Challenge
Starting from version 9.5 Update 3, Veeam Backup & Replication lets you deploy and manage Veeam Agent for Microsoft Windows on computers in your infrastructure. Veeam Agent for Microsoft Windows 2.1 adds full support for mission-critical Microsoft Failover Clusters, SQL Server-based Microsoft Failover Clusters, SQL Always On Availability Groups, and Exchange Database Availability Groups.This is a “how to” step-by-step guide on backing up your Windows Failover Cluster.
Cause
Before you beginСonsider the following:
- Backup of failover clusters is supported in Veeam Agent for Microsoft Windows managed by Veeam Backup & Replication only. You cannot process a failover cluster by Veeam Agent for Microsoft Windows operating in the standalone mode.
- Agent license with a server counter must be installed in Veeam Backup & Replication to enable Agent Management features including failover cluster support.
- Veeam Agent for Microsoft Windows supports Windows Server Failover Clusters running Microsoft Windows Server 2008 R2 and later.
- Workgroup clusters, multi-domain clusters, Windows Server 1709 core edition clusters, and mixed OS version clusters are not supported.
- Cluster Shared Volumes (CSV) are skipped automatically during the backup.
- The following Microsoft Failover Cluster applications are supported: Microsoft SQL Server Failover Cluster Instances (Microsoft SQL Server 2008 SP4 or newer), Microsoft SQL AlwaysOn Availability Groups, Microsoft Exchange Database Availability Groups (Microsoft Exchange Server 2010 or newer).
- Microsoft Exchange Database Availability Groups should be protected by independent managed by backup server jobs with server type.
- When using Basic Availability Groups on Microsoft SQL Server 2016 Standard Edition, consider that secondary replica node cannot be backed up because the secondary node is not readable. See the following article for more information.
- AlwaysOn Availability Groups based on multiple Failover Cluster Instances are not supported.
- AlwaysOn Clusterless Availability Groups are not supported.
- Managed by backup server job with failover cluster type does not support the file-level backup mode.
- Recovery must be performed via Veeam Backup & Replication console.
Solution
Microsoft Windows Server Failover Clusters, Microsoft SQL Server Failover Cluster Instances, SQL Always On Availability Groups
1. Create an Active Directory protection group
Protection Group is a container or folder to organize hosts you are willing to protect
To create a protection group navigate to Inventory, select Physical & Cloud Infrastructure node and hit Add Group button at the ribbon. Select Microsoft Active Directly objects as a type for this Protection Group.
At the Active Directory step specify the domain name and account (if required) and select the Active Directory cluster name object.
The AD cluster name account can be found in Active Directory Users and Computers or Failover Cluster Manager.
Make sure that you didn’t exclude any required host at the Exclusion step of the wizard.
Specify the common master account and set custom credentials for particular hosts if needed.
Check the box to Install backup agent at the Options step of the wizard to install backup agents automatically during the rescan (installation can be performed manually if required).
Rescan the protection group to discover the newly added cluster account, all its nodes and install backup agent to every node. Once the rescan is completed you should see the cluster and all child nodes under the protection group
NOTE To avoid installation issues, make sure that all required network and DNS requirements are fulfilled.
2. Configure the backup job
Once you install the backup agent on every node, you can create a cluster job. Navigate to Home node, click the Backup Job button at the ribbon and select Backup > Windows computer.
Failover Clusters must be processed by backup server jobs with the Failover cluster type.
At the Computers step of the wizard hit Add and select cluster account or the parent protection group for this cluster.
NOTE: each cluster node will consume a Server License host counter:
Select the backup mode. Only volume-level backup and entire system backup modes are available for the failover cluster job.
Define the appropriate application-aware processing options at the Guest Processing step of the wizard. For example, SQL log backup/truncation settings would be applicable to a SQL cluster.
Start the job. The cluster job will track node changes and perform log backup respectively.
Microsoft Exchange Database Availability Group
Microsoft Exchange Data Availability Group does not support concurrent VSS snapshots at multiple DAG nodes. As a result, cluster-aware backup job logic is not applicable for this application. This is why Microsoft Exchange Database Availability Groups should managed by backup server jobs with the server type regardless of the DAG configuration (AD-detached IP-less DAG or a regular DAG with a cluster administrative access point).1. Create a protection group
Exchange protection group should not contain an Active Directory cluster name object but it should contain individual Exchange Active Directory computer accounts.
You are welcome to use other protection group types to add Exchange nodes as individual computers or as computers from a CSV file.
Further discovery and installation steps are similar to the configuration steps described above for the standard Windows Server Failover Cluster.
2. Job configuration
In contrast to a standard Windows Server Failover Cluster, Exchange Database Availability Groups should be processed node by node via separate server jobs managed by server.
NOTE If you have an Exchange node which contains only passive database copies, then it’s sufficient to perform the backup of this node only:
Application-aware processing should be enabled for a consistent backup, log truncation and application-item restore possibility.
Specify the backup schedule in a way that these backup jobs start at different times and do not overlap, e.g. you may schedule the jobs one by one.