroki Tateno updated OPENJPA-913:
----------------------------------
Attachment: proposed-fix.patch
Also, I attached possible fix of the issue.
The root cause of the issue is sometimes BrokerImpl.fireLifecycleEvent
is called without BrokerImpl lock. So, I added lock()/unlock() to
BrokerImpl.fireLifecycleEvent. It may be overkill fix. If it's overkill,
it's better that adding broker.lock()/broker.unlock() to around
calling fireLifecycleEvent() in StateManagerImpl.dirty().
> A deadlock issue happens when DirtyListener is used
> ---------------------------------------------------
>
> Key: OPENJPA-913
> URL: https://issues.apache.org/jira/browse/OPENJPA-913
> Project: OpenJPA
> Issue Type: Bug
> Components: kernel
> Affects Versions: 1.2.0
> Environment: Linux x86
> Reporter: Hiroki Tateno
> Attachments: proposed-fix.patch, testcase.zip
>
>
> A deadlock issue happens when OpenJPA entity manager is
> concurrently called and DirtyListener is used and
> DirtyListener.beforeDirty calls an entity manager related
> operation. To call OpenJPA entity manager concurrently,
> we need to define openjpa.MulthThreaded option as true.
> Following is test scenario.
> 1. Thread A calls entityMangaer.refresh() repeatedly.
> In refresh() method, entityManager acquires BrokerImpl
> lock. And then, entityManager acquires LifecycleEventManager
> lock to call lifecycle callback.
> 2. Thread B calls persistedObject.getAItems() (getting
> collection items).
> 3. In enhanced getItems() method, entityManager tries
> to mark it as "dirty". Before marking, callback listener
> DirtyListener.beforeDirty is called. In this point,
> LifecycleEventManager lock is acquired without acquiring
> BrokerImpl lock.
> 4. In the testcase, beforeDirty calls persistedObject.getAItems().
> And then, entity manager tries to acquire BrokerImpl lock.
> But, sometimes BrokerImpl lock is already acquired by Thread B.
> So, a deadlock issue happens. deadlock stack is as follows.
> ====
> [java] "Thread-1" prio=1 tid=0x09e98b28 nid=0x7fcc waiting on condition [0xb15f3000..0xb15f4130]
> [java] at sun.misc.Unsafe.park(Native Method)
> [java] at java.util.concurrent.locks.LockSupport.park(LockSupport.java:118)
> [java] at java.util.concurrent.locks.AbstractQueuedSynchronizer.parkAndCheckInterrupt(AbstractQueuedSynchronizer.java:716)
> [java] at java.util.concurrent.locks.AbstractQueuedSynchronizer.acquireQueued(AbstractQueuedSynchronizer.java:746)
> [java] at java.util.concurrent.locks.AbstractQueuedSynchronizer.acquire(AbstractQueuedSynchronizer.java:1076)
> [java] at java.util.concurrent.locks.ReentrantLock$NonfairSync.lock(ReentrantLock.java:184)
> [java] at java.util.concurrent.locks.ReentrantLock.lock(ReentrantLock.java:256)
> [java] at org.apache.openjpa.kernel.BrokerImpl.lock(BrokerImpl.java:4168)
> [java] at org.apache.openjpa.kernel.BrokerImpl.beginOperation(BrokerImpl.java:1770)
> [java] at org.apache.openjpa.kernel.BrokerImpl.isActive(BrokerImpl.java:1742)
> [java] at org.apache.openjpa.kernel.StateManagerImpl.beforeRead(StateManagerImpl.java:964)
> [java] at org.apache.openjpa.kernel.StateManagerImpl.accessingField(StateManagerImpl.java:1501)
> [java] at model.A.getAItems(A.java)
> [java] at model.ADirtyListener.beforeDirty(ADirtyListener.java:24)
> [java] at org.apache.openjpa.event.LifecycleEventManager.fireEvent(LifecycleEventManager.java:423)
