IC83998: SUSPENDED THE GM & GC BETWEEN SIDE B AND SIDE C USING TPCRM WAIT FOR SUSPENDED

APAR status

• Closed as program error.

Error description

• The GM session get suspended but GC stay at Copy Pending on all
  device pairs.
  
   csmMessage.log shows:
  [Server]  IWNR6000I [May 31, 2012 9:42:00 AM] Starting all pairs
  in role pair H1-J2 ...
  [Server]  IWNR1041I [May 31, 2012 9:42:01 AM] The command start
  was successfully issued to all pairs under role pair H1-J2 for
  session GM_Discount_site3.
  [Server]  IWNR1950I [May 31, 2012 9:42:02 AM] Session
  GM_Discount_site3 changed from the Preparing state to the
  Prepared state.
  [Server]  IWNH1037E [May 31, 2012 9:43:59 AM] Device
  DS8000:BOX:2107.XXXXX managed by hardware connection
  ESSNI:ip-address::2107.XXXXX is no longer accessible.
  [Server]  IWNH1037E [May 31, 2012 9:43:59 AM] Device
  DS8000:BOX:2107.XXXXX managed by hardware connection
  ESSNI:ip-address::2107.XXXXX is no longer accessible.
  [Server]  IWNR1955E [May 31, 2012 9:43:59 AM] Tivoli Storage
  Productivity Center for Replication Server (server name) has
  encountered communication errors with storage system
  DS8000:BOX:2107.XXXXX.
  [UNKNOWN]  IWNH0007I [May 31, 2012 9:44:02 AM] The hardware
  device has been stopped for location
  ESSNI:ip-address::2107.XXXXX.
  [Server]  IWNH1037E [May 31, 2012 9:44:02 AM] Device
  DS8000:BOX:2107.XXXXX managed by hardware connection
  ESSNI:ip-address::2107.XXXXX is no longer accessible.
  [Server]  IWNH1037E [May 31, 2012 9:44:02 AM] Device
  DS8000:BOX:2107.XXXXX managed by hardware connection
  ESSNI:ip-address::2107.XXXXX is no longer accessible.
  [UNKNOWN]  IWNH0007I [May 31, 2012 9:44:05 AM] The hardware
  device has been stopped for location
  ESSNI:ip-address::2107.XXXXX.
  [UNKNOWN]  IWNR3081I [May 31, 2012 9:44:05 AM] High availability
  for the server (server name) has been shutdown.
  [UNKNOWN]  IWNR1901I [May 31, 2012 9:44:08 AM] Copy Services
  Manager has stopped successfully.
  ************ Start Display Current Environment ************
  Replication Manager - build: k20111212-1654 version: 4.2.2.1
  Hostname: (server name) -- (server name)/ip-address
  ************* End Display Current Environment *************
  [Server]  IWNR3080I [May 31, 2012 9:44:31 AM] High availability
  for the server (server name) is now running.
  [Server]  IWNH0001I [May 31, 2012 9:44:32 AM] The hardware
  device started.
  [Server]  IWNH1600I [May 31, 2012 9:44:32 AM] The SNMP event
  listener has been successfully started and is now ready for SNMP
  traps.
  [Server]  IWNH0001I [May 31, 2012 9:44:32 AM] The hardware
  device started.
  [Server]  IWNH0002I [May 31, 2012 9:44:38 AM] The hardware
  device started for location ESSNI:ip-address::2107.XXXXX.
  [Server]  IWNH0002I [May 31, 2012 9:44:43 AM] The hardware
  device started for location ESSNI:ip-address::2107.XXXXX.
  [Server]  IWNR1900I [May 31, 2012 9:44:44 AM] Copy Services
  Manager version 4.2.2.1 has started successfully.
  [Server]  IWNH1038I [May 31, 2012 9:44:46 AM] Device
  DS8000:BOX:2107.XXXXX managed by hardware connection
  ESSNI:ip-address::2107.XXXXX is now accessible.
  

Local fix

• There is a known defect which is fixed in 4.2.2.2.
  

Problem summary

• After a suspend command the hardware does not show GC pairs
  suspended but TPC-R does show that each role pair was suspended.
  

Problem conclusion

• This is caused by a race condition after issuing Pause to the
  master on the hardware. TPC-R kicks off the pause poller that
  queries the master for pause completion. The flag to tell the
  code not to suspend the GC pairs caused the Pause Poller to end
  up suspending the GC pairs AFTER the session had already been
  restarted.
  

Temporary fix

• Use dscli to manually suspend the pairs
  

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

  TPC FOR REPL SE

• Fixed component ID

  5608TRMSV

Applicable component levels

• R420 PSY

     UP