OC Groups view - slow performance issues in cumulative update (CU) 6

Products

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

Issue/Introduction

• Groups view performance issues after upgrading to CU6 from CU5.
• MS SQL Server cpu utilization seems high as well.
• Selecting group view and clicking alarms or maintenance ends up freezing up any browser.

data_engine.log shows
de: COM Error [0x80040e31] IDispatch error #3121 - [Microsoft OLE DB Provider for SQL Server] Query timeout expired

...seems related to the problem.  It makes no difference if you use the old default data_engine provider or the OLEDB provider, both generate the "timeout" message in the log.

When working on a lot of dashboards - the response time is sometimes terrible. 

It's almost the impression that some new code is "locking" the sql database.

Environment

• Release: 20.4 CU6

Cause

• Defect in CU 6

Resolution

1. Installed updated ump_operator_console 20.49 to the Operator Console machine (ump_operatorconsole-20.4.9.zip)
   (can be found here https://support.broadcom.com/web/ecx/solutiondetails?aparNo=99111691&os=MULTI-PLATFORM)
2. Increased SQL Server Max memory
3. Increased maximum memory for the discovery_server
4. data_engine command_timeout set back to 150
   • This can be added to the data_engine <setup> section.
     command_timeout = 300
     This setting helps to avoid timeout issues during data_engine INSERTs to the backend UIM database.
5. Deployed new discovery_server v20.44T1 to the robot that owns the discovery_server probe (it is attached to this article).