7.5 Backup and restore Backup and restore processes in a clustered environment require some special attention. Application data can usually be accessed only from one node exclusively at a time. According to that, backup of the data has to be run from the node where the data is accessible. The backup data on a typical backup server such as IBM Tivoli Storage Manager is bound to a node name or the host name of the data source. Therefore, a backup solution for a cluster service must performed so that correct saved data is located when required. This section provides a conceptual solution based on best practice. In this solution, the cluster service backup data is stored under a common name for that data. This common name is independent from the physical node on which the cluster service is running. Data is backed up under this common name from the node that runs the service actively during backup. With the IBM Tivoli Storage Manager backup and archive client, it is possible to have multiple logical configurations on one physical machine. These configurations are registered in the IBM Tivoli Storage Manager server as nodes. Each configuration can back up different data sets. One data set is created for the physical node's local file systems on internal disks. Further configurations with individual node names are created for each cluster service, for example, one for sesap and one for sedb. Multiple copies of the IBM Tivoli Storage Manager