Troubleshooting ESXi host 'Install failed' state due to physical CPU contention in VMware NSX

Products

• VMware NSX

Issue/Introduction

• An ESXi host exhibits an install failed state within the VMware NSX manager UI.
  

• Symptoms preceding this state include ESXi host slowness, accessibility issues and unresponsiveness.

Environment

VMware NSX 

Cause

• Severe physical host resource contention causes the observed install failed state.
• Extreme CPU contention and high CPU ready times prevent the hypervisor from scheduling essential host processes.
• Resource deadlocks cause the NSX installation or host fails to report the correct information during routine health check from NSX manager.

Resolution

 

• It is critical to verify host performance by accessing the vSphere client and navigating to the affected host.
• Examine the performance charts for CPU utilization and CPU ready time.
  
  
• Analyze the performance metrics to identify the timeframe when CPU contention spiked.
• High CPU ready values confirm that virtual machines are waiting for physical CPU scheduling.
• Perform a controlled reboot of the affected ESXi host to clear the resource deadlock.
• Monitor the ESXi host status in the NSX UI following the reboot to ensure the installation status report correctly.
• Examine the host logs to identify the processes or virtual machines that caused the initial CPU contention spike.
• Remediate the underlying cause of the contention to avert recurrence.

 

Additional Information

ESXi hosts showing high CPU Utilization

ESXi host CPU usage alarms and VM performance degradation in VDI environments due to vCPU over provisioning