How to configure NetBackup Policies for Clustered clients

Description

This article supplements existing NetBackup documentation for use in High Availability environments. It provides a general guide for configuring NetBackup client policies for clusters to establish a reliable backup strategy. This article covers only the configuration of NetBackup client policies.

Considerations:

Under certain conditions, a Clustered Application Fail-Over may occur during backup due to the extra load caused by simultaneous backups on each host node. Careful planning is required when configuring NetBackup Schedules in a clustered environment to avoid overlapping backup times between host nodes and Virtual Server resources.

Using ALL_LOCAL_DRIVES directive with clusters is not recommended because of issues with Snapshot technologies and Shared cluster disks.

To fully backup the client:

• The NetBackup client should be installed on each individual cluster node.

A. To backup the cluster nodes (for DR of the OS on that node if needed):

1. Create a separate policy for each individual node in the cluster.
2. Specify the actual hostname of each cluster node in the 'Client Name' field on the 'Clients' tab.
3. In the selection list, specify the Local boot disk(s), such as C:\ and D:\ etc., and the Shadow Copy Components:\ directive.
4. Configure the Schedule so that these policies do not run at the same time as the virtual server backups, or the backups of other cluster nodes.

B. To back up the cluster shared data (for user-data recovery in case of data loss):

1. Create an individual policy for each virtual server in the cluster.
2. Specify the Virtual Server hostname in the 'Client Name' field on the 'Clients' tab.
3. Specify the Shared Disk associated with the Virtual Server.
4. Configure the Schedule so that these policies do not run at the same time as the individual cluster node backups or other Virtual Server backups in the cluster.

Note:

1. When backing up individual cluster nodes, the Shadow Copy Components:\ directive includes the System State components and the EFI System partition of that node. There is no need to add the System_State:\ directive separately in the policy selection list.
2. The Shadow Copy Components:\ directive will also back up DFSR paths if DFSR is hosted in a cluster. Usually, DFS data is hosted on the shared disk of the cluster, which you do not want to back up during the backup of individual cluster nodes. Therefore, you should add the top-level DFS directory to the exclude list of the policy as follows:

Shadow Copy Components:\User Data\Distributed File System Replication\DfsrReplicatedFolders