2 node ESXi cluster fails to automatically migrate VMs when placed into maintenance mode
Article ID: 394680
Updated On:
Products
Issue/Introduction
When attempting to place an ESXi host into Maintenance Mode in a 2-node cluster, the operation hangs (typically at 2% or 19%) and virtual machines do not automatically migrate to the remaining host. This occurs even when VMware Distributed Resource Scheduler (DRS) is set to "Fully Automated."
The following errors appear in Monitor > DRS > Faults:
The operation is not allowed in the current state.vSphere HA admission control is preventing the operation.
In the vpxd.log of vCenter Server, you may observe: info vpxd sub=DAS MinPoweredOnCapacity check failed, requiredCpuCap : ####, requiredMemCap : #### neededCpuCap : 1, neededMemCap : -4
Environment
VMware vCenter Server
Cause
This issue occurs because vSphere HA Admission Control is enabled. Admission Control requires at least three hosts in a cluster to maintain failover redundancy (N+1). In a 2-node environment, placing one host in maintenance mode leaves only one host active, which violates HA failover capacity requirements and causes DRS to block automatic migrations.
Resolution
To allow hosts to enter maintenance mode automatically in a 2-node cluster, you must temporarily disable Admission Control or use specific vSAN evacuation settings.
- Select the Cluster in the vSphere Client inventory.
- Navigate to Configure > vSphere Availability > Edit.
- Select the Admission Control tab.
- Set Define host failover capacity by to Disabled.
- Click OK.
- If using vSAN, ensure you select Ensure Accessibility when prompted for the data evacuation mode. Selecting "Full Data Migration" will fail as there is no third host to receive the data components.
Note: Disabling Admission Control removes the guarantee that VMs can restart in the event of a host failure during the maintenance window. Re-enable the setting once maintenance is complete.
Additional Information
Feedback
