vSAN Objects Inaccessible due to Cluster Quorum Loss while Hosts are in Maintenance Mode
search cancel

vSAN Objects Inaccessible due to Cluster Quorum Loss while Hosts are in Maintenance Mode

book

Article ID: 442639

calendar_today

Updated On:

Products

VMware vSAN

Issue/Introduction

  • Multiple Virtual Machines (VMs) on a vSAN datastore show as Inaccessible.
  • The vCenter UI reports vSAN object health errors and potential network partitions.
  • Issues occurred during or immediately after a vSAN encryption reconfiguration or cluster-wide maintenance activity.
  • Attempts to boot VMs may fail or show no underlying storage components.

Environment

  • vSAN 8.x

Cause

This issue occurs when a vSAN cluster loses quorum. In this specific scenario, the cluster-wide inaccessibility was triggered because all (or a majority of) ESXi hosts were placed into Maintenance Mode (MM) simultaneously during a reconfiguration process.

When a host enters Maintenance Mode, its contribution to the cluster quorum is restricted. If the number of active, non-maintenance hosts falls below the required threshold to maintain a majority of votes, vSAN objects become inaccessible to protect data integrity, as the cluster can no longer reach a consensus on object states.

Resolution

To restore object accessibility, the cluster must regain quorum by ensuring a sufficient number of hosts are active and participating in the cluster.

  1. Exit Maintenance Mode: Identify all hosts currently in Maintenance Mode and exit them. Priority should be given to hosts containing critical metadata and data components.
  2. Verify Cluster Membership: Once hosts have exited Maintenance Mode, verify the cluster state from the ESXi CLI using: esxcli vsan cluster get Ensure the 'Local Node State' is Agent or Master and the 'Health' is Healthy.
  3. Monitor Object Health: In the vSphere Client, navigate to Monitor > vSAN > Skyline Health.
    • Trigger a Retest.
    • Check the vSAN Object Health category to ensure objects transition from 'Inaccessible' to 'Healthy' or 'Reduced Availability'.
  4. Wait for Resync: If the reconfiguration involved data movement (such as encryption changes), monitor the Resyncing Objects tab. Accessibility should return as soon as quorum is established, even if background resyncing is still in progress.

Additional Information