> [java] at org.apache.openjpa.event.LifecycleEventManager.fireEvent(LifecycleEventManager.java:289)
> [java] - locked <0x51b4e4b0> (a org.apache.openjpa.event.LifecycleEventManager)
> [java] at org.apache.openjpa.kernel.BrokerImpl.fireLifecycleEvent(BrokerImpl.java:693)
> [java] at org.apache.openjpa.kernel.StateManagerImpl.fireLifecycleEvent(StateManagerImpl.java:364)
> [java] at org.apache.openjpa.kernel.StateManagerImpl.dirty(StateManagerImpl.java:1596)
> [java] at org.apache.openjpa.kernel.StateManagerImpl.dirty(StateManagerImpl.java:1539)
> [java] at org.apache.openjpa.util.Proxies.dirty(Proxies.java:66)
> [java] at org.apache.openjpa.util.ProxyCollections.beforeAdd(ProxyCollections.java:57)
> [java] at org.apache.openjpa.util.java$util$HashSet$proxy.add(Unknown Source)
> [java] at business.Test$2.run(Test.java:80)
> [java] at java.lang.Thread.run(Thread.java:595)
> [java] "Thread-0" prio=1 tid=0x09e9d010 nid=0x7fcb waiting for monitor entry [0xb1674000..0xb1674db0]
> [java] at org.apache.openjpa.event.LifecycleEventManager.fireEvent(LifecycleEventManager.java:272)
> [java] - waiting to lock <0x51b4e4b0> (a org.apache.openjpa.event.LifecycleEventManager)
> [java] at org.apache.openjpa.kernel.BrokerImpl.fireLifecycleEvent(BrokerImpl.java:693)
> [java] at org.apache.openjpa.kernel.StateManagerImpl.fireLifecycleEvent(StateManagerImpl.java:364)
> [java] at org.apache.openjpa.kernel.StateManagerImpl.clearFields(StateManagerImpl.java:2647)
> [java] at org.apache.openjpa.kernel.StateManagerImpl.beforeRefresh(StateManagerImpl.java:1239)
> [java] at org.apache.openjpa.kernel.BrokerImpl.refreshInternal(BrokerImpl.java:2835)
> [java] at org.apache.openjpa.kernel.BrokerImpl.refresh(BrokerImpl.java:2781)
> [java] at org.apache.openjpa.kernel.DelegatingBroker.refresh(DelegatingBroker.java:1078)
> [java] at org.apache.openjpa.persistence.EntityManagerImpl.refresh(EntityManagerImpl.java:694)
> [java] at business.Test$1.run(Test.java:64)
> [java] at java.lang.Thread.run(Thread.java:595)
> ====
> Intially the problem is reproduced on OpenJPA 1.x. But,
> I verified the problem could be reproduced with latest
> OpenJPA head.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (OPENJPA-913) A deadlock issue happens when
DirtyListener is used
Posted by "Hiroki Tateno (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/OPENJPA-913?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Hiroki Tateno updated OPENJPA-913:
----------------------------------
Attachment: testcase.zip
I uploaded a simple testcase testcase.zip.
====
step by step test procedure)
1. update database configuration in persistence.xml
2. run "ant test"
====
> A deadlock issue happens when DirtyListener is used
> ---------------------------------------------------
>
> Key: OPENJPA-913
> URL: https://issues.apache.org/jira/browse/OPENJPA-913
> Project: OpenJPA
> Issue Type: Bug
> Components: kernel
> Affects Versions: 1.2.0
> Environment: Linux x86
> Reporter: Hiroki Tateno
> Attachments: testcase.zip
>
>
> A deadlock issue happens when OpenJPA entity manager is
> concurrently called and DirtyListener is used and
> DirtyListener.beforeDirty calls an entity manager related
> operation. To call OpenJPA entity manager concurrently,
> we need to define openjpa.MulthThreaded option as true.
> Following is test scenario.
> 1. Thread A calls entityMangaer.refresh() repeatedly.
