Since upgrading to 2020.2 (2020/08/12) our Server Team has reported spikes in CPU utilization that coincide with Orion Asset Inventory polling.
When triggering a manual poll for Asset Inventory I can identify a 50%+ CPU spike lasting for 5+ minutes.
The issue only presents on Nodes polled by polling engines (2 of them) that reside in a domain separate from the primary polling engine.
The current deployment has been in place for 5 years, which leads me to the obvious question: Was something changed in the way 2020.2 polls for Asset inventory?
