IC89994: PERFORMANCE DEGRADATION OF REPLICATION PROCESSING WHILE DELETING TARGET FILES FOR A NODE

APAR status

• Closed as program error.

Error description

• The Tivoli Storage Manager server REPLICATE NODE process can
  experience performance degradation when deleting large numbers
  of files for a replication target server.  This is more likely
  to occur when particular node/filespace is not replicating to
  completion daily, causing a large backlog of deletion work
  that needs to happen in order to sync the node across both
  servers.
  
  If the REPLICATE NODE process listed by QUERY PROCESS shows
  millions of objects under the "Files deleted" category, and
  the process has been running for longer than 24 hours, this
  issue likely applies.  For example:
  
   Process Process Description  Status
    Number
   ------- -------------------  ----------------------------
         8     Replicate Node  Replicating node(s) NODE.
                               File spaces complete: 11.
                               File spaces identifying and
                                replicating: 1.
                               File spaces replicating: 0.
                               File spaces not started: 0.
                               Files current: 27,419,278.
                               Files replicated: 6,775 of
                                6,779.
                               Files updated: 223,514 of
                                223,553.
               ------------>   Files deleted: 3,499 of
                                81,190,399.
                               Amount replicated: 2,990 MB
                                of 2,990 MB.
                               Amount transferred: 1,312 MB.
               ------------>   Elapsed time: 12 Day(s),
                                22 Hour(s), 8 Minute(s).
  
  Customer/L2 Diagnostics:
   None
  
  Initial Impact:
  Medium
  
  Tivoli Storage Manager Versions Affected:
  All 6.3.0.000 or higher Tivoli Storage Manager servers.
  
  Additional Keywords:
  ZZTSM63 replication deletion slow perf performance 87953
  

Local fix

• Initially, the problematic node/filespace should be removed
  from replication processing to prevent impacting replication
  processing of other nodes/filespaces.
  
  Depending on the number of deletions that the target server
  needs to make, it may be faster to remove the problematic
  node/filespace (DEL FI, REMOVE NODE) on the target system
  and start a fresh replication process.
  
  
  Note that if deduplication is being used on the target,
  removing the problematic node and/or filespace may take an
  extended period of time to complete.  An alternative may be
  to rename the node/filespace and delete it at a later time.
  

Problem summary

• ****************************************************************
  * USERS AFFECTED: All Tivoli Storage Manager server users of   *
  *                 replication.                                 *
  ****************************************************************
  * PROBLEM DESCRIPTION: See ERROR DESCRIPTION.                  *
  ****************************************************************
  * RECOMMENDATION: Apply fixing level when available. This      *
  *                 problem is currently projected to be fixed   *
  *                 in level 6.3.4. Note that this is            *
  *                 subject to change at the discretion of IBM.  *
  ****************************************************************
  *
  

Problem conclusion

• This problem was fixed.
  

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

  TSM SERVER

• Fixed component ID

  5698ISMSV

Applicable component levels

• R63A PSY

     UP

• R63H PSY

     UP

• R63L PSY

     UP

• R63S PSY

     UP

• R63W PSY

     UP

• R63Z PSY

     UP