> In refresh() method, entityManager acquires BrokerImpl
> lock. And then, entityManager acquires LifecycleEventManager
> lock to call lifecycle callback.
> 2. Thread B calls persistedObject.getAItems() (getting
> collection items).
> 3. In enhanced getItems() method, entityManager tries
> to mark it as "dirty". Before marking, callback listener
> DirtyListener.beforeDirty is called. In this point,
> LifecycleEventManager lock is acquired without acquiring
> BrokerImpl lock.
> 4. In the testcase, beforeDirty calls persistedObject.getAItems().
> And then, entity manager tries to acquire BrokerImpl lock.
> But, sometimes BrokerImpl lock is already acquired by Thread B.
> So, a deadlock issue happens. deadlock stack is as follows.
> ====
> [java] "Thread-1" prio=1 tid=0x09e98b28 nid=0x7fcc waiting on condition [0xb15f3000..0xb15f4130]
> [java] at sun.misc.Unsafe.park(Native Method)
> [java] at java.util.concurrent.locks.LockSupport.park(LockSupport.java:118)
> [java] at java.util.concurrent.locks.AbstractQueuedSynchronizer.parkAndCheckInterrupt(AbstractQueuedSynchronizer.java:716)
> [java] at java.util.concurrent.locks.AbstractQueuedSynchronizer.acquireQueued(AbstractQueuedSynchronizer.java:746)
> [java] at java.util.concurrent.locks.AbstractQueuedSynchronizer.acquire(AbstractQueuedSynchronizer.java:1076)
> [java] at java.util.concurrent.locks.ReentrantLock$NonfairSync.lock(ReentrantLock.java:184)
> [java] at java.util.concurrent.locks.ReentrantLock.lock(ReentrantLock.java:256)
> [java] at org.apache.openjpa.kernel.BrokerImpl.lock(BrokerImpl.java:4168)
> [java] at org.apache.openjpa.kernel.BrokerImpl.beginOperation(BrokerImpl.java:1770)
> [java] at org.apache.openjpa.kernel.BrokerImpl.isActive(BrokerImpl.java:1742)
> [java] at org.apache.openjpa.kernel.StateManagerImpl.beforeRead(StateManagerImpl.java:964)
> [java] at org.apache.openjpa.kernel.StateManagerImpl.accessingField(StateManagerImpl.java:1501)
> [java] at model.A.getAItems(A.java)
> [java] at model.ADirtyListener.beforeDirty(ADirtyListener.java:24)
> [java] at org.apache.openjpa.event.LifecycleEventManager.fireEvent(LifecycleEventManager.java:423)
> [java] at org.apache.openjpa.event.LifecycleEventManager.fireEvent(LifecycleEventManager.java:289)
> [java] - locked <0x51b4e4b0> (a org.apache.openjpa.event.LifecycleEventManager)
> [java] at org.apache.openjpa.kernel.BrokerImpl.fireLifecycleEvent(BrokerImpl.java:693)
> [java] at org.apache.openjpa.kernel.StateManagerImpl.fireLifecycleEvent(StateManagerImpl.java:364)
> [java] at org.apache.openjpa.kernel.StateManagerImpl.dirty(StateManagerImpl.java:1596)
> [java] at org.apache.openjpa.kernel.StateManagerImpl.dirty(StateManagerImpl.java:1539)
> [java] at org.apache.openjpa.util.Proxies.dirty(Proxies.java:66)
> [java] at org.apache.openjpa.util.ProxyCollections.beforeAdd(ProxyCollections.java:57)
> [java] at org.apache.openjpa.util.java$util$HashSet$proxy.add(Unknown Source)
> [java] at business.Test$2.run(Test.java:80)
> [java] at java.lang.Thread.run(Thread.java:595)
> [java] "Thread-0" prio=1 tid=0x09e9d010 nid=0x7fcb waiting for monitor entry [0xb1674000..0xb1674db0]
