Concurrency during REORG

You can specify certain options that might prevent timeouts and deadlocks when you run the REORG utility.

To improve concurrency for REORG operations:

• If you encounter timeouts or deadlocks when you use REORG with the SHRLEVEL CHANGE option, run the REORG utility with the DRAIN ALL option.
  The default is DRAIN WRITERS, which is done in the log phase. The specification of DRAIN ALL indicates that both writers and readers are drained when the MAXRO threshold is reached.
• If application have long running readers that do not commit often or have overlapping claims, specify the FORCE option of the REORG utility.
  You can specify FORCE READERS to cancel read claims or FORCE ALL to cancel both read and write claims when REORG requests a drain on the last RETRY process. If you specify FORCE NONE, no read or write claims are canceled when REORG requests a drain.
• Consider the DRAIN ALL option in environments where a lot of update activity occurs during the log phase.
  With this specification, no subsequent drain is required in the switch phase.