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

Products

• VMware vSphere ESXi

Issue/Introduction

Users may encounter a warning banner in the vSphere Client stating: Insufficient configured resources to satisfy the desired vSphere HA failover level on the cluster <cluster-name> in <datacenter-name>

In some instances, it may appear that resource usage is not being reflected properly, or that the cluster has sufficient resources despite the alert. This issue is often reported after attempting to reconfigure HA or disabling/enabling HA without a change in the status.

Cause

This alert can trigger when the Desired failover capacity (Define host failover capacity by > Cluster resource Percentage in Admission Control) exceeds the Current available failover capacity.

In the logs (vpxd.log), the failure can be identified by comparing the requested reservations against available resources:

• Example Log Entry: Desired: CPU-reservation: 293960, Mem-reservation: 3906338 vs. Current: CPU-reservation: 294372, Mem-reservation: 3890128

The primary cause is typically a misalignment between the Percentage of Cluster Resources Reserved and the actual number of host failures the cluster is intended to tolerate (N+X redundancy). For example, in a 10-host cluster:

• A 50% reservation requires 5 full hosts to be kept empty.
• If the goal is to tolerate 4 host failures (N+4), the setting only needs to be 40%.
• If the workload consumes more than 50% of the resources, the 50% "Desired" mark cannot be met, even if the cluster can safely support 4 host failures.

Resolution

To resolve this alert, align the Admission Control percentage with your specific host failure requirements and current cluster size.

1. Calculate the Required Percentage:
   Use the formula: (Number of Host Failures to Tolerate / Total Number of Hosts) * 100.
   • Example: For 4 host failures in a 10-host cluster: (4 / 10) * 100 = 40%.
2. Adjust Admission Control Settings:
   • Navigate to the Cluster in the vSphere Client.
   • Go to Configure > vSphere Availability > vSphere HA is Turned ON > Edit >  Admission Control > Define host failover capacity by > Cluster resource Percentage > Override calculated failover capacity.
   • Adjust the Reserved failover CPU capacity and Reserved failover Memory capacity to the calculated percentage (e.g., 40%).
3. Monitor Capacity: Once the setting is adjusted, verify that the "Current Failover Capacity" remains above the new threshold. If the banner persists at the lower percentage, it indicates that the VM workload is consuming too many resources to support the desired level of redundancy, and additional hardware or VM rightsizing may be required.

Note: Before adjusting these values, consult with your Architecture team to ensure the new redundancy level (e.g., switching from 50% to 40%) aligns with your internal Service Level Agreements (SLAs).

Although the HA admission control policy was updated to "Cluster Resource Percentage" via the UI, the internal calculation engine of vCenter Server continued to use the previous configuration. 

• Restarting the vpxd service clears the stale in-memory cache and forces the vCenter Server internal calculation engine to successfully apply the newly configured Cluster Resource Percentage policy.

         service-control --restart vpxd

 

Additional Information

Reference: Cluster Resources Percentage Admission Control