Troubleshooting
Problem
If there is a significant delay between the end of SWITCHPLAN and the start of eligible jobs like UPDATESTATS then the Tivoli Workload Scheduler (TWS) topology configuration might need attention.
Symptom
Once the TWS production plan has been extended and initiated it is expected that job processing will continue.
Diagnosing The Problem
Inspect the workstation definitions to determine if they are sharing the master's mailman process. If the agents do not belong to a sub-domain and have not been assigned a mailman serverID then the master's mailman server is being used by the Fault Tolerant Agent (FTA).
Resolving The Problem
Assign mailman serverIDs for all agents so that no agent uses the same mailman process as the master. Also, consider creating sub-domains and assigning agents to the sub-domains (and use mailman serverIDs in the sub-domains.)
In one known case there were over 200 FTAs using the master's mailman server. The FINAL jobstream was taking over an hour to complete and the lag time between SWITCHPLAN and UPDATESTATS was as long as 30 minutes. After modifying the workstation definitions so that sub-domains were used and no workstation was using the master's mailman server the FINAL jobstream jobs had a combined duration of a few minutes and the noted lag time was reduced to two minutes.
Related Information
Product Synonym
Maestro;TWS;TWA
Was this topic helpful?
Document Information
Modified date:
17 June 2018
UID
swg21668658