Problem: How to determine which job caused high CPU usage on a specific compute node.
Using the IIS console 'Activity > Resources' narrow down the window of the event that occurred on the specific compute node.
Parse the 'dslogger' of the specific compute node to find the process that had the high CPU within the date/time range found in step 1.
Parse the 'dslogger' of the head node for the ID and date/time range to find the name of the sequence job.
With the job name identified then I can research why the job consumed so much CPU.
|Who would benefit from this IDEA?||IIS Adminstrators that need to find root cause of systems resource usages that exceed the expected limits|
How should it work?
Add a compute node drop down on the 'Activity' page just like the 'Dashboard' has it for resources. Now the jobs that used the resources for that specific date/time range are identified
Add a 'Generate SQL' button to create the SQL used for the current selection. This SQL can now be used manually to get the raw data for further analysis.
|Priority Justification||The usage is always changing as new jobs are added and other modified. The constant need for proactive monitoring and root cause identification is a daily effort. This will speed up the root cause analysis and improve the system stability.|
|Customer Name||Erik Zollinger (Albertsons Companies Inc.)|
NOTICE TO EU RESIDENTS: per EU Data Protection Policy, if you wish to remove your personal information from the IBM ideas portal, please login to the ideas portal using your previously registered information then change your email to "email@example.com" and first name to "anonymous" and last name to "anonymous". This will ensure that IBM will not send any emails to you about all idea submissions