Error: vSAN Skyline Health reports Inaccessible Objects due to stale entries

Products

• VMware vSAN

Issue/Introduction

vSAN Skyline Health reports "Inaccessible" object alerts even when all Virtual Machines are healthy. This occurs when orphaned objects remain in the vSAN CMMDS database from historical hardware failures or network partitions. In some cases, these objects cause the host to hang when entering maintenance mode or may cause objtool deletions to deadlock.

Symptoms:

• Skyline Health reports: Inaccessible
• Objects show as 'unknown' type in Skyline Health GUI.
• Manual deletion via objtool delete hangs indefinitely at Issuing delete ioctl.
• Host decommission or Maintenance Mode tasks fail or time out.

Environment

VMware vSAN (All versions)

Cause

Orphaned entries exist in the CMMDS database where the configuration cleanup task failed during historical events. This creates a metadata mismatch between the CMMDS control plane and the DOM storage layer, resulting in kernel deadlocks during standard management operations.

Resolution

1. Identify inaccessible objects and check for friendly names:

   esxcli vsan debug object list --health=inaccessible
   

   Review the output for the target Object UUID.

2. Determine the query path based on the object list output:

   • If the friendly name IS NOT displayed: Skip standard DOM resolution. Identify an active component UUID and its owning host from the object layout output in Step 1. SSH into that host and query the metadata directly at the component layer:

     /usr/lib/vmware/osfs/bin/objtool getAttr -c <Component_UUID> --bypassDOM
     

3. Attempt to purge the stale object:

   /usr/lib/vmware/osfs/bin/objtool delete -u <Object_UUID> -f

4. Escalate if unresolved: If the object remains inaccessible or IOCTL calls deadlock, contact Broadcom Support for guidance on manual CMMDS metadata truncation. See Contact Broadcom Support.