Persistence table usage displays incorrect values on UI

Products

• VMware Avi Load Balancer

Issue/Introduction

• Persistence Table Usage graph reaches 100% utilization when the number of client session increases. This value doesn't come down even if the client sessions are timed out or no new sessions are not initiated by the clients.
• The persistent table usage graph under the VS analytics on the UI is pinned at 100%.

Environment

• Avi Version 22.1.x
• Avi Version less than 30.2.4.
• All VSs with persistence enabled and Extra Shared Config Memory configured.
  
  • https://techdocs.broadcom.com/us/en/vmware-security-load-balancing/avi-load-balancer/avi-load-balancer/30-2/load-balancing-overview/service-engine-group/extra-shared-memory/configuring-extra-shared-memory.html

Cause

• There is currently an issue with how we calculate the persistence table usage.
• The current calculation derives the usage from shm_conn_memory_mb which is carved out at the SE bootup time and does not include the extra_shared_config_memory configured on the SE Group settings.
• Functionally, the persistence entries can grow beyond the shm_conn_memory_mb value, if there is more memory available i.e the extra_shared_config_memory.
• However, the current calculation does not account for that.

Resolution

• If there is Extra Shared Config Memory configured in the SE Group settings, there is no functional impact to the operation of the VS. 
• Only the UI displays incorrect values.
• This issue has been resolved in v30.2.4 which is the next maintenance release.
• On v22.1.x, if the analytics graph is pinned at 100%, please ensure that you have extra_shared_config_memory enabled as per the article linked above.