Windows operating systems

Recovering from non-standard error conditions

Problems with iSCSI devices can prevent you from performing an instant access or instant restore operation.

About this task

When an ESX server cannot access a datastore on an iSCSI disk, a VMware message is issued to indicate that a "permanent device loss" error occurred. You should be offered an option to either cancel or retry the iSCSI connection attempt. Choose the option to try the operation again to see whether the error is transient and if recovery is possible. If the retry is not successful, try the following troubleshooting steps. If they are successful, then try the instant restore or instant access operation again.

Procedure

  1. Examine the ESX server Task and Event log for an All Paths Down (APD) error. It can take time for this error to display in the logs, but it must be present before you continue to the next steps. If you do not wait for the error before you attempt more troubleshooting, you might bring the ESX server down.
  2. Power off the virtual machine.
  3. Rescan the HBA. Rescanning the HBA on the ESX server might reactivate the failed device. If VMware kernel locks prevent you from rescanning the HBA, perform the following steps:
    1. In the vCenter interface, select the ESX host.
    2. Click Configuration.
    3. Right click iSCSI Software Adapter and select Properties.
    4. Click Static Discovery.
    5. Delete any static addresses and click Close.
    6. Rescan the HBA.