Insufficient configured resources to satisfy the desired vSphere HA failover level on the cluster

Products

• VMware vCenter Server

Issue/Introduction

Symptoms:

• Dedicated host fail over policy in enabled in vSphere HA cluster. 
• The cluster summary displays a message  "Insufficient configured resources to satisfy the desired vSphere HA failover level on the cluster"
• In vpxd.log after enabling verbose level (Configure Logging Options), you see entries similar to: 

2018-09-25T14:20:48.495Z verbose vpxd[7F9A72862700] [Originator@6876 sub=DAS opID=DrsHaConfigWithTOCDialog-apply-1351484-ngc:70073575-2c-04] Failover resources reservation is 64587 cpu and 767809 memory and desired resources for admission control are 64750 cpu and 766785 memory with failover level 1

Note: The preceding log excerpts are only examples. Date, time, and environmental variables may vary depending on your environment.

Environment

VMware vCenter Server 6.5.x

Cause

This issue occurs if the maximum capacity host is not selected for dedicated host fail over policy

Resolution

This issue is resolved in vCenter Server 6.5 Update 1d, available at VMware Downloads

With the fix,

• For Dedicated host policy, vSphere HA will tolerate maximum host capacity by 5% or configured Overhead values before generating Insufficient resource message
• To configure Overhead values, add below advance parameters to vCenter:

vpxd.das.memOverheadInMB
vpxd.das.cpuOverheadInMhz

• The above values can be derived as difference between desired resources for admission control and resources reservation noticed in vpxd.log entry similar to the symptom section of this article
• To add values in web client:

Select “vCenter”, select “vCenter Servers”, select the appropriate vCenter Server and select the “Manage” tab, select “Settings” and “Advanced Settings

 

• These advanced options were rephrased in vsphere65u2 to be as following:

config.vpxd.das.memToleranceInMB
config.vpxd.das.cpuToleranceInMhz

Workaround:
To workaround this issue, use any one of the below options

• Use other admission control policies (slot or percentage based policy)
• Use host with maximum CPU and Memory among all the hosts as dedicated host.