PM86667: EYUVC1278E EYUVC1293E NO RESPONSE FROM CMAS, TIMEOUT ON WUI WLMAWORK VIEW WHEN WORKLOAD OPTIMIZED

A fix is available

Subscribe

You can track all active APARs for this component.

APAR status

  • Closed as program error.

Error description

  • You use the WUI to look at Active Workloads. Some of your
    workloads have implemented Sysplex Optimized Workload
    management. Your WUI browser hangs, and eventually you receive
    a series of messages similar to the following:
    .
    EYUVC1278E Unable to obtain data for requested View.
               GET request failed. (Environerror, Reqtimeout).
    EYUVC1293E No response received from CMAS xxxxxxxx.
               Reason: Execution Time Out )225).
    .
    In the AUX trace for the CMAS where the WUI is connected, you
    see trace records showing the timeout
    .
    TASK  MTD  PREV TRAN OBJ LEVL PT-ID   DEBUG  UOW  CMAS/USR ENVR
    12345 CSLT XLOP LSRT COM EXCP     16 TIMEOUT CPSM xxxxxxxx CMAS
    .
    In the AUX trace for the CMAS mentioned in the EYUVC1293E
    message, you see trace records showing
    .
    TASK: 12345 METHOD:CTSD   PRIOR:CTST  DEBUG:EXCEPT
                POINTID:     3            TRANID:LNMS
                MAJOBJ:COM    ENVRN:CMAS  TYPE:EXCEPTION
                UOW(CPSM):    SYSTM:yyyy  NAME:xxxxxxxx
    .
    The POINTID of 3 out of CTSD corresponds to
    CMPI_QUE_MAX_EXCEEDED.
    .
    Additional Keywords/Symptoms:
    EYUVC1278 EYUVC1293 CTSD_QUE_TOO_BIG XQIQ_MAX_ENTRY_LENGTH
    CMPI_QUE_MAX 4096
    KIXREVEPH
    

Local fix

Problem summary

  • ****************************************************************
    * USERS AFFECTED: All CICSPlex SM V5R1M0 Users                 *
    ****************************************************************
    * PROBLEM DESCRIPTION: -  When more than 26 workloads in the   *
    *                         same CICSplex are managed by a CMAS, *
    *                         errors will occur when WLMAWORK      *
    *                         records are retrieved from the CMAS  *
    *                         through the API, WUI, CMCI or CICS   *
    *                         Explorer.                            *
    *                                                              *
    *                         - If the request is entered through  *
    *                           an API program or WUI server       *
    *                           connected to the CMAS, then        *
    *                           invalid data may be returned and a *
    *                           storage overlay may occur.         *
    *                                                              *
    *                         - If the request is entered through  *
    *                           an API program or WUI server       *
    *                           connected to another CMAS, then    *
    *                           the request will hang until the    *
    *                           time out limit is reached, and no  *
    *                           data will be returned by the CMAS. *
    *                                                              *
    *                      -  When WLMATGRP or WLMAWORK records    *
    *                         are retrieved through the API, WUI,  *
    *                         CMCI or CICS Explorer, CPSM queue    *
    *                         resources will be orphaned.          *
    ****************************************************************
    * RECOMMENDATION: After applying the PTF that resolves this    *
    *                 APAR, all CMASes must be restarted.  Note    *
    *                 that the restarts do not need to occur at    *
    *                 the same time.                               *
    ****************************************************************
    -  When a get request for WLMAWORK records is received for a
       CICSplex, method EYU0WABW (WABW) runs in all CMASes that
       manage the CICSplex.  WABW builds a WLMAWORK record for each
       workload that the CMAS manages, and adds it to a buffer.
       When the buffer is full, WABW writes the buffer as a CPSM
       queue record.
    
       On entry to WABW, a calculation is made to determine how many
       WLMAWORK records will fit into a 4096 byte buffer.  This
       value is supposed to be used to determine how many WLMAWORK
       records are added to the buffer.  A logic error in WABM
       causes this value to be incorrectly modified, depending upon
       whether the WLM optimiation status of the workload is
       INACTIVE, ACTIVE or PARTIAL.  If the status is INACTIVE, then
       one record will be written to the buffer before it is written
       to the queue.  If the status is ACTIVE or PARTIAL, then all
       records will be written the buffer before it is written to
       the queue.
    
       The 4096 byte buffer allocated in the working storage for
       WABW is sufficient to contain 26 WLMAWORK records.  If the
       status of the workloads is ACTIVE or PARTIAL, and more than
       26 workloads are managed by the CMAS, then a buffer overrun
       will occur.  This may result in corrupted records and an
       overlay of storage.  Additionally, if the CMAS received the
       request from another CMAS, it must return the queue records
       to the originating CMAS.  When this occurs, method EYU0CTSD
       (CTSD) will fail the transmission because the queue record is
       over the 4096 byte maximum size.  When CTSD fails the
       transmission, it will issue an exception trace record with a
       trace point ID of 3 and a debug text of "EXCEPT".  Since no
       response is returned to the originating CMAS, the originating
       CMAS, and he user request, will wait until the time out limit
       expires.
    
    -  When a get request for WLMATGRP records is received for a
       workload, method EYU0WABG (WABG) runs in all CMASes that
       manage the workload.  If WABG determines that there are
       WLMATGRP records for the workload, it allocates a CPSM queue
       for temporary use.  On exit, WABG does not free the queue,
       which results in it and its resources being orphaned.
    
       When WABW runs in a CMAS to process a get request for
       WLMAWORK records, it calls WABG directly for each workload
       that the CMAS manages.  For each workload that has WLMATGRP
       records, a CPSM queue and its resources will be orphaned.
    

Problem conclusion

  • -  WABW has been updated to not incorrectly modify the queue
       record counter.
    
    -  WABG has been updated to correctly free the temporary queue
       on exit if it was allocated.
    

Temporary fix

  • FIX AVAILABLE BY PTF ONLY
    

Comments

APAR Information

  • APAR number

    PM86667

  • Reported component name

    CICS TS Z/OS V5

  • Reported component ID

    5655Y0400

  • Reported release

    80M

  • Status

    CLOSED PER

  • PE

    NoPE

  • HIPER

    NoHIPER

  • Special Attention

    NoSpecatt

  • Submitted date

    2013-04-09

  • Closed date

    2013-04-16

  • Last modified date

    2013-05-03

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

    PM86664

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

    UK93505

Modules/Macros

  •    EYU0WABG EYU0WABW
    

Fix information

  • Fixed component name

    CICS TS Z/OS V5

  • Fixed component ID

    5655Y0400

Applicable component levels

  • R80M PSY UK93505

       UP13/04/20 P F304

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.



Rate this page:

(0 users)Average rating

Document information


More support for:

z/OS family

Software version:

5.1

Reference #:

PM86667

Modified date:

2013-05-03

Translate my page

Machine Translation

Content navigation