Backup Exec supports various cluster configurations of between two and eight nodes on a fibre channel SAN, with locally attached storage devices, or with storage devices on a shared SCSI bus. You can use any combination of these configurations.
If you are using a cluster on a fibre channel SAN or with storage devices on a shared SCSI bus and failover occurs, depending on the capability of your various SAN components, media might be orphaned in the tape drive until the failed node becomes active again.
If end-of-job markers were not written to the media before the failover occurred, the media may be marked as unappendable by the Backup Exec engine when the next append backup job is run. The media remains unappendable until it is overwritten (or erased, or the retention period expires, etc.).
If the storage device is a robotic library, you can review the Robotic Library Inventory report to discover if the media was marked unappendable by the Backup Exec engine. If the Full column reports a 3, the Backup Exec engine has marked the media as unappendable.
To add or remove hot-swappable devices in a cluster, run the Hot-swap Device Wizard on all Backup Exec Cluster nodes. If a server is not updated to recognize a new device, any job that is targeted to that device may fail.
Examples of various cluster configurations are available.