Microsoft Exchange
Veeam Backup and Replication supports a variety of Exchange configurations including DAG deployments. For more details refer to the corresponding section of the User Guide.
General
Only process one Exchange cluster within one backup job. This provides better handling for tasks like jobs cancelations or on-demand backups before a cluster maintenance is performed.
Virtualized Exchange Database Availability Groups (DAG)
Preparation
DAG clustered configurations may require adjusting cluster timeouts to avoid failovers during backup as per KB1744.
Job configuration
Only include the minimal number of DAG nodes required to backup all Exchange databases within the cluster to the Veeam backup job. This reduces the impact on production machines as well as the consumed storage, as a reduced number of copies of the same databases are backed up.
Ideally you can create a single passive-only DAG node containing passive copies of all Exchange databases that need backup and only backup this single VM.
For backup and restore of Exchange servers to work properly application aware image processing option has to be enabled in the job properties. For more details refer to the corresponding section of the User Guide.
Granular item restore
When mounting Exchange database Veeam Explorer for Exchange replays relevant log files which may significantly increase time needed for mount operation in case there is a lot of logs to replay. As lagged DAG technology relies on keeping lots of Exchange logs expect Veeam Explorer taking significant amount of time to mount EDBs when performing item restore from lagged DAG mailbox servers.
Log Truncation Scenarios
There are different scenarios for log truncation, depending on the job configuration. The following table provides an overview when log truncation will take place and when not.
Validated setup:
- Exchange DAG Cluster with 3 nodes.
- Each node has 2 volumes with at least one Exchange database each.
- Backups perfomed with Veeam Agent for Windows managed by VBR.
Job configuration | DAG with IP | DAG IP-less |
---|---|---|
Veeam Protection Group (Cluster), entire computer selected | truncated | not supported |
All nodes added to one job, entire computer selected | truncated | truncated |
Single Node, Entire Computer | ||
Only active DBs | truncated | truncated |
Only passive DBs | truncated | truncated |
Active and passive DBs | truncated | truncated |
Single Node, All Volumes | ||
Only active DBs | truncated | truncated |
Only passive DBs | truncated | truncated |
Active and passive DBs | truncated | truncated |
Single Node, 1 of 2 DB volumes selected | ||
Only active DBs on this volume | not truncated | not truncated |
Only passive DBs on this volume | not truncated | not truncated |
Active and passive DBs on this volume | not truncated | not truncated |
As seen in the table above, log truncation will only take place if all volumes holding Exchange DBs are included during the backup job run.
This behaviour can be applied to VM based backups when exclusion of virtual disks is used.
References
- Tips for DAG Exchange Backup and Replication in vSphere
- Backup of Database Availability Groups with Veeam Agent for Windows
- Support for DAG Log Truncation