PH56881: NULLPOINTEREXCEPTION DURING RECOVERY OF WS-AT SUBORDINATE IN WSATRECOVERYCOORDINATOR.FROMLOGDATA CAUSES RECOVERY TO FAIL

APAR status

• Closed as program error.

Error description

• WebSphere Application Server v8.5.5.22 startup fails and the
  following errors are thrown in the SystemOut.log:
  
  WTRN0000E: An internal error occurred in method reconstruct in
  class com.ibm.ws.tx.jta.TransactionImpl; the exception stack
  trace follows: javax.transaction.SystemException
  at com.ibm.ws.Transaction.wstx.WSATRecoveryCoordinator.fromLogD
  ata(WSATRecoveryCoordinator.java:128)
  at com.ibm.ws.tx.jta.TransactionImpl.reconstructCoordinators(Tr
  ansactionImpl.java:1609)
  at com.ibm.tx.jta.impl.TransactionImpl.reconstruct(TransactionI
  mpl.java:745)
  at com.ibm.ws.tx.jta.RecoveryManager.handleTranRecord(RecoveryM
  anager.java:1015)
  at com.ibm.tx.jta.impl.RecoveryManager.replayTranLog(RecoveryMa
  nager.java:499)
  at com.ibm.tx.jta.impl.RecoveryManager.run(RecoveryManager.java
  :2160)
  at java.lang.Thread.run(Thread.java:830)
  Caused by: java.lang.NullPointerException
  at com.ibm.ws.Transaction.wstx.WSATRecoveryCoordinator.fromLogD
  ata(WSATRecoveryCoordinator.java:118)
  
  WTRN0025E: Recovery of Transaction failed
  javax.transaction.SystemException
  at com.ibm.ws.tx.jta.TransactionImpl.reconstructCoordinators(Tr
  ansactionImpl.java:1618)
  at com.ibm.tx.jta.impl.TransactionImpl.reconstruct(TransactionI
  mpl.java:745)
  at com.ibm.ws.tx.jta.RecoveryManager.handleTranRecord(RecoveryM
  anager.java:1015)
  at com.ibm.tx.jta.impl.RecoveryManager.replayTranLog(RecoveryMa
  nager.java:499)
  at com.ibm.tx.jta.impl.RecoveryManager.run(RecoveryManager.java
  :2160)
  at java.lang.Thread.run(Thread.java:830)
  
  WSVR0009E: Error occurred during startup
  com.ibm.ws.exception.RuntimeError:
  javax.transaction.SystemException:
  javax.transaction.SystemException
  at com.ibm.ws.tx.util.WASTMHelper.asynchRecoveryProcessingCompl
  ete(WASTMHelper.java:176)
  at com.ibm.tx.util.TMHelper.asynchRecoveryProcessingComplete(TM
  Helper.java:57)
  at com.ibm.tx.jta.impl.RecoveryManager.recoveryFailed(RecoveryM
  anager.java:1464)
  at com.ibm.tx.jta.impl.RecoveryManager.run(RecoveryManager.java
  :2172)
  at java.lang.Thread.run(Thread.java:830)
  Caused by: javax.transaction.SystemException:
  javax.transaction.SystemException
  at com.ibm.tx.jta.impl.RecoveryManager.replayTranLog(RecoveryMa
  nager.java:535)
  at com.ibm.tx.jta.impl.RecoveryManager.run(RecoveryManager.java
  :2160)
  

Local fix

Problem summary

• ****************************************************************
  * USERS AFFECTED:  WebSphere Application Server users          *
  *                  of Web Services-Atomic Transactions (WS-    *
  *                  AT)                                         *
  ****************************************************************
  * PROBLEM DESCRIPTION: NullPointerException occurs during      *
  *                      recovery in                             *
  *                      WSATRecoveryCoordinator.fromLogData     *
  ****************************************************************
  * RECOMMENDATION:                                              *
  ****************************************************************
  When using the WS-AT protocol, the WebSphere Application Server
  implementation does not provide the WS-AT Coordinator to the
  subordinate until prepare time.  A subordinate transaction may
  unilaterally rollback, for instance due to a client inactivity
  timeout, in which case the subordinate will not have a
  reference to the superior's Coordinator.
  When the subordinate transaction branch rolled back due to a
  timeout, it misidentified the 1PC rollback on timeout as being
  the second phase of a 2PC rollback when handling a subsequent
  Exception while distributing the rollback to its enlisted
  reosurces and mistakenly logged the subordinate branch.
  On a subsequent restart of the subordinate application
  server, the logged subordinate transaction branch was read in,
  but no Coordinator for the superior was found, resulting in the
  NullPointerException.
  

Problem conclusion

• The WS-AT code was modified so that it did not mis-identify the
  1PC rollback as a 2PC rollback when handling Exceptions from
  its own enlisted resources.
  
  The fix for this APAR is targeted for inclusion in fix pack
  8.5.5.25 and 9.0.5.18. For more information, see 'Recommended
  Updates for WebSphere Application Server':
  https://www.ibm.com/support/pages/node/715553
  

Temporary fix

Comments

APAR Information

• APAR is sysrouted FROM one or more of the following:

• APAR is sysrouted TO one or more of the following:

Fix information

• Fixed component name

  WEBS APP SERV N

• Fixed component ID

  5724H8800

Applicable component levels