IBM Support

PI41732: HIGH CPU USAGE DETECTED IN CMAS

A fix is available

Subscribe

You can track all active APARs for this component.

 

APAR status

  • Closed as program error.

Error description

  • A problem occurred where the CMAS region began using high CPU.
    Looking at the CMAS dump, a loop is occurring between XSRR,
    XEPC, XSRA, XEPC trying to acquire a lock.  A task was waiting
    for the exclusive lock MCST_LOCK.  It's task structure may
    resemble:
    .
    Task Meth Load-Pt  OSSB     Stack    Mal      MODB     MOEB
    xxxx XLOP 000B3F28 306C5740 306C5768 00000000 2F0F0E60 2EC9B140
    xxxx XDNR 30051DA0 306C5740 306C58B0 306ACD6C 2F0F6360 2FF81060
    xxxx XDP4 3005A888 306C5740 306C63B8 306C6270 2F0F6360 2FF81060
    xxxx MOMS 3008EDA8 306C5740 306C6D68 306C66C0 2F0F6360 2FF81060
    xxxx XLSI 2F34B5B0 306C5740 306C79D8 306C76D0 2F0F0E60 2EC9B140
    xxxx XLSD 2F32D1A0 2FF23000 2FF23170 306C7D60 2F0F0E60 2EC9B140
    xxxx MMID 30202620 2FF23000 2FF24A30 306A9678 2F0FA760 305EB140
    xxxx MMIC 30207888 2FF23000 2FF252C0 2FF24C88 2F0FA760 305EB140
    xxxx XSRA 2F383FD8 2FF23000 2FF25960 2FF253E8 2F0F4160 000CCC90
    xxxx XSLA 2F37A4A0 2FF23000 2FF25C08 2FF25A88 2F0F4160 000CCC90
    xxxx XSWX 2F37CC60 2FF23000 2FF25E38 2FF25D20 2F0F4160 000CCC90
    .
    A second task is trying to install a monitor definition and its
    taxk structure resembles:
    .
    Task Meth Load-Pt  OSSB     Stack    Mal      MODB     MOEB
    xxxx XLOP 000B3F28 306D6740 306D6768 00000000 2F0F0E60 2EC9B140
    xxxx XLST 2F34D1E0 306D6740 306D68B0 306B54BC 2F0F0E60 2EC9B140
    xxxx MMIT 30201D70 306D6740 306D7088 306AAA64 2F0FA760 305EB140
    xxxx MMIC 30207888 306D6740 306D7900 306D7220 2F0FA760 305EB140
    xxxx MMIB 302069C8 306D6740 306D7FA0 306D7BA0 2F0FA760 305EB140
    xxxx XCLA 2F3A1208 306D6740 306D8350 306D80D0 2F0F5260 2EC9B660
    xxxx XSRR 2F36C548 306D6740 306D88E8 306D87A8 2F0F4160 000CCC90
    A third task that currently owns the MCST_LOCK has the
    following task structure:
    .
    Task Meth Load-Pt  OSSB     Stack    Mal      MODB     MOEB
    xxxx XLOP 000B3F28 306BB740 306BB768 00000000 2F0F0E60 2EC9B140
    xxxx XLST 2F34D1E0 306BB740 306BB8B0 306ACE7C 2F0F0E60 2EC9B140
    xxxx MMIT 30201D70 306BB740 306BC088 306AC084 2F0FA760 305EB140
    xxxx MMIC 30207888 306BB740 306BC900 306BC220 2F0FA760 305EB140
    xxxx MMIB 302069C8 306BB740 306BCFA0 306BCBA0 2F0FA760 305EB140
    xxxx MADS 3023CDB8 306BB740 306BD350 306BD0D0 2F0FA760 305EB140
    xxxx XSSP 2F382790 306BB740 306BD718 306BD6F8 2F0F4160 000CCC90
    .
    Both of these task are looping trying to install MONDEF MON0001
    with PERIODEF PDEF0002 but found that the MONDEF MON0001 was
    already installed with PERIODEF PDEF0001.  This prevents the
    MCST_LOCK from being freed allowing any other task to acquire
    it.
    .
    .
    Additional Symptom(s) Search Keyword(s): KIXREVGJT
    

