We do not support changing the execution time of grooming or purging.... That's 
why these were not exposed via override.

Partitioning and grooming is hard coded to 12:00am, and Purging happens two 
hours later at 2am.

These are not designed to run at the same time.  We partition, then we groom, 
then, we purge what was groomed.


If you don't mess with these times, and either is failing, then I'd review the 
logs....

select * from InternalJobHistory order by InternalJobHistoryId DESC

If these are failing, it is almost always a SQL performance issue, or a massive 
amount of data - meaning it is trying to groom/purge too much, because there is 
too much noise.




From: [email protected] [mailto:[email protected]] On 
Behalf Of Gareth Miles
Sent: Wednesday, May 24, 2017 7:29 AM
To: [email protected]
Subject: [msmom] SCOM DB locks at specific time

Hi

I'm getting massive locking on the OperationsManager DB at 02:00 local time.
I have both the 'Discovery Data Grooming' and 'Partitioning and Grooming' set 
to run at 02:00.

I've run both  p_DataPurging and p_PartitioningAndGrooming manually and both 
run through successfully without any issues.

I've attached the locking history from 00:00 this morning to 04:00.
You can see the main locking starts at 02:01 with the proc p_AlertDataPurging.
I don't see it listed in this post so I don't know if it's part of grooming or 
not?

The PID 370 from p_AlertCreateProcess which runs a select from the Alert table 
and attempts to create the proc p_AlertCreateProcess
This is the main lock for the period of the locking, until we shut down all 
management server services at 02:48.
If you filter the column 'Query' in the spreadsheet on p_AlertCreateProcess 
you'll see what I mean.

I've also run p_AlertGrooming, this runs in a few seconds without issue.

Any help would be appreciated.

Kind Regards
Gareth Miles






Reply via email to