> [java] at org.apache.openjpa.event.LifecycleEventManager.fireEvent(LifecycleEventManager.java:272)
> [java] - waiting to lock <0x51b4e4b0> (a org.apache.openjpa.event.LifecycleEventManager)
> [java] at org.apache.openjpa.kernel.BrokerImpl.fireLifecycleEvent(BrokerImpl.java:693)
> [java] at org.apache.openjpa.kernel.StateManagerImpl.fireLifecycleEvent(StateManagerImpl.java:364)
> [java] at org.apache.openjpa.kernel.StateManagerImpl.clearFields(StateManagerImpl.java:2647)
> [java] at org.apache.openjpa.kernel.StateManagerImpl.beforeRefresh(StateManagerImpl.java:1239)
> [java] at org.apache.openjpa.kernel.BrokerImpl.refreshInternal(BrokerImpl.java:2835)
> [java] at org.apache.openjpa.kernel.BrokerImpl.refresh(BrokerImpl.java:2781)
> [java] at org.apache.openjpa.kernel.DelegatingBroker.refresh(DelegatingBroker.java:1078)
> [java] at org.apache.openjpa.persistence.EntityManagerImpl.refresh(EntityManagerImpl.java:694)
> [java] at business.Test$1.run(Test.java:64)
> [java] at java.lang.Thread.run(Thread.java:595)
> ====
> Intially the problem is reproduced on OpenJPA 1.x. But,
> I verified the problem could be reproduced with latest
> OpenJPA head.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Resolved: (OPENJPA-913) A deadlock issue happens when
DirtyListener is used
Posted by "Hiroki Tateno (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/OPENJPA-913?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Hiroki Tateno resolved OPENJPA-913.
-----------------------------------
Resolution: Fixed
Fix Version/s: 2.0.0
1.1.1
Abe's change resolved the issue.
> A deadlock issue happens when DirtyListener is used
> ---------------------------------------------------
>
> Key: OPENJPA-913
> URL: https://issues.apache.org/jira/browse/OPENJPA-913
> Project: OpenJPA
> Issue Type: Bug
> Components: kernel
> Affects Versions: 1.2.0
> Environment: Linux x86
> Reporter: Hiroki Tateno
> Fix For: 1.1.1, 2.0.0
>
> Attachments: proposed-fix.patch, testcase.zip
>
>
> A deadlock issue happens when OpenJPA entity manager is
> concurrently called and DirtyListener is used and
> DirtyListener.beforeDirty calls an entity manager related
> operation. To call OpenJPA entity manager concurrently,
> we need to define openjpa.MulthThreaded option as true.
> Following is test scenario.
> 1. Thread A calls entityMangaer.refresh() repeatedly.
> In refresh() method, entityManager acquires BrokerImpl
> lock. And then, entityManager acquires LifecycleEventManager
> lock to call lifecycle callback.
> 2. Thread B calls persistedObject.getAItems() (getting
> collection items).
> 3. In enhanced getItems() method, entityManager tries
> to mark it as "dirty". Before marking, callback listener
> DirtyListener.beforeDirty is called. In this point,
> LifecycleEventManager lock is acquired without acquiring
> BrokerImpl lock.
> 4. In the testcase, beforeDirty calls persistedObject.getAItems().
> And then, entity manager tries to acquire BrokerImpl lock.
> But, sometimes BrokerImpl lock is already acquired by Thread B.
> So, a deadlock issue happens. deadlock stack is as follows.