Local fix

Problem summary

  • ****************************************************************
    * USERS AFFECTED: All CICSPlex SM V5R1M0 and V5R2M0 Users      *
    ****************************************************************
    * PROBLEM DESCRIPTION: A CMAS region may experience high CPU   *
    *                      usage caused by a loop after installing *
    *                      a MONDEF with a PERIODEF, which is      *
    *                      already active under another PERIODEF.  *
    *                                                              *
    *                      After CMAS shutdown, you may find the   *
    *                      similar messages in the joblog:         *
    *                                                              *
    *                      EYUXL0033I xxxxxxxx Attempting to PURGE *
    *                      TRANID(MCCM), TASKID(xxxxx), METHOD     *
    *                      (XSSP), CALLER(MCCD).                   *
    *                                                              *
    *                      EYUMM0601E xxxxxxxx Monitoring Time     *
    *                      Definition (xxxxxxxx) failed to install *
    *                      for Context(xxxxxxx) Scope(xxxxxxxx)    *
    *                      Reason: 3 -- MMICMMIB.                  *
    ****************************************************************
    * RECOMMENDATION: After applying the PTF that resolves this    *
    *                 APAR, all CMASes must be recycled to pick up *
    *                 the new code.  Note that regions do not need *
    *                 to be brought down and restarted at the same *
    *                 time.                                        *
    ****************************************************************
    During a MONDEF installation, method MMIB is called to create
    MDLE (Monitor Definition List Entry) and MDMS (Monitor Master
    Definition).
    
    If the monitor definition has already been installed on a MAS
    region, a duplicate key MDMS element is found in the master list
    . Method MMIB calls MADS to discard the installed definition. As
    part of MADS process, MTPL (Time Period Definition List Element)
    is discarded.
    
    If the installing MDMS is exactly the existing one, the unique
    MTPL element is discarded.
    
    Later, if another installation for the MONDEF is performed on
    the MAS region with another active PERIODEF, MADS is called one
    more time to discard the installed definition. Since the unique
    MTPL element has already been discarded, MADS_DEF_NOT_FOUND is
    returned to MMIB. According to the current code logic, method
    MMIB loop executes the creation and discard process, which
    causes high CPU usage in the CMAS region.
    

Problem conclusion

  • The following methods have been changed:
    
    - Method MMIB (EYU0MMIB) has been updated to re-add MTPL element
      to the Time Period Definition List after the discard process
      if the installing MDMS is exactly the same as the existing one
    
    - Method MADS (EYU0MADS) has been updated to write an exception
      trace entry if definition is not found
    
    - Method MMKP (EYU0MMKP) has been updated to write an exception
      trace entry if MTPL element is not found
    

Temporary fix

  • *********
                * HIPER *
                *********
    FIX AVAILABLE BY PTF ONLY
    

Comments

APAR Information

  • APAR number

    PI41732

  • Reported component name

    CICS TS Z/OS V5

  • Reported component ID

    5655Y0400

  • Reported release

    80M

  • Status

    CLOSED PER

  • PE

    NoPE

  • HIPER

    YesHIPER

  • Special Attention

    NoSpecatt

  • Submitted date

    2015-05-26

  • Closed date

    2015-06-23

  • Last modified date

    2015-07-01

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

    PI39377

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

    UI28715 UI28716

Modules/Macros

  • EYU0MADS EYU0MMIB EYU0MMKP
    

Fix information

  • Fixed component name

    CICS TS Z/OS V5

  • Fixed component ID

    5655Y0400

Applicable component levels

  • R80M PSY UI28715

       UP15/06/27 P F506 ½

  • R90M PSY UI28716

       UP15/06/27 P F506 ½

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":"BU058","label":"IBM Infrastructure w\/TPS"},"Product":{"code":"SSGMGV","label":"CICS Transaction Server"},"Component":"","ARM Category":[],"Platform":[{"code":"PF025","label":"Platform Independent"}],"Version":"5.1","Edition":"","Line of Business":{"code":"LOB35","label":"Mainframe SW"}},{"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":"5.1","Edition":"","Line of Business":{"code":"","label":""}}]

Document Information

Modified date:
01 July 2015