Library Control Path Port Failover may be configured with one drive hosting the active logical link to the library controller and a second drive configured as the passive logical link to the library controller. The library connections will share physical links with the drives but the library will have its own identity on the SAN. If data path port failover is also configured the library control path will follow the data path on link failures.
The HPE StoreEver ESL G3 tape libraries provide an advanced library control path failover capability. The library will monitor the drive that is hosting the library controller and the ports on the drive. The library receives notification from the drive of any interface change events and can quickly detect whether that change indicates a fault. The library monitors the drive and can detect when a drive has been removed or otherwise becomes inaccessible. Upon detecting a fault that would cause loss of communication with the library controller the library will automatically move the library control path identity (worldwide names) and all settings (such as reservations and prevent/allow settings) over to the alternate control path drive and activate the control path on the alternate drive. When properly configured, the change is minimally disruptive to the host and does not require any configuration changes on the host or in the backup application.
The HPE StoreEver MSL tape libraries provide a basic control path failover capability. In the event of a failure of the drive hosting the library control path the library web interface may be used to take the control path drive offline and the library control path identity (worldwide names) and all settings, (such as reservations and prevent/allow settings), will be moved over to the alternate master drive automatically. When properly configured the change is minimally disruptive to the host and does not require any configuration changes on the host or in the backup application.
If no library commands are sent while the port is being reconfigured, the port change happens with virtually no disruption to the SAN. Commands sent while the port is being reconfigured cannot be processed and will fail. Applications that retry commands are able to use the library following reconfiguration of the port. Applications that do not retry can be restarted remotely without making any hardware configuration changes.