vSAN datastore not showing full capacity after using 'Shutdown Cluster' option.

Products

• VMware vSAN 8.x

Issue/Introduction

Symptoms:

• 'Shutdown Cluster' was option used to power-cycle the cluster:
  
  
  
  
• All vSAN nodes show up as Master node creating a cluster per node.
  Run command : 'esxcli vsan cluster get' 
  
  
• Total datastore capacity is equivalent to formatted disks of one single host.

 

Environment

VMware vSAN 8.x

Cause

• Executing 'Shutdown Cluster' results in following advanced setting being set to True:
  /VSAN/IgnoreClusterMemberListUpdates
  
  
• Enablement of this setting results in vSAN nodes ignoring the cluster member list updates and recognize each of the host as a Master node.
  
  
• Resulting effect is of all vSAN nodes partitioned after power-cycling.
  
  

Resolution

1. Set vSAN Advanced setting /VSAN/IgnoreClusterMemberListupdates to false for all vSAN nodes by running command:
   # esxcfg-advcfg -s 0 /VSAN/IgnoreClusterMemberListupdates 
   
   
2. Create a new vSAN cluster with vSAN ESA enabled
   Right click on datacenter > 'New Cluster' > Toggle the radio button 'against vSAN and check the box 'Enable vSAN ESA'.
   
   
   
   
3. Add all the hosts to the new cluster by 'drag and drop'
   (Alternatively, right click on the cluster > 'Add host' > Select all the hosts)
   
   
4. Click on vSAN Cluster > Go to 'Configure' tab > Disk Management.
   Select each listed host > Click on 'Claim disks' .           
   
   
5. Allow disk claim to complete for all vSAN nodes. 
   
   
6. Validate the total capacity of the vSAN datastore to be equivalent to summation of the individual formatted disks of all the nodes.