Troubleshooting

This article presents troubleshooting for the Nutanix-AHV Plugin.

If the proxy VM boots to an initramfs shell after a failed backup/restore it is probably due to the fact that some of the disks that were mounted on the proxy VM are still attached.

To solve this issue the user has to manually remove the supernumerary disks manually through the PRISM console.

Another way to remove the disk is to issue REST API commands. A simple bash script like the following example will do the trick:

Note

This script is pretty aggressive and will remove all disks from index 1 to 10 without asking anything. It should be used carefully and adapted if the initial proxy VM configuration has more than one disk on the SCSI bus.

for i in {1..10}
do

curl -k -X POST --header 'Content-Type: application/json' --header 'Accept: application/json' -d '{
  "uuid": "$proxyVmUuid",
  "vm_disks": [
    {
      "disk_address": {
        "device_bus": "SCSI",
        "device_index": '$i'
      },
      "is_cdrom": false
    }
  ]
}' 'https://$HOST:$PORT/PrismGateway/services/rest/v2.0/vms/$proxyVmUUID/disks/detach' -u '$USER:$PASSWORD'

done

The script will try to unmount all disks from index 1 up to 10. Index 0 is skipped since the proxy VM first disk is the one containing the VM data.