When an instant restore operation fails after the VM is
powered on, manual cleanup and repair tasks are required.
An instant restore operation that fails with storage vMotion
running creates either of the following situations:
- The instant restore operation generates an error message.
- The instant restore operation suspends indefinitely and the VM
is not responsive.
To determine the cause of the problem, perform a detailed query
of the VM by using the following command:
dsmc q vm * -vmrestoretype=instantrestore -detail
In
the output that is produced by this command, for each VM in the output,
look for the line that contains
Action Needed.
Use the following
Action Needed paragraphs to recover
from failed instant restore operation, depending on the
Action
Needed status.
Action Needed: Cleanup
In the output of
the query vm * -vmrestoretype=instantrestore -detail command,
verify that the storage vMotion status is successful (vMotion
Status: Successful) and that all VM disks are physical
disks (Disk Type: Physical). This status
confirms that the VM was restored and cleanup of orphaned components,
such as iSCSI mounts, is needed.
This type of failure occurs
as a result of either of the following situations:
- The instant restore failed and Storage vMotion is running. VMware
vSphere continues the vMotion process.
- Storage vMotion finished successfully, but the automatic cleanup
of the iSCSI mounts fails.
To clean up any orphaned components, run the
restore
vm command with the
-VMRESToretype=VMCLeanup parameter.
For example:
dsmc restore vm original_vmname -vmname=new_vm_name -VMRESToretype=VMCLeanup
Action Needed: Repair
In the output of the query
vm * -vmrestoretype=instantrestore -detail command, verify
that the iSCSI device that is attached to the VM is dead (status is Disk
Path: Dead).
This type of failure occurs as a
result of one of the following three situations:
- The VM that is used as a data mover or the physical data mover
machine failed.
- A network failure occurred between the data mover and the ESX
host or the data mover and the Tivoli® Storage
Manager server.
- The Data Protection for VMware Recovery Agent Service failed.
The iSCSI device must be returned to an active state before any
other instant operation is attempted.
To attempt
to recover from a data mover failure, complete the following steps:
- Investigate that cause of the failure and restart the data mover
machine if it does not start automatically. This action starts an
automatic recovery of the mounted iSCSI disks.
- In the output of the query vm * -vmrestoretype=instantrestore
-detail command, verify that the VM disks are active (Disk
Path: Active). This status means that the VM was restored
and is available for use.
- Restart storage vMotion in the vSphere client and monitor its
progress in the vSphere client status bar.
- If storage vMotion processing completed successfully, run the restore
vm command with the -vmrestoretype=VMCLeanup parameter
to clean up the iSCSI disks. For example:
dsmc restore vm original_vmname -vmname=new_vm_name -VMRESToretype=VMCLeanup
To attempt recovery after a network
failure, complete the following steps:
- Repair the network issue so that communication between the data
mover and the ESX host, and the data mover and the Tivoli Storage
Manager server resumes.
- In the output of the query vm * -vmrestoretype=instantrestore
-detail command, verify that the VM disks are active (Disk
Path: Active). This status means that the VM was restored
and is available for use.
- If the network failure did not cause storage vMotion to time out,
no action is required.
- If the network failure caused storage vMotion to time out, and
the error message indicates that the source disk is not responding,
restart storage vMotion in the vSphere client. When storage vMotion
processing completes, run the restore vm command
with the -vmrestoretype=VMCLeanup parameter to
clean up the iSCSI disks. For example:
dsmc restore vm original_vmname -vmname=new_vm_name -VMRESToretype=VMCLeanup
To attempt recovery after a Data
Protection for VMware Recovery Agent service failure, complete the
following steps:
- Investigate that cause of the failure and restart the Data Protection
for VMware Recovery Agent service if it does not start automatically.
This action starts an automatic recovery of the mounted iSCSI disks.
- In the output of the query vm * -vmrestoretype=instantrestore
-detail command, verify that the VM disks are active (Disk
Path: Active). This status means that the VM was restored
and is available for use.
- If the Data Protection for VMware Recovery Agent service failure
did not cause storage vMotion to time out, no action is required.
- If the Data Protection for VMware Recovery Agent service failure
caused storage vMotion to time out, and the error message indicates
that the source disk as not responding, restart storage vMotion in
the vSphere client. When storage vMotion processing completes, run
the restore vm command with the -vmrestoretype=VMCLeanup parameter
to clean up the iSCSI disks. For example:
dsmc restore vm original_vmname -vmname=new_vm_name -VMRESToretype=VMCLeanup
Full cleanup
If you are not able to recover
from a failure and want to remove the VM and its components, run the
restore
vm with the
-vmrestoretype=VMFULLCLeanup parameter.
For example:
dsmc restore vm original_vmname -vmname=new_vm_name -VMRESToretype=VMFULLCLeanup
A
VMFULLCLeanup operation
forces removal of the VM and all of its components, regardless of
the state of the virtual machine. Do not start a full clean up operation
while vMotion is still migrating a virtual machine.