ModelUpdates and AlarmGlobalUpdates queues are very large and are delaying alarm processing
Article ID: 186397
CA Service Operations Insight (SOI)
We are seeing SOI processing alarms with huge delays in alarm updates. Reviewing further we see that the following job queues have very large Enqueue counts:
AlarmGlobalUpdates is at over 2 million ModelUpdates is over 32 million
What can we do to reduce these numbers and get SOI functioning properly again ?
Release : 4.2
Component : Service Operations Insight (SOI) Manager
The queue counts relate to the ca_ssa* tables and it is these tables that will need clearing.
Ran the following queries with SOI Manager down and ensuring a backup of the SAMStore has been taken.
This set of queries will cleanup any items still remaining from deleted notebooks:
delete from ca_ssa_ci_timestamp where id not in (select id from ca_ssa_ci_detail) delete from ca_ssa_tags where id not in (select id from ca_ssa_notebooks) delete from ca_ssa_notebooks_timestamp where id not in (select id from ca_ssa_notebooks)
This set of queries will cleanup any items left over from synchronization tasks:
delete from ca_ssa_tags where tagname like '%synchronization.inprogress%' delete from ca_ssa_tags where tagname like '%synched.pending%' delete from ca_ssa_tags where tagname='synched.known.mdrs'
The queries should not need to be run regularly but can be included as part of scheduled SOI Maintenance if desired.