"Host faults in EMM dry run" error in ESXi precheck

Products

• VMware SDDC Manager / VCF Installer

Issue/Introduction

• "Host faults in EMM dry run" error is reported against a specific ESXi host during precheck:
  
  Description: Checks for host faults while performing EMM dry run for host <hostname>.
Error Description: A host faults were found while performing a dry run enter in maintenance mode validation.
Impact: High: Do not perform upgrade without addressing this issue.
Remediation: There are insufficient memory resources on standby hosts (if any) to meet the requirements of a given operation.  
  
  
• HA admission control has been disabled following vCenter Upgrade fails on SDDC Manager with error "Error : Error during enter MAINTENANCE check due to InsufficientResourcesFault" but issue persists.
  
  
• Below logs are observed in vCenter /var/log/vmware/vpxd/vpxd.log:
  
  MinPoweredOnCapacity check failed, requiredCpuCap : <requiredcpucap>, requiredMemCap : <requiredmemcap>, neededCpuCap :  <neededcpucap>, neededMemCap : <neededmemcap>


• The mentioned ESXi host can be put into maintenance mode manually from vCenter with no issue.

Environment

SDDC manager 9.X

Cause

This EMM dry run failure usually occurs on heterogeneous cluster, i.e. one ESXi host is having different hardware spec than the rest of the ESXi hosts. 

In this particular case, the cpu and memory resources of the mentioned host is larger than the aggregate of the rest of the hosts in the cluster causing failure in DRS EMM calculation even after disabling admission control. As a result, DRS determines the ESXi host cannot be put into maintenance mode.

Resolution

As long as the host can be manually put into maintenance mode from vCenter, the cluster can be upgraded following below steps as a workaround:

Note: Please don't proceed if the ESXi host cannot be manually put into maintenance mode from vCenter.

1. Initiate upgrade from SDDC manager, SDDC manager would be able to put the rest of the hosts into maintenance and upgrade one by one.
2. Once the upgrade workflow fails for the mentioned host due to EMM failure, manually put the host into maintenance mode from vCenter.
3. Reinitiate the upgrade from SDDC manager, the remediation of the mentioned host should start.