IBM Support

PH00543: ABEND04E 00E50013 IN DSNRUC02 LEADS TO CHECKPOINT PROCESSOR 003.RCRSC HANG WHEN TRYING TO RECOVER FROM THE CANCEL.

A fix is available

Subscribe

You can track all active APARs for this component.

APAR status

  • Closed as program error.

Error description

  • ABEND04E in DSNRUC02 when executing recovery routine
    after an abend 00E50013. This leads to checkpoint processor
    003.RCRSC hang. An 8 byte overlay in latch word. Checkpoint
    processor thread remains suspended due to latch not getting
    resolved. Can lead to storage pool abend:
    .
    04E-00E2000CDSNILCUR,M=DSNTFRCV,LOC=DSNIRCSB.DSNSVSVB+00A54
    due to total 2GB of PDAs being allocated.
    
    Additional keywords and symptoms:
    ******************************************
    ABEND0C4 PIC00011 in DSNVDTA LOGREC entry
    00E10000 - overlay string in registers of LOGREC entry
    
    DB2OVRLAY/K DB2DSHR
    

Local fix

  • If this leads to restart issue where Db2 cannot be brought
    up then special usermod will be required to expand pool
    to allow for recovery processing.
    The USERMOD code changes the storage size, which Db2 uses for
    Db2 internal control block PDA, from 2GB to 16GB so that
    Db2 can successfully restart to process log records.
    
    To avoid the storage pool from exceeding limit, customers should
    monitor for the following message and take action:
    .
    DSNJ016E   csect-name WARNING - SYSTEM CHECKPOINT PROCESSOR MAY
    HAVE STALLED. LAST CHECKPOINT WAS TAKEN date-time.
    ...
    No local fix to avoid the overlay that can occur.
    

Problem summary

  • ****************************************************************
    * USERS AFFECTED:                                              *
    * All Db2 12 for z/OS users.                                   *
    ****************************************************************
    * PROBLEM DESCRIPTION:                                         *
    * Storage overlay of latch owner word                          *
    * causes the system checkpoint processor                       *
    * to wait endlessly.                                           *
    *                                                              *
    * Message MSGDSNJ016E issued for the                           *
    * stalled system checkpoint processor.                         *
    ****************************************************************
    * RECOMMENDATION:                                              *
    * Apply corrective PTF when available                          *
    ****************************************************************
    A storage overlay pattern of '80E10000'X was seen in the latch
    owner word for a latch in the GXID control block.  The same
    storage overlay pattern was also seen in the CAPL control
    block.
    
    A soft cancel abend, RC00E50013, occurred in CSECT DSNRUC02.
    This abend was retried, but when returning to the mainline
    code path in DSNRUC02, registers were not set up correctly
    in Db2 12 for z/OS.  Zero was incorrectly used for a pointer,
    which caused '00E10000'X to be incorrectly moved into a latch
    owner word in the GXID control block.
    
    This caused the system checkpoint processor to wait
    endlessly for the latch. Since system checkpoints were
    not being taken, there was a significant delay on the next
    restart of Db2.  Db2 restart would need to read logs from at
    least the last checkpoint.
    

Problem conclusion

  • To fix the storage overlay, DSNRUC02 was fixed to ensure the
    registers are set up correctly in the retry routine.
    
    DB2OVRLAY/K
    

Temporary fix

  • *********
    * HIPER *
    *********
    

Comments

APAR Information

  • APAR number

    PH00543

  • Reported component name

    DB2 OS/390 & Z/

  • Reported component ID

    5740XYR00

  • Reported release

    C10

  • Status

    CLOSED PER

  • PE

    NoPE

  • HIPER

    YesHIPER

  • Special Attention

    NoSpecatt / Xsystem

  • Submitted date

    2018-07-16

  • Closed date

    2018-09-27

  • Last modified date

    2018-11-01

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

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

    UI58787

Modules/Macros

  • DSNRUC02 DSNRUC01
    

Fix information

  • Fixed component name

    DB2 OS/390 & Z/

  • Fixed component ID

    5740XYR00

Applicable component levels

  • RC10 PSY UI58787

       UP18/10/12 P F810 ¢

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.



Document information

More support for: DB2 for z/OS

Software version: 12.0

Reference #: PH00543

Modified date: 01 November 2018