> ====
> [java] "Thread-1" prio=1 tid=0x09e98b28 nid=0x7fcc waiting on condition [0xb15f3000..0xb15f4130]
> [java] at sun.misc.Unsafe.park(Native Method)
> [java] at java.util.concurrent.locks.LockSupport.park(LockSupport.java:118)
> [java] at java.util.concurrent.locks.AbstractQueuedSynchronizer.parkAndCheckInterrupt(AbstractQueuedSynchronizer.java:716)
> [java] at java.util.concurrent.locks.AbstractQueuedSynchronizer.acquireQueued(AbstractQueuedSynchronizer.java:746)
> [java] at java.util.concurrent.locks.AbstractQueuedSynchronizer.acquire(AbstractQueuedSynchronizer.java:1076)
> [java] at java.util.concurrent.locks.ReentrantLock$NonfairSync.lock(ReentrantLock.java:184)
> [java] at java.util.concurrent.locks.ReentrantLock.lock(ReentrantLock.java:256)
> [java] at org.apache.openjpa.kernel.BrokerImpl.lock(BrokerImpl.java:4168)
> [java] at org.apache.openjpa.kernel.BrokerImpl.beginOperation(BrokerImpl.java:1770)
> [java] at org.apache.openjpa.kernel.BrokerImpl.isActive(BrokerImpl.java:1742)
> [java] at org.apache.openjpa.kernel.StateManagerImpl.beforeRead(StateManagerImpl.java:964)
> [java] at org.apache.openjpa.kernel.StateManagerImpl.accessingField(StateManagerImpl.java:1501)
> [java] at model.A.getAItems(A.java)
> [java] at model.ADirtyListener.beforeDirty(ADirtyListener.java:24)
> [java] at org.apache.openjpa.event.LifecycleEventManager.fireEvent(LifecycleEventManager.java:423)
> [java] at org.apache.openjpa.event.LifecycleEventManager.fireEvent(LifecycleEventManager.java:289)
> [java] - locked <0x51b4e4b0> (a org.apache.openjpa.event.LifecycleEventManager)
> [java] at org.apache.openjpa.kernel.BrokerImpl.fireLifecycleEvent(BrokerImpl.java:693)
> [java] at org.apache.openjpa.kernel.StateManagerImpl.fireLifecycleEvent(StateManagerImpl.java:364)
> [java] at org.apache.openjpa.kernel.StateManagerImpl.dirty(StateManagerImpl.java:1596)
> [java] at org.apache.openjpa.kernel.StateManagerImpl.dirty(StateManagerImpl.java:1539)
> [java] at org.apache.openjpa.util.Proxies.dirty(Proxies.java:66)
> [java] at org.apache.openjpa.util.ProxyCollections.beforeAdd(ProxyCollections.java:57)
> [java] at org.apache.openjpa.util.java$util$HashSet$proxy.add(Unknown Source)
> [java] at business.Test$2.run(Test.java:80)
> [java] at java.lang.Thread.run(Thread.java:595)
> [java] "Thread-0" prio=1 tid=0x09e9d010 nid=0x7fcb waiting for monitor entry [0xb1674000..0xb1674db0]
> [java] at org.apache.openjpa.event.LifecycleEventManager.fireEvent(LifecycleEventManager.java:272)
> [java] - waiting to lock <0x51b4e4b0> (a org.apache.openjpa.event.LifecycleEventManager)
> [java] at org.apache.openjpa.kernel.BrokerImpl.fireLifecycleEvent(BrokerImpl.java:693)
> [java] at org.apache.openjpa.kernel.StateManagerImpl.fireLifecycleEvent(StateManagerImpl.java:364)
> [java] at org.apache.openjpa.kernel.StateManagerImpl.clearFields(StateManagerImpl.java:2647)
> [java] at org.apache.openjpa.kernel.StateManagerImpl.beforeRefresh(StateManagerImpl.java:1239)
> [java] at org.apache.openjpa.kernel.BrokerImpl.refreshInternal(BrokerImpl.java:2835)
> [java] at org.apache.openjpa.kernel.BrokerImpl.refresh(BrokerImpl.java:2781)
> [java] at org.apache.openjpa.kernel.DelegatingBroker.refresh(DelegatingBroker.java:1078)
> [java] at org.apache.openjpa.persistence.EntityManagerImpl.refresh(EntityManagerImpl.java:694)
> [java] at business.Test$1.run(Test.java:64)
> [java] at java.lang.Thread.run(Thread.java:595)
> ====
> Intially the problem is reproduced on OpenJPA 1.x. But,
> I verified the problem could be reproduced with latest
> OpenJPA head.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.