[VMC on AWS] High CPU and Memory usage for Large NSX Edges within the SDDC

Products

• VMware Cloud on AWS

Issue/Introduction

High NSX Edge resource utilization is reported in vCenter when Large Edge appliances are deployed.

Alert: "Virtual machine memory usage"


Edge CPUs are pinned at 100% usage.

The hosts running the Edge appliances show as being oversubscribed.

Cause

The CPU and memory are fully reserved. vCenter is unable to determine the actual CPU usage of the edges which results in 100% usage being reported.

Resolution

This is expected when Large appliances are deployed within the SDDC.

Additional Information

The VMC team is committed to proactive system health management via a robust monitoring framework. This includes the oversight of CPU and memory usage, alongside critical alarms from NSX appliances. Alerting mechanisms are in place for resource consumption exceeding a 90% threshold and for any critical alarm events, ensuring system stability. In instances where CPU consumption surpasses defined limits, the VMC team intervenes to maintain optimal performance, and customers will be notified of such events through the Notification Gateway (NGW).