Dr protected io services failover troubleshooting – HP Matrix Operating Environment Software User Manual

Page 65

Advertising
background image

DR Protected IO services failover troubleshooting

In addition to the failover issues addressed in this User Guide that are common to both logical
servers and IO services, the following failover issues apply to IO services only:

Failed to activate IO service in a Recovery Group

Possible Causes:

Storage resources are not available.

The IO service is in an invalid state for activation.

The IO service does not exist.

The Matrix infrastructure orchestration Windows service is not running.

There is a communication failure with IO.

Failed to deactivate IO service in a Recovery Group

Possible Causes:

The Matrix infrastructure orchestration Windows service is not running.

The IO service does not exist.

The IO service is in an invalid state for deactivation.

There is a communication failure with IO.

Matrix recovery management error messages specific to IO services

Failed to get the IO service.

Error message

Matrix recovery management was not able to get the IO service information from IO.

Cause

Check the Matrix recovery management and IO log files for more details on the failure. Restart the
Matrix infrastructure orchestration Windows service and retry the operation.

Action

Service activation/deactivation cannot be performed at this time since

the service is in IN_PROGRESS state.

Error message

The IO service was being operated on when the Matrix recovery management activation or
deactivation operation was performed (modification, activation, or deactivation was in progress).

Cause

Wait for the IO service to exit the IN_PROGRESS state and retry the activation or deactivation
operation.

Action

Hyper-V based IO service or logical server fails activation

Symptom: Hyper-V based Matrix service or logical server fails activation with the following error
in Matrix infrastructure orchestration or Matrix OE visualization jobs screen:

Unable to bring resource online. (Virtual Machine Management error code: -1)

Possible cause: The Device Manager in the Hyper-V cluster node lists the disk devices previously
failed over to the remote site as pseudo devices. This can cause Windows disk management not
to see the devices that have been made available as read-write as a result of storage failover. One
of the reasons for this symptom may be the way multi-path I/O (MPIO) re-presents the same LUN
after a failover and failback.

Suggested Actions: The following are two workarounds:

Perform a reboot of the cluster nodes that are failing to see the disks, even after the storage
replication group has been failed over to the local site (as part of Matrix RM failover) to fix

Hyper-V based IO service or logical server fails activation

65

Advertising