The DPdirect log has deadlock messages

If the DPdirect log file has deadlock messages, then replication will still occur correctly, but it will take more time as several attempts will be made to replicate the data. The log file will also regularly grow and need to be truncated.

Resource deadlocks typically occur when the top-of-hour processing and replication from other systems need to access the same tables

To solve this problem

Add a topOfHourDelay to the datapump configuration at the source system(s) for the table reported to cause the deadlock. The delay needs to:

• Be specified in minutes.

• Start at 5 minutes before the top-of-hour.

For example:

use xosapp

go

datapumpDPSysKeyTable 'accum', 'hour', -3, 'time', 'tagName', 'datasetId', NULL, NULL, 'DPkeyhour', @priority=7, @quote=1, @separator='\t', @bulkload=0, @catchUp=100, @topOfHourDelay=15

go