vCenter HA status is Healthy but Standby Nodes show "lwsmd.service failure" errors on Console
Article ID: 447265
Updated On:
Products
Issue/Introduction
In the vSphere Client, the vCenter HA compliance dashboard displays Mode: Enabled, State: Healthy, and all cluster nodes (Active, Passive, Witness) show a green checkmark status of Up.
However, upon opening the Direct Console User Interface (DCUI) or Web Console for the Passive or Witness virtual machines during or shortly after a reboot/reconfiguration, the following error messages are actively displayed on the black boot screen:
Environment
VMware vCenter Server 8.0
Cause
This is caused by a temporary backend synchronization delay between the nodes during the boot-up sequence.
When the secondary virtual machines start up, a brief network or timing lag causes the authentication service (lwsmd) to initially time out and log a failure on the console. However, the system automatically retries and stabilizes in the background, allowing the vCenter HA cluster to successfully return to a Healthy status.
Resolution
Follow the standard VCHA reconstruction procedure to clear any underlying database locks:
Remove VCHA: Navigate to vCenter HA > click REMOVE VCENTER HA. This cleanly deletes the secondary Passive and Witness VMs from your inventory without affecting your live management engine.
Re-run Setup: Click SET UP VCENTER HA to initiate a fresh deployment.
Optimize Resource Placement: During the resource assignment step, ensure you place the Witness or Passive node on an ESXi host and datastore with optimal performance and verified NTP time synchronization.
Allow Sync Time: After deployment, allow 10 to 15 minutes for the backend PostgreSQL database and authentication tokens to finish their initial full handshake. The cluster status will automatically shift to Healthy.
Feedback
