IBM Support

OA48280: TSO ADDRESS SPACE HANGS WITH ENQUEUE CONTENTION FOR MAJOR NAME SYSZTSOC MINOR IKJOBTAINMSG SCOPE=STEP

A fix is available

Subscribe

You can track all active APARs for this component.

 

APAR status

  • Closed as program error.

Error description

  • TSO batch job executing a REXX EXEC hangs with Enqueue
    contention on SYSZTSOC IKJOBTAINMSG.   The REXX EXEC
    establishes an EMCS CONSOLE that is defined to receive HARDCOPY
    (ie. SYSLOG) from every system in the SYSPLEX (MSCOPE=*ALL).
    This generated a large volume of message traffic to this EMCS
    console since there were 13 systems in the sysplex.  The
    REXX EXEC then issues a command (SETOMVS) followed by a
    GETMSG to retrieve messages.
    
    The error is seen when the system is constrained causing any
    delay to the command response expected by the REXX EXEC.  The
    EMCS dataspace quickly fills up with messages from all other
    systems in the sysplex and becomes suspended.  IKJCNANT will
    start to clear the dataspace by receiving messages from the
    dataspace so new messages can arrive.  The amount of storage
    required by the IKJCNANT task trying to deal with the sysplex
    message traffic causes it to run out of private storage
    resulting in an abend878 'out of private storage' condition in
    the TSO address space. IKJCNANT receives a return code x'18'
    from the MCSOPMSG GETMSG request indicating a system problem
    so IKJCNANT attempts to deactivate the console, cleanup and
    abort the REXX exec.
    
    IKJCNANT calls IKJCNASR to deactivate the console without
    releasing the SYSZTSOC IKJOBTAINMSG enqueue then waits
    in IKJCNASR+X'272' because there is an outstanding GETMSG
    request (CONSOLE_NUMBER_OF_REQUESTS is equal to 0001).  The
    GETMSG request will never complete (i.e. the count will never
    go to 0) because it is waiting for the SYSZTSOC IKJOBTAINMSG
    enqueue.  If the address space remains in this state (with an
    active console listening to all hardcopy message with
    MSCOPE=*ALL) then the system can eventually fail with
    abend878-10 in the CONSOLE address space with CONSOLEs running
    out of private storage due to excessive message traffic.
    
    Additional symptoms:
    CONSOLE ABEND878 RC10
    IEE286I CONSOLE DELAYED MESSAGE BUFFER FULL
    IXC431I GROUP SYSMCS MEMBER JOB CONSOLE ASID 000A STALLED AT
    
    Verification Steps in dump of hung user:
    1.  IP RESOURCE ANALYSIS should show:
        RESOURCE #0034:
        NAME=MAJOR=SYSZTSOC MINOR=IKJOBTAINMSG SCOPE=STEP
        RESOURCE #0034 IS HELD BY:
        JOBNAME=JOBA MASID=0096  TCB=00990E88
        RESOURCE #0034 IS REQUIRED BY:
        JOBNAME=JOBA MASID=0096  TCB=009AE438
    2.  TCB 00990E88 will be in a WAIT in IKJCNASR+x'272'
        (HTE7790 level)  Reg9 at time of WAIT will point
        to a control block with the ID of CONSOLE.
        CONSOLE+x'5C' will contain 80000001
        TCB 009AE438 will be waiting for the ENQUEUE
    

Local fix

  • Specify MSCOPE=*  instead of MSCOPE=*ALL to reduce the
    amount of message traffic
    .
    A ++APAR (AA48280) is available for HTE7780 (V1R13) and
    HTE7790 (V2R1) from L2 support.
    

