event : vsan detected an unrecoverable media or checksum error in a component of the disk group.

[cai123]

 I have located the physical disk where the component is located through the RVC script and identified which virtual machine it belongs to. What do I need to do now?  Can removing the physical disk solve this?

[TheBobkin]

Hello cai123,

 

"Can removing the physical disk solve this?"

 

Unfortunately this depends on whether bad data-blocks were propagated to other data-components or not - typical symptoms of this would be VMs that are failing to take snapshot-based backups (or manual snapshots), failure to clone or SvMotion specific VMs/vmdks etc. .

If no VMs have those symptoms then likely okay here and can just replace the disk that had medium errors, if you do have symptoms mentioned then would be advisable to open a case with VMware by Broadcom Global Support to properly assess the situation and advise accordingly.

[Octopus]

1. Attempt Object Repair (If Data Still Exists Elsewhere)
If the object is still available via its mirrored/redundant copy: You can trigger a manual repair: vsan.repair_object -force <object_path>

2. Remove the Failed Disk (If It’s Confirmed Bad)
If the disk has SMART failures, read errors, or can’t be recovered:
Place the host into maintenance mode
Use “Ensure Accessibility” if you don’t have enough free space
Or “Full data migration” if you want full rebuild (requires space)
Remove the disk group using vCenter or ESXCLI: esxcli vsan storage diskgroup remove -d <disk-uuid>

Then you can physically replace the disk and recreate the disk group.
Do not just pull the disk from the server without properly removing it from vSAN — it can destabilize the cluster.