PK17143: ABEND 0C4 DURING SERVER TERMINATION

APAR status
Closed as fixed if next.

Error description
Abend 0C4 when server was terminating due to some other problem.
In this case, the following variable was set to a port number
that was already in use and therefore, the server was not able
to initialize:
protocol_iiop_port

While it is expected for the server to terminate, the
termination should not cause an 0C4 abend.

From the dump:
Traceback for the TCB with Completion Code 0C4:
PU Addr   PU Offset  Entry         E Addr
4B183EA0  +0000041A                4B183EA0
4B1832C8  +0000011E                4B1832C8
07BE63C8  -00000106  CEEPGTFN      07BE6268
00D116E8  +00000932  CEEOPCMM      00D116E8

From systrace:
008F 009DED90   PGM   010 072C0400 CB11B762  00040010 00000000
                                                      4B11B000

PSW address: 4B11B762 - the storage is no longer there...  If
you look at the page not found 4B11B000, you'll see it's not
there either.

From detailed failing TCB view, there is still an RB active that
requested the storage that is no longer available.

GTF trace showed the storage in question was obtained by BBOCTL
task and was freed later by BBOBOA.

When address space is terminating, MVS is doing a cleanup of the
threads.  There is a window where main WAS task exits (without
executing TTIMER CANCEL code in the termination path) and other
threads are still running before mvs terminates them.

This problem is already fixed in base of V5.1 (MD08764)
Local fix
Fix the original error that caused server termination.  Ignore
0C4 abend.
Problem summary
****************************************************************
* USERS AFFECTED: All users of WebSphere Application Server    *
*                 version 5.1.0 for z/OS                       *
****************************************************************
* PROBLEM DESCRIPTION: ABEND0C4/ABENDS0C4 after TCP Listeners  *
*                      initialization fails with minor code    *
*                      C9C21152.                               *
****************************************************************
* RECOMMENDATION:                                              *
****************************************************************
A failure to initialize the TCP listeners, which causes the
WebSphere Application Server to end abnormally with minor code
C9C21152, might also result in ABEND0C4/ABENDS0C4 failures on
other threads.
Problem conclusion Temporary fix Comments
The subsequent ABEND0C4/ABENDS0C4 failures can be ignored.
Those failures are fallout from the C9C21152 failure, which
caused the main task to terminate.  There is a period of time
after the main task terminates, but before the address space is
cleaned up, where other threads are still executing.  It is
during this window where those other threads might experience
an ABEND0C4/ABENDS0C4 due to the aborted main task.

This APAR is being closed FIN with concurrence from the
submitting customer. A solution to this problem will be
delivered in a WebSphere Application Server for z/OS release
within the next 18 months.
APAR information
APAR number PK17143
Reported component name WEBSPHERE FOR Z
Reported component ID 5655I3500
Reported release 500
Status CLOSED FIN
PE NoPE
HIPER NoHIPER
Special Attention NoSpecatt
Submitted date 2005-12-22
Closed date 2006-04-17
Last modified date 2006-04-17

APAR is sysrouted FROM one or more of the following:

APAR is sysrouted TO one or more of the following:

Modules/Macros

Publications Referenced

Fix information

Applicable component levels
R500 PSN    UP


Document Information


Current web document: swg1PK17143.html
Product categories: Software > Application Servers > Distributed Application & Web Servers > WebSphere Application Server for z/OS
Operating system(s):
Software version: 500
Software edition:
Reference #: PK17143
IBM Group: Software Group
Modified date: Apr 17, 2006