IT29648: APPLICATIONS MAY HANG DURING TAKEOVER IN AUTOMATED TSA HADR ENVIRONMENT DURING FIX PACK UPGRADE

Db2 Version 11.1 Mod 4 Fix Pack 5 for Linux, UNIX, and Windows
Db2 Version 11.1 Mod 4 Fix Pack 6 for Linux, UNIX, and Windows
Db2 Version 11.1 Mod 4 Fix Pack 6 for Linux, UNIX, and Windows

APAR status

• Closed as program error.

Error description

• When performing a rolling fixpack upgrade in V11.1 in a
  automated TSA HADR environment, Application/database
  connectivity may hang during a takeover from the downlevel
  primary to the uplevel standby database due to peer window.
  
  Hang duration will usually be equal to the value of
  HADR_PEER_WINDOW.  We should not trigger peer window in this
  rolling update scenario.
  
  Example, from the db2diag.log from uplevel Standby, we could see
  that takeover completed successfully within a minute, but still
  application connection hangs until we see the message "Peer
  window ends. Peer window expired.".
  
  2019-07-01-11.52.10.587133-300 E14508358A508        LEVEL: Event
  PID     : 66584616             TID : 3600           PROC :
  db2sysc 0
  INSTANCE: db2inst1              NODE : 000           DB   :
  SAMPLE
  APPHDL  : 0-9                  APPID:
  *LOCAL.db2inst1.190701165210
  AUTHID  : db2inst1              HOSTNAME: StandbyHost
  EDUID   : 3600                 EDUNAME: db2agent (SAMPLE) 0
  FUNCTION: DB2 UDB, base sys utilities,
  sqeDBMgr::StartUsingLocalDatabase, probe:13
  START   : Received TAKEOVER HADR command.
  
  2019-07-01-11.52.17.728295-300 I14535698A520        LEVEL:
  Warning
  PID     : 66584616             TID : 10796          PROC :
  db2sysc 0
  INSTANCE: db2inst1              NODE : 000           DB   :
  SAMPLE
  HOSTNAME: StandbyHost
  EDUID   : 10796                EDUNAME: db2hadrs.0.0 (SAMPLE) 0
  FUNCTION: DB2 UDB, High Availability Disaster Recovery,
  hdrStbyTkHandlePrimaryDone, probe:46590
  MESSAGE : Rolling upgrade: Standby is on old version.  Closing
  connection to
            avoid shipping new log records to standby
  
  2019-07-01-11.52.17.729368-300 E14536785A450        LEVEL: Event
  PID     : 66584616             TID : 10796          PROC :
  db2sysc 0
  INSTANCE: db2inst1              NODE : 000           DB   :
  SAMPLE
  HOSTNAME: StandbyHost
  EDUID   : 10796                EDUNAME: db2hadrs.0.0 (SAMPLE) 0
  FUNCTION: DB2 UDB, High Availability Disaster Recovery,
  hdrSetHdrState, probe:10000
  CHANGE  : HADR state set to HDR_P_DISCONN_PEER (was HDR_P_PEER),
  connId=3
  
  2019-07-01-11.52.17.817201-300 I14540828A443        LEVEL: Info
  PID     : 66584616             TID : 10796          PROC :
  db2sysc 0
  INSTANCE: db2inst1              NODE : 000           DB   :
  SAMPLE
  HOSTNAME: StandbyHost
  EDUID   : 10796                EDUNAME: db2hadrp.0.1 (SAMPLE) 0
  FUNCTION: DB2 UDB, High Availability Disaster Recovery,
  hdrStbyTkHandlePrimaryDone, probe:46630
  MESSAGE : Standby has completed takeover (now primary).
  
  
  We could only see these messages until 11.57
  
  2019-07-01-11.57.11.036030-300 I14558583A537        LEVEL: Error
  PID     : 66584616             TID : 10796          PROC :
  db2sysc 0
  INSTANCE: db2inst1              NODE : 000           DB   :
  SAMPLE
  HOSTNAME: StandbyHost
  EDUID   : 10796                EDUNAME: db2hadrp.0.1 (SAMPLE) 0
  FUNCTION: DB2 UDB, High Availability Disaster Recovery,
  hdrSendRedirectMsgToOneAddress, probe:31070
  MESSAGE : ZRC=0xFFFFFFFF=-1
  DATA #1 : <preformatted>
  The HADR primary was not able to form a TCP connection with the
  standby: 10.27.98.91:60044.
  .
  .
  .
  2019-07-01-11.57.18.057736-300 I14559121A430        LEVEL:
  Warning
  PID     : 66584616             TID : 10796          PROC :
  db2sysc 0
  INSTANCE: db2inst1              NODE : 000           DB   :
  SAMPLE
  HOSTNAME: StandbyHost
  EDUID   : 10796                EDUNAME: db2hadrp.0.1 (SAMPLE) 0
  FUNCTION: DB2 UDB, High Availability Disaster Recovery,
  hdrEduAcceptEvent, probe:20202
  MESSAGE : Peer window ends. Peer window expired.
  

Local fix

• As a workaround, you can temporarily disable peer window before
  performing rolling update.
  

Problem summary

• ****************************************************************
  * USERS AFFECTED:                                              *
  * ALL                                                          *
  ****************************************************************
  * PROBLEM DESCRIPTION:                                         *
  * See Error Description                                        *
  ****************************************************************
  * RECOMMENDATION:                                              *
  * Upgrade to Db2 11.1 Mod 4 Fixpack 5 or higher                *
  ****************************************************************
  

Problem conclusion

• First fixed in Db2 11.1 Mod 4 Fixpack 5
  

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:

Fix information

• Fixed component name

  DB2 FOR LUW

• Fixed component ID

  DB2FORLUW

Applicable component levels

• RB10 PSN

     UP