PRTG 8.2.1.1766 - Windows 2003 server/2 CPU/4 GIG RAM - recently added about 4,000 snmp sensors (with auto-discovery) equally about 8,200 total sensors. CPU utilization climbed and stayed elevated, triggering PRTG CPU alerts. PRTGserver.exe was at 1.2 GIG memory usage and CPU was between 75 - 99 % for over an hour, even though most new sensors were "green". Had to delete new sensors and stop/start the server to resume "normal" CPU usage. Are there limit thresholds or best practice procedures for auto-discovery. We have MANY more sensors to add and don't want to crash the server.
Excessive CPU after adding sensors
Modified on 2025-06-10 12:06:15 +0200
Attention: This article is a record of a conversation with the Paessler support team. The information in this conversation is not updated to preserve the historical record. As a result, some of the information or recommendations in this conversation might be out of date.
Disclaimer:
The information in the Paessler Knowledge Base comes without warranty of any kind. Use at your own risk. Before applying any instructions please exercise proper system administrator housekeeping. You must make sure that a proper backup of all your data is available.
The information in the Paessler Knowledge Base comes without warranty of any kind. Use at your own risk. Before applying any instructions please exercise proper system administrator housekeeping. You must make sure that a proper backup of all your data is available.
Hello,
the Auto-Discovery can be quite stressing, especially if it finds a lot of SNMP Traffic Sensors. We recommend not to use daily (or even hourly) auto-discoveries in such scenarios. I'm afraid apart from this, there aren't many options (if the sensors have to be added, they have to be added). Maybe you could also consider an upgrade to a 4-core CPU, that might also help.
Best Regards.
Nov, 2010 - Permalink