UIM environment recommended action Items

Products

• DX Unified Infrastructure Management (Nimsoft / UIM)
• CA Unified Infrastructure Management On-Premise (Nimsoft / UIM)
• CA Unified Infrastructure Management SaaS (Nimsoft / UIM)

Issue/Introduction

• List of Action Items/Recommendations for customer's UIM Environment

Environment

• Release: 20.4 or higher

Cause

• System healthy and highly-operational for many years
• Change/loss of UIM administrator

Resolution

Recommended action items for the customer's UIM Environment:

- MS SQL Server UIM backend database is currently at 2012 SP4 and that version is unsupported so the DB needs to be updated to either 2016 or 2019 ASAP...

- Possible migration of select data based on time range after archiving the original DB for safekeeping.

- Currently, partitioning is working as expected-this has been confirmed.

- Running the large tables query showed over 50 or so tables in the tens of millions or hundreds of millions row range. 50M rows or higher slows down queries tremendously.

- Large tables/table indexes (see list below) should be reduced and/or defragmented.

ALTER INDEX ALL ON CM_COMPUTER_SYSTEM REBUILD; 
ALTER INDEX ALL ON CM_COMPUTER_SYSTEM_ATTR REBUILD;
ALTER INDEX ALL ON CM_DEVICE REBUILD; 
ALTER INDEX ALL ON CM_DEVICE_ATTRIBUTE REBUILD; 
ALTER INDEX ALL ON CM_CONFIGURATION_ITEM REBUILD;
ALTER INDEX ALL ON CM_CONFIGURATION_ITEM_METRIC REBUILD; 
ALTER INDEX ALL ON CM_CONFIGURATION_ITEM_DEFINITION REBUILD; 
ALTER INDEX ALL ON CM_CONFIGURATION_ITEM_METRIC_DEFINITION REBUILD; 
ALTER INDEX ALL ON CM_NIMBUS_ROBOT REBUILD;
ALTER INDEX ALL ON CM_COMPUTER_SYSTEM_ORIGIN REBUILD; 
ALTER INDEX ALL ON CM_CONFIGURATION_ITEM_ATTRIBUTE REBUILD; 
ALTER INDEX ALL ON CM_RELATIONSHIP_CI_CI REBUILD;
ALTER INDEX ALL ON CM_RELATIONSHIP_CI_CS REBUILD;
ALTER INDEX ALL ON CM_RELATIONSHIP_CS_CI REBUILD;
ALTER INDEX ALL ON CM_DISCOVERY_NETWORK REBUILD; 
ALTER INDEX ALL ON S_QOS_DATA REBUILD;
ALTER INDEX ALL ON NAS_TRANSACTION_SUMMARY REBUILD;
ALTER INDEX ALL ON NAS_TRANSACTION_LOG REBUILD; 
ALTER INDEX ALL ON NAS_ALARMS REBUILD;

- The UIM Administrator needs to have SA or SA-like access to MS SQL Server Management Studio.

- Orphaned S_QOS_DATA entries need to be pruned from the S_QOS_DATA table for better performance as this table is accessed frequently, because the table has grown to over ~4M rows.

- A DAILY defragmentation job should have been set up but this should be checked. The job should be scheduled off hours and should be defragmenting the following table indexes:

- All OLD hub instances completely eradicated from the environment by upgrading to-> the current DX UIM version and/or upgraded to the highest hub GA version available

- Monitoring governance: Disabled processes process monitoring on (failover server) and there was a low frequency monitoring interval 1-2 mins on 1352 robots which was creating a huge amount of QOS which was being stored even for process state up/down which is questionable in terms of usability. Alarming should be sufficient instead of saving the QOS data for process up/down state. data being published to the DB for processes probe process state up/down was quite extensive and flooding hub xxxx. This should be further reduced.