Insufficient Resources Fault when using Mobility Pre-Checks for Passthrough Virtual Machines

Products

• VMware vCenter Server

Issue/Introduction

Provide guidance to configure the infrastructure for Stun Time Limit when using Mobility Pre-Checks.

Symptoms:

Failed to find any candidate host for Virtual Machine(VM) configuration, estimated stun time ... second(s) should be less than ... seconds(s) of stun time tolerated.

Cause

When using Mobility Pre-Checks (DRS Cluster Advanced Option PassthroughRequireDrs = 1), the Stun Time Limit (default 100 seconds) must be greater than the Estimated Stun Time. Host evaluation order or Cluster health may change over time, causing unexpectedly high Estimated Stun Times for a VM configuration. The fault reported may reflect the last host evaluated, and that host may have an unhealthy vMotion network. This results in an Estimated Stun Time that is not representative of the rest of the Cluster.

Resolution

Currently there is no resolution to the issue.

Workaround:
The Estimated Max Stun Time shown in the VM configuration dialog is from the perspective of the currently assigned Host. It is recommended for all Hosts in a Cluster to have similarly configured host vMotion network bandwidth. With a uniform vMotion network, set the Stun Time Limit to accommodate the Workload's vMotion Downtime toleration. If the Workload's vMotion Downtime toleration is less than the Estimated Max Stun Time for the VM/Host configuration, upgrade your Host's vMotion network configuration OR remove the VM devices that are contributing to the Stun Time.