Starting with vSphere 6.0, in vSphere HA configurations, there is a new storage-related feature, VM Component Protection (VMCP), that protects VMs from possible storage issues.
VMCP can manage two different types of condition:
- PDL: This occurs when the storage array issues an SCSI sense code indicating that the device is unavailable (for example, a failed LUN)
- APD: Usually, related to an underlying storage/networking issue, which is different from a PDL because the host doesn't have enough information to determine if the device loss is temporary or permanent
A typical response could be to restart the VM in the case of a PDL, because this condition may indicate that the storage device does not expect the device to return any time soon. However, you can configure PDL with different responses:
- Disabled: No action will be taken to the affected VMs
- Issue events: No action will be taken against the affected VMs; however, the administrator will be notified when a PDL event has occurred
- Power off and restart VMs: All affected VMs will be terminated on the host, and vSphere HA will attempt to restart the VM on hosts that still have connectivity to the storage device
An APD condition is more of an unknown situation; when an APD occurs, a timer starts. After 140 seconds, the APD is declared, and the device is marked as APD time out. There are different types of responses for APD:
- Disabled: Same as before, no action will be taken against the affected VMs
- Issue events: Same as before, no action will be taken against the affected VMs. However, the administrator will be notified when a PDL event has occurred
- Power off and restart VMs (conservative): vSphere HA will not attempt to restart the affected VMs unless it has determined there is another host that can restart the VMs
- Power off and restart VMs (aggressive): vSphere HA will terminate the affected VMs even if it cannot determine that another host can restart the VMs
Note that there is also a response recovery option to retry before APD times out.
All settings are also available with the vSphere Client (HTML5), as shown in the following screenshot:
For more information, see the following:
- Permanent Device Loss (PDL) and All-Paths-Down (APD) in vSphere 5.x and 6.x at https://kb.vmware.com/kb/2004684
- Lost or degraded connectivity to the storage device at https://kb.vmware.com/kb/1009553