All VM are in "invalid" state on vSAN

Products

• VMware vSphere ESXi
• VMware vSAN

Issue/Introduction

• VM's are in a invalid state on one or more hosts. 
• vCenter is down 
• esxcli vsan health cluster -w showing network partition
• esxcli vsan cluster get   showing incorrect member count

Cause

With the vCenter down and unable update the cluster membership

Hosts are missing unicastagent entries to from table

• esxcli vsan cluster unicastagent list 
• Unicast table will not see an entry for the host you are preforming the command on; this is normal. 

Resolution

To correct the issue add the missing entries on hosts unable to see the partitioned host. 

1. You will need the following:
   
   
   • The ip of the vmk for vsan traffic of the missing host.
   • The vsan UUID of the missing host. You can get this by run the following command on the missing hosts
     
     
     • esxcli vsan cluster get
       
       
     • This will be the Local Node UUID
       
       
2. After information is collected you will run the following command on the hosts missing entries
   
   
   • esxcli vsan cluster unicastagent add -a <ip of missing host> -u <uuid of missing host> -t node -U true 
     
     
3. Then check to see if the cluster membership is corrected and the network partition is cleared.
   
   • esxcli vsan cluster get
     
     
   • esxcli vsan health cluster -w

Additional Information

• vSAN -- Cluster down. VMs are showing as "invalid". Network partition caused by an invalid/incomplete unicast agent list on vSAN host(s)
• esxcli reference
• esxcli vsan commands