My first discovered-deadlock :)

About one week ago I was struggling at work with a really strange scalability issue on the back-end system I'm responsible of.
Once a system looks like frozen and does not react it could be possible it has been somehow somewhere deadlocked.
Thus I asked the JVM about the presence of any deadlocked threads, and the answer has been:

Found one Java-level deadlock:
=============================
"Thread-11":
waiting to lock monitor 0x08078ac4 (object 0xa5c85f98, a oracle.jdbc.driver.LogicalConnection),
which is held by "RMICallHandler-29"
"RMICallHandler-29":
waiting to lock monitor 0x08078e04 (object 0x67154b48, a oracle.jdbc.pool.OracleImplicitConnectionCache),
which is held by "Thread-11"

Java stack information for the threads listed above:
===================================================
"Thread-11":
at oracle.jdbc.driver.LogicalConnection.close(LogicalConnection.java:194)
- waiting to lock <0xa5c85f98> (a oracle.jdbc.driver.LogicalConnection)
at oracle.jdbc.pool.OracleImplicitConnectionCache.closeCheckedOutConnection(OracleImplicitConnectionCache.java:1148)
at oracle.jdbc.pool.OracleImplicitConnectionCacheThread.runTimeToLiveTimeout(OracleImplicitConnectionCacheThread.java:153)
- locked <0x67154b48> (a oracle.jdbc.pool.OracleImplicitConnectionCache)
at oracle.jdbc.pool.OracleImplicitConnectionCacheThread.run(OracleImplicitConnectionCacheThread.java:67)
"RMICallHandler-29":
at oracle.jdbc.pool.OracleImplicitConnectionCache.reusePooledConnection(OracleImplicitConnectionCache.java:1422)
- waiting to lock <0x67154b48> (a oracle.jdbc.pool.OracleImplicitConnectionCache)
at oracle.jdbc.pool.OracleConnectionCacheEventListener.connectionClosed(OracleConnectionCacheEventListener.java:61)
- locked <0x673947e8> (a oracle.jdbc.pool.OracleConnectionCacheEventListener)
at oracle.jdbc.pool.OraclePooledConnection.callImplicitCacheListener(OraclePooledConnection.java:519)
at oracle.jdbc.pool.OraclePooledConnection.logicalCloseForImplicitConnectionCache(OraclePooledConnection.java:440)
at oracle.jdbc.pool.OraclePooledConnection.logicalClose(OraclePooledConnection.java:456)
at oracle.jdbc.driver.LogicalConnection.closeInternal(LogicalConnection.java:221)
at oracle.jdbc.driver.LogicalConnection.close(LogicalConnection.java:194)
- locked <0xa5c85f98> (a oracle.jdbc.driver.LogicalConnection)
at oracle.oc4j.sql.spi.ManagedConnectionImpl._cleanup(ManagedConnectionImpl.java:371)
at oracle.oc4j.sql.spi.ManagedConnectionImpl.cleanup(ManagedConnectionImpl.java:339)
at com.evermind.server.connector.ConnectionContext.cleanup(ConnectionContext.java:893)
- locked <0xa5d4ee08> (a com.evermind.server.connector.ConnectionContext)
at com.evermind.server.connector.ConnectionContext.cleanupIfUnreferenced(ConnectionContext.java:836)
- locked <0xa5d4ee08> (a com.evermind.server.connector.ConnectionContext)
at com.evermind.server.connector.ConnectionContext.jtaTransactionEnded(ConnectionContext.java:733)
- locked <0xa5d4ee08> (a com.evermind.server.connector.ConnectionContext)
at com.evermind.server.connector.ConnectionContext.afterCompletion(ConnectionContext.java:542)
at com.evermind.server.ApplicationServerTransaction.callSynchronizationAfterCompletion(ApplicationServerTransaction.java:1184)
at com.evermind.server.ApplicationServerTransactionManager.freeResources(ApplicationServerTransactionManager.java:394)
at com.evermind.server.ApplicationServerTransaction.doCommit(ApplicationServerTransaction.java:282)
- locked <0xa5d5bab8> (a com.evermind.server.ApplicationServerTransaction)
at com.evermind.server.ApplicationServerTransaction.commit(ApplicationServerTransaction.java:162)
at com.evermind.server.ApplicationServerTransactionManager.commit(ApplicationServerTransactionManager.java:472)
at com.evermind.server.ejb.EJBTransactionManager.end(EJBTransactionManager.java:132)
at com.evermind.server.ejb.interceptor.system.TxRequiredInterceptor.invoke(TxRequiredInterceptor.java:57)
at com.evermind.server.ejb.interceptor.InvocationContextImpl.proceed(InvocationContextImpl.java:119)
at com.evermind.server.ejb.interceptor.system.DMSInterceptor.invoke(DMSInterceptor.java:52)
at com.evermind.server.ejb.interceptor.InvocationContextImpl.proceed(InvocationContextImpl.java:119)
at com.evermind.server.ejb.InvocationContextPool.invoke(InvocationContextPool.java:55)
at com.evermind.server.ejb.StatelessSessionEJBObject.OC4J_invokeMethod(StatelessSessionEJBObject.java:87)
at DgComponentSession_RemoteProxy_k3k2ci.create(Unknown Source)
at sun.reflect.GeneratedMethodAccessor126.invoke(Unknown Source)
at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:25)
at java.lang.reflect.Method.invoke(Method.java:585)
at com.evermind.server.rmi.RmiMethodCall.run(RmiMethodCall.java:53)
at com.evermind.util.ReleasableResourcePooledExecutor$MyWorker.run(ReleasableResourcePooledExecutor.java:303)
at java.lang.Thread.run(Thread.java:595)

Found 1 deadlock.

Once I saw this output message I could not belive to my own two eyes, I had finally found a deadlock in a third-party piece of code! I still belive it is a quite rare thing to see so clearly, so I felt quite happy and excited when I saw the mesage above for the first time.

Just a silly post to share this rare little-thing with others.

F.

JTR 4.0 final

Today I'm going to write this post in english. It's about JTR again. Finally I've been able to publish the final 4.0 release. Achieving this release has been a pretty cumbersome effort actually, but it has been funny and instructive in many ways.

I've published a JTR 4.0 final availability article on The Server Side this week, and it appeared on the main page yesterday. The full article can be reached here.

Hope JTR 4.0 will receive enough attention from the comunity ;). We'll see...

F.

Today I'm going to write this post in english. It's about JTR again. Finally I've been able to publish the final 4.0 release. Achieving this release has been a pretty cumbersome effort actually, but it has been funny and instructive in many ways.

I've published a JTR 4.0 final availability article on The Server Side this week, and it appeared on the main page yesterday. The full article can be reached here.