Problem summary

  • ****************************************************************
    * USERS AFFECTED: All users of z/OS TSOE CONSOLE command       *
    ****************************************************************
    * PROBLEM DESCRIPTION: TSO Address Space may hang with ENQ     *
    *                      contention on Majorname='SYSZTSOC' and  *
    *                      Minorname='IKJOBTAINMSG' during a TSO/E *
    *                      EMCS CONSOLE session.                   *
    ****************************************************************
    * RECOMMENDATION:                                              *
    ****************************************************************
    TSO batch job executing a REXX EXEC hangs with Enqueue
    contention on SYSZTSOC IKJOBTAINMSG. (This is referred to
    as the cache enqueue.)  The REXX EXEC establishes an EMCS
    CONSOLE that is defined to receive HARDCOPY (ie. SYSLOG) from
    every system in the SYSPLEX (MSCOPE=*ALL).  This generated a
    large volume of message traffic to this EMCS console since
    there were 13 systems in the sysplex.  The REXX EXEC then
    issues a command (SETOMVS) followed by a GETMSG to retrieve
    messages.
    -
    The error is seen when the system is constrained causing any
    delay to the command response expected by the REXX EXEC.  The
    EMCS dataspace quickly fills up with messages from all other
    systems in the sysplex and becomes suspended.  IKJCNANT will
    start to clear the dataspace by receiving messages from the
    dataspace so new messages can arrive.  The amount of storage
    required by the IKJCNANT task trying to deal with the sysplex
    message traffic causes it to run out of private storage
    resulting in an abend878 'out of private storage' condition in
    the TSO address space. IKJCNANT receives a return code x'18'
    from the MCSOPMSG GETMSG request indicating a system problem
    so IKJCNANT attempts to deactivate the console, cleanup and
    abort the REXX exec.
    -
    IKJCNANT calls IKJCNASR to deactivate the console without
    releasing the SYSZTSOC IKJOBTAINMSG enqueue then waits
    in IKJCNASR+X'272' because there is an outstanding GETMSG
    request (CONSOLE_NUMBER_OF_REQUESTS is equal to 0001).  The
    GETMSG request will never complete (i.e. the count will never
    go to 0) because it is waiting for the SYSZTSOC IKJOBTAINMSG
    enqueue at IKJCNAGW+x'168' under the EXEC task.  If the
    address space remains in this state (with an active console
    listening to all hardcopy message with MSCOPE=*ALL) then the
    system can eventually fail with abend878-10 in the CONSOLE
    address space with CONSOLEs running out of private storage
    due to excessive message traffic.
    

Problem conclusion

  • IKJCNANT will release the console cache enqueue before trying
    to deactivate a console.
    

Temporary fix

  • *********
    * HIPER *
    *********
    SEE LOCAL FIX:
    SPECIFY MSCOPE=*  INSTEAD OF MSCOPE=*ALL TO REDUCE THE
    AMOUNT OF MESSAGE TRAFFIC
    

Comments

APAR Information

  • APAR number

    OA48280

  • Reported component name

    TSO/E SCHEDULAR

  • Reported component ID

    566528502

  • Reported release

    790

  • Status

    CLOSED PER

  • PE

    NoPE

  • HIPER

    YesHIPER

  • Special Attention

    NoSpecatt

  • Submitted date

    2015-07-03

  • Closed date

    2015-08-06

  • Last modified date

    2015-09-01

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

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

    UA78520 UA78521 UA78522

Modules/Macros

  • IKJCNANT
    

Fix information

  • Fixed component name

    TSO/E SCHEDULAR

  • Fixed component ID

    566528502

Applicable component levels

  • R7A0 PSY UA78520

       UP15/08/19 P F508

  • R780 PSY UA78521

       UP15/08/19 P F508

  • R790 PSY UA78522

       UP15/08/19 P F508

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.

[{"Business Unit":{"code":"BU054","label":"Systems w\/TPS"},"Product":{"code":"SG19M","label":"APARs - z\/OS environment"},"Component":"","ARM Category":[],"Platform":[{"code":"PF025","label":"Platform Independent"}],"Version":"790","Edition":"","Line of Business":{"code":"","label":""}},{"Business Unit":{"code":null,"label":null},"Product":{"code":"SG19O","label":"APARs - MVS environment"},"Component":"","ARM Category":[],"Platform":[{"code":"PF025","label":"Platform Independent"}],"Version":"790","Edition":"","Line of Business":{"code":"","label":""}}]

Document Information

Modified date:
01 September 2015