PI80769: JVMSERVER STALLS IN CICS AFTER JVM DUMP

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• A JVMServer in CICS doesn't work anymore after suffering
  a JVM dump. The CICS region must be recycled in order to
  resolve the situation.
  
  The CICS Joblog shows the following sequence of messages:
  ---------------------------------------------------------
  
  08.50.55 BPXM023I JVMDUMP039I Processing dump event "abort",
           detail "" at 2017/03/21 08:50:55 - please wait.
  
  09.25.58 +CEE3798I ATTEMPTING TO TAKE A DUMP FOR ABEND U4094
  
  09.30.59 BPXP018I THREAD xxxxxx, IN PROCESS yyyyyy, ENDED
           WITHOUT BEING UNDUBBED WITH COMPLETION CODE 84000FFE
  
  09.31.34 +DFHSJ0005 aaaaaaaa AN UNEXPECTED ABEND HAS OCCURRED
           ON THE INITIAL PTHREAD TCB FOR JVMSERVER: zzzzzz
  
  
  JVMERR shows the following failure  :
  -------------------------------------
  017/03/21 08:51:30.044099 CET D ¬LE-Thread-0? ¬DFHSJSC:2661
                            @Event: ipt_control_loop()
  A Posix signal was received by the JVM server. SIGQUIT may be
  used to obtain a Java dump. Signals causing termination
  (SIGKILL, SIGTERM, SIGABRT etc) should NEVER be issued. CICS
  may become unstable and the JVM server will be terminated in
  an uncontrolled
  

Local fix

• not available
  

Problem summary

• ****************************************************************
  * USERS AFFECTED: All CICS users                               *
  ****************************************************************
  * PROBLEM DESCRIPTION: CICS JVM server suffers hang and        *
  *                      possible abend 0C2 in DFHKETCB.         *
  ****************************************************************
  * RECOMMENDATION: .                                            *
  ****************************************************************
  A CICS JVM server pthread terminates and uses storage which
  has previously been FREEMAINed. This can lead to hangs and
  an abend0C2 on an SPKA instruction in DFHKETCB.
  

Problem conclusion

• DFHSJIN has been changed to set a pointer to nulls when
  the control block that it addresses is freed.
  DFHSJSA has been changed to check for a null pointer before
  accessing the control block.
  

Temporary fix

Comments

APAR Information

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

  PI79050

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

  UI47135 UI47136 PI83322

Modules/Macros

• DFHSJIN  DFHSJSA
  

Fix information

• Fixed component name

  CICS TS Z/OS V5

• Fixed component ID

  5655Y0400

Applicable component levels

• R800 PSY UI47135

     UP17/05/15 P F705

• R900 PSY UI47136

     UP17/05/15 P F705

Fix is available

• Select the PTF appropriate for your component level. You will be required to sign in. Distribution on physical media is not available in all countries.