How the NetWare client works in a cluster

In a cluster environment, the NetWare Client is installed on each node in the cluster. For installation requirements and procedures, see Installing the NetWare client.

IP protocol must be enabled and a route must exist between the NetWare Client and NetBackup master and media servers; the master and media servers, however, do not have to be on the same subnet as the NetWare Client.

Cluster volumes are mounted on a single physical server. When a node failover occurs, the cluster volume automatically remounts on another node in the cluster.

The NetWare Client is server volume-centric, meaning these clients only display volumes that are mounted on, or “owned by”, a server. When you create a backup, whether it be a scheduled backup or a user-directed backup, you can only back up volumes that are owned by the server.

To make the NetWare Client work in a cluster environment you must either create targets or allow scheduled access for all virtual nodes on all physical servers. You should also understand the following limitations:

Backup completion status could be Successful for a volume if the cluster volume was mounted. Or, it could appear as a 1 or a 71, depending on whether you are using single streaming or multi-streaming for backups. You must, therefore, check logs to ensure only the expected volumes were skipped (see Understanding status limitation).

The first time a virtual node fails over, its associated volume will be backed up in full, which could lead to greater media usage.

NetBackup does not track what virtual node controls each shared volume. Therefore, you and the NetBackup system administrator must be able to track what physical node controlled what virtual node, or restores will become a lengthy process.


More Information