Overall system memory utilization going high on Edge with NAPP usage

Products

• VMware NSX

Issue/Introduction

• The overall system memory utilization gradually increases on edge nodes after upgrading to version 4.1 with NAPP.
• The below alarm will be observed on the NSX manager:
  
  2024-09-13T06:42:06.949Z nsxmgr NSX 84830 MONITORING [nsx@6876 alarmId="xxxxx-xxx-xxxx-xxxx-xxxxxxx" alarmState="OPEN" comp="nsx-manager" entId="xxxxx-xxxx-xxxx-xxxx-xxxxxxxxx" errorCode="MP701099" eventFeatureName="edge_health" eventSev="CRITICAL" eventState="On" eventType="edge_memory_usage_very_high" level="FATAL" nodeId="xxxxx-xxx-xxxx-xxxx-xxxxxxx" subcomp="monitoring"] The memory usage on Edge node xxxxx-xxx-xxxx-xxxx-xxxxxxx has reached 90% which is at or above the very high threshold value of 90%.


• The below log will be observed in syslog on the edge node:
  
  2024-09-13T07:59:36.520Z nsxedge NSX 3797 - [nsx@6876 comp="nsx-edge" subcomp="node-mgmt" username="root" level="CRITICAL" eventFeatureName="edge_health" eventType="edge_memory_usage_very_high" eventSev="critical" eventState="On" entId="xxxxx-xxx-xxxx-xxxx-xxxxxxx"] The memory usage on Edge node xxxxx-xxx-xxxx-xxxx-xxxxxxx has reached 90% which is at or above the very high threshold value of 90%.


• In the top-mem.log on the edge node, we will see the datapathd process as the top process, but the datapath mempool utilization would be normal.

Environment

VMware NSX

Cause

This is a known issue with NAPP usage in the NSX-T 4.1 version.
When NAPP is onboarded, requests made from NAPP to the NSX Edge can cause a memory leak in the Edge datapathd process.

Resolution

The issue is fixed on NSX-T 4.1.2.5 version.
https://docs.vmware.com/en/VMware-NSX/4.1.2.5/rn/vmware-nsx-4125-release-notes/index.html