NSX reports the following capacity alarm:
Description: The number of objects defined in the system for System-wide Prefix-lists has reached [###], which is above the minimum capacity threshold of 70.0%. The maximum number of System-wide Prefix-lists supported is [###].
Recommended Action: Navigate to the capacity page under the NSX overview UI for the respective feature (i.e., Networking, Security, Inventory, and System) and review current usage versus threshold limits. If the current usage is expected, consider increasing the minimum threshold values. If the current usage is unexpected, review the network policies configured to decrease usage at or below the minimum threshold.
Note: The maximum number of System-wide Prefix-lists supported may vary based on the size of the deployed NSX Manager node.
Run the following API call to validate the capacity usage:
https://<NSX_Manager_IP>/policy/api/v1/infra/capacity/dashboard/usage?category=networking
From the NSX Manager support bundle, inspect theadaptor_ufo_dumpfile to confirm prefix-list
less adaptor_ufo_dump | sed -n '/PrefixListInternalConfigMsg/,/ComputeCollectionMsg/p' | grep display_name -c
less adaptor_ufo_dump | sed -n '/PrefixListInternalConfigMsg/,/ComputeCollectionMsg/p' \| grep display_name | sort | uniq -c | sort -nr | head -n 20
VMware NSX 4.2.X
The capacity usage accounting is inflated because system-default prefix-lists are automatically assigned to every Tier-0 router and VRF instance. These default lists significantly contribute to the overall count, alongside any user-created prefix-lists.
This issue is resolved in Upcoming NSX versions.
Changes made in upcoming NSX releases: The "System-wide prefix-lists" limit has been removed from the Capacity Dashboard across all NSX Manager sizes.
Workaround :
The system-default prefix-lists cannot be deleted.
To suppress the alarm as a workaround: