How to Monitor ESP Slowdowns
Article ID: 252758
Updated On:
Products
Issue/Introduction
The normal the time between when a job is ready and when it's scheduled is a second or two. However there are times a job's resources are satisfied and the job is readied, until the job is scheduled, can take 2,3, 4 up to 10+ minutes
Environment
Release : ESP Workload Automation
Cause
When using ESP Enclaves and WLM service classes, the workload on event initiators is running with low priority.
Resolution
After setting up Enclaves in ESP there is a need to focus on the WLM setup that might affect ESP Initiators priority. The thing to understand here is that the particular WLM service class setting will not affect just the job after it is submitted, but it will also affect ESP processing on the ESP event initiator that leads to the job's submission.
The idea of implementation of WLM Enclaves in ESP is to attribute the workload processing to a particular user instead of ESP started task. It starts with ESP joining the enclave for the particular user in the early stages of Event initiator processing.
Check/set these parameters in WLM option 6. Classification Rules
A. SST : ESP use option 1 Create (a new type)
B. SI (Subsystem Instance) : espp From your SUBSYS initparm in ESP - You can set this pretty high
C. UI (User ID) : produser
D. TC (Transaction Class) : ESPEVENT You can set this pretty high
E. TN (Transaction Name representing ESP Application name) : bigappl
Additional Information
Enclaves in ESP Workload Automtation.
Feedback
