KBI 311092 High CPU And/Or Skipped Tasks In Argent Main Engines
Version
Argent Advanced Technology – All versions
Date
Tuesday, 21 Oct 2014
Summary
Environments with a large percentage of remote servers being monitored may experience high CPU utilization and a lot of skipped messages in Argent Main Engine trace logs which degrades performance tremendously
Technical Background
Argent performance monitoring utilizes the standard Microsoft PDH (Performance Data Helper) to monitor performance objects
However, when a server is located in a distant location (more than 6 hops down), it may take up to 10 minutes to successfully extract performance data objects
This in turn clogs the “Shared Monitoring Engine Process” if that’s the mode of monitoring selected on the Relator
This in turns generates high CPU utilization and tasks to be skipped
Resolution
The best solution is to create Daughter Engine(s) or Trusted Agent(s)
If this is not possible, then divide the Monitoring Group into two – one Monitoring Group for the remote servers and one Monitoring Group for the local servers
The Relator monitoring performance objects on local servers can continue to utilize the “Shared Monitor Engine Process In Pool” option
The Relator for Remote servers should have “Spawn New Monitor Engine Process” option
Please review performance troubleshooting KBI: