PK91816: Concurrent PE Server subtasks that monitor members of the same DSG can cause deadlocks and timeouts

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• When two PE server subtasks are started on two separate LPARs
  occassionally they can cause deadlocks or timeouts. Example of
  deadlock messages seen in DB2mstr log:
  DSNT375I  #DB2U PLAN=OCP60E21 WITH  725
                CORRELATION-ID=POOLW4T10011
                CONNECTION-ID=CP1EE09
                LUW-ID=NETFBA.LUDB2U.C435D982B6DF=57914
                THREAD-INFO=$CEPLT:*:*:*
                IS DEADLOCKED WITH PLAN=OCP60E21 WITH
                CORRELATION-ID=POOLW4T10004
                CONNECTION-ID=CP1EE09
                LUW-ID=NETFBA.LUDB2U.C435D982B7AC=57915
                THREAD-INFO=$CEPLT:*:*:*
         ON MEMBER DB2U
  DSNT501I  #DB2U DSNILMCL RESOURCE UNAVAILABLE  726
                CORRELATION-ID=POOLW4T10011
                CONNECTION-ID=CP1EE09
                LUW-ID=NETFBA.LUDB2U.C435D982B6DF=258043
                REASON 00C90088
                TYPE 00000302
                NAME OCPDB21 .OCPTSVO .X'00700472'
  .
    Also timeout messages occur with message DSNT376I and
    DSNT501I  #DB4U DSNILMCL RESOURCE UNAVAILABLE 063
                   CORRELATION-ID=
                   CONNECTION-ID=RRSAF
                   LUW-ID=NETFBA.LUDB4U.C435D264853F=0
                   REASON 00C900BA
                   TYPE 00002002
    and DSNT501I with REASON 00C9008E and TYPE 00000304
  

Local fix

• Please apply PTF when available.
  

Problem summary

• ****************************************************************
  * USERS AFFECTED: OMEGAMON XE for DB2 PE/DB2PM                 *
  *                 users of component:                          *
  *                 -   PE Server subtask                        *
  ****************************************************************
  * PROBLEM DESCRIPTION: When several PE Server subtasks are     *
  *                      simultaneously running on different     *
  *                      LPARs and are monitoring members of     *
  *                      the same DB2 data sharing group,        *
  *                      they can occasionally cause deadlocks   *
  *                      or timeouts.                            *
  ****************************************************************
  * RECOMMENDATION: Apply this PTF                               *
  *                                                              *
  ****************************************************************
  PROBLEM SUMMARY:
  The problem can occur if several PE Server subtasks:
  - Are simultaneously monitoring members of the same DB2 data
    sharing group, and
  - Are running on different LPARs, and
  - Are trying to insert or update rows of the DB2PM.PARAMETER
    table at the same time.
  
  The DB2PM.PARAMETER table maintains the status of running PE
  Server subtasks (master or slave). A status specifies which
  PE Server subtask is allowed to collect and handle DB2 connect
  data.
  Concurrent updates to the DB2PM.PARAMETER table might cause
  several PE Server subtasks running in "master" status, which
  results in deadlocks and timeouts against the KO2PLAN plan.
  
  PROBLEM CONCLUSION:
  A PE Server subtask that modifies the DB2PM.PARAMETER table
  locks the table until all necessary SQL statements are
  performed. The table is released by a corresponding COMMIT
  statement. During this time requests from other PE Server
  subtasks are blocked.
  

Problem conclusion

• A PE Server subtask that modifies the DB2PM.PARAMETER table
  locks the table until all necessary SQL statements are
  performed. The table is released by a corresponding COMMIT
  statement. During this time requests from other PE Server
  subtasks are blocked.
  
  The service level of the
  OMEGAMON collector PE Server subtask
  FPEV0000I  DATA SERVER INITIALIZATION
  FPEV0114I  PTF BUILD IS 09.310for V 4.1
  FPEV0114I  PTF BUILD IS 09.310 for V 4.2
  
  KEYWORDS :
  DSNT375I IS DEADLOCKED WITH PLAN=KO2PLAN
  

Temporary fix

Comments

APAR Information

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

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

  UK51855 UK51856

Modules/Macros

• B390SPA1 B390SPA2
  

Fix information

• Fixed component name

  OM XE DB2PE/PM

• Fixed component ID

  5655OPE00

Applicable component levels

• R410 PSY UK51855

     UP09/11/17 P F911

• R420 PSY UK51856

     UP09/11/17 P F911

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.