Aria Operations Manager is not providing accurate data when managing capacity mgmt and resizing VMs.
Many of the datacenters are gray in the Optimize > Capacity page
Restarting the Aria Operations cluster resolves the issue temporarily, but it returns within a week or two.
The issue started occurring shortly after upgrading to Aria Operations 8.16.x
Environment
Aria Operations 8.16.x Aria Operations 8.17.x
Cause
The upgrade to Aria Operations 8.16 introduced new functionality to the NSX management pack which causes the internal database usage of BGP Service objects to grow unexpectedly large.
The oversized database objects cause performance issues for the Aria Operations cluster which eventually results in thread locks when persisting the collected metrics for the BGP Service objects.
The thread locks result in the Aria Operations metric queue to overflow.
When the metric queue overflows, metrics that have been collected but not processed are discarded
Delete the oversized BGP Service objects from the inventory list with the Delete object button in the inventory toolbar using this documentation: Inventory: List of Objects