DRS will not let VMs run on host in a vSAN cluster with vSAN File Services Enabled
Article ID: 377975
Updated On:
Products
Issue/Introduction
When on a vSAN cluster with vSAN File Services enabled, DRS will not migrate VMs to certain hosts. When VMs are migrated manually, DRS will immediately move them back.
In Skyline Health, there will be alarms for vSAN File Services that certain nodes don't have File Services enabled.
In eam.log there can be entries similar to the following:
2024-09-12T15:56:53.762Z | WARN | vlsi | Workflow.java | 156 | [OvfValidator->Validate:http://localhost:1080/external-tp/http1/napa1-<fqdn>/443/1f22b33fede62411a179f316d455502fd30d6c4c/vsanHealth/fileService/ovf/7.0.3.1000-20036589/VMware-vSAN-File-Services-Appliance-7.0.3.1000-20036589_OVF10.ovf:56917d1ebd1a7d5c] NEXT WORK ITEM : Failed to instantiate com.vmware.eam.exception.CannotAccessOVF: Cannot access OVF at http://localhost:1080/external-tp/http1/<fqdn> /443/1f22b33fede62411a179f316d455502fd30d6c4c/vsanHealth/fileService/ovf/7.0.3.1000-20036589/VMware-vSAN-File-Services-Appliance-7.0.3.1000-20036589_OVF10.ovf
Environment
VMware vSAN 7.0 +
Cause
A per the following documentation: Developing and Deploying vSphere Solutions, vServices, and ESX Agents
If a node is not in a 'ready' state DRS will not consider it eligible to run VMs. A node reaches 'ready' state when all ESX agent virtual machines have been provisioned and powered on. This includes vSAN File Services containers.
"ESX agent virtual machines are similar to services in Windows or Linux. They start when the
operating system starts and they stop when it shuts down. The behavior of ESX agent virtual
machines is transparent to the user. A vSphere host reaches the ready state when the ESXi
operating system has started and all ESX agent virtual machines have been provisioned and
powered on."
Resolution
Resolve underlying File Services Issue.
In most cases this occurs when hosts are added to a vSAN File Services enabled cluster after the containers have already been deployed. vSAN FS needs to deploy the containers all at the same time. To resolve this, disable vSAN File Services, and then re-enable it.
(Note: Disabling and Re-enabling vSAN File Services does not delete any data from the shares, nor will it remove the configurations. It will just re-deploy the container VMs so that each host has one so the host can be marked as 'ready' as per DRS. However, access to the shares will be interrupted while the containers are offline, so you may want to consider a maintenance window or non peak hours in that event.)
If there is no internet connectivity and need to upload the OVF FIles, please see the following documentation: Configure File Services (See 'Manual Approach' Method.)
OVF Downloads can be found in the Broadcom support portal: VMware vSphere downloads, VMware Converter, OEM custom images, patches and addons in the Broadcom Support Portal
Additional Information
Feedback
