Best Practices for Backing up and Restoring Database Availability Group

Best Practices for Backing up and Restoring Database Availability Group

Database Availability Group (DAG) setup in Exchange Server provides high availability and site resilience with automatic database-level recovery in case of database, server, or network failure. However, it isn’t an alternative or substitute to backups. You still need to back up your mailbox database and other essential data from the DAG member servers and store them on at least two different storage devices at two geographically distant locations, preferably in another branch or office.

But how do you backup Database Availability Group and what are the best practices (or guidelines) to backup and restore DAG? In this article, we have answered these questions with examples and illustrations.?

Why do you need to back up the DAG?

Database Availability Group or member servers of DAG may degrade, fail, or not boot due to many reasons, such as:

  • Sudden power failure
  • Abrupt or unexpected shutdowns
  • Server crash
  • Hardware failure
  • Human errors, such as accidental deletion
  • Virus or malware intrusion
  • Ransomware or other cyber attacks
  • Network problems

In most scenarios, users won’t notice a change in DAG environment. But you must restore the failed member server immediately to ensure high availability. However, if the DAG fails due to multiple member servers’ failure, boot issues, network problems, etc., users can’t access their mailbox—leading to downtime. In such cases, a backup can help you quickly restore the Exchange Servers and reconfigure the DAG.

Best Practices to Backup and Restore Exchange Database Availability Group?

Let’s assume we have a DAG setup with three member servers—each having two copies of three databases across three servers

No alt text provided for this image

The DAG configuration ensures no two copies of the same database are stored on one server. Instead, the copies are distributed among all member servers to provide site resilience and high availability. However, if two member servers fail one after another or simultaneously for any reason (unlikely), the server may no longer have access to all the database copies. Therefore, with one member server, at least one database copy will be missing or unavailable.

To restore the DAG, you must set up new Exchange Servers and add them to the DAG. After that, you need to restore the lost third-database copy from the backup or use an Exchange database recovery software, such as Stellar Repair for Exchange.

Best Practices to Follow While Backing Up DAG

Here are some best practices you should follow while taking or creating DAG backups.

  • ?You can use a DAG member to store the backups. You can manually copy the database copy, which is not stored on the member server. For instance, the EXCH01 server holds DB01 and DB02 databases (refer to the above illustration). You can copy the DB03 database from the other two member servers and store it on the EXCH01 server. This way, you can have a backup of the non-replicated database copy on the server that you can use to restore the mailbox database after setting up new member servers.

No alt text provided for this image

  • Although you can copy or create backups from the active or passive copies of the databases from any member server, it is highly recommended that you backup from the passive copies to avoid load on the active databases or production server. Copying or creating a backup from passive copy helps keep the load on the server minimal. ?
  • Create VSS-based backups using the built-in Windows Backup Server (WSB) or an Exchange-aware third-party backup utility.
  • Always create a ‘full’ backup for the first time, followed by incremental backups. You may store these backups on DAG member servers.
  • Keeping backups only on the DAG member server isn’t recommended. Instead, use a separate node for the backup activities.
  • Clear old backups that are no longer needed to prevent discrepancies and conserve storage.
  • Back up all database copies at the same time. You may also create a schedule to back up the databases (excluding active copies).?

Best Practices to Follow While Restoring the DAG Backup

Follow these best practices when restoring databases in a DAG environment.

  1. You must restore local backup from the same DAG member server where it was created.
  2. Restore the database backup on the server hosting the active copy of the database.
  3. If there’s a passive copy on the server where you want to restore the database from DAG backup, set the copy to active state and then restore the database. After the restore operation, you can set the database copy to inactive or passive state. ?

Final Thoughts

Backups are critical to avoid permanent data loss and prevent extended downtimes when a disaster strikes. If you are using the Exchange Server, whether standalone or in a DAG setup, it is highly recommended to maintain regular backups. Database Availability Group also requires regular backups to prevent downtimes if multiple member servers fail due to logical or physical damage (unlikely but not impossible).

However, if the backup isn’t available, does not work, or turns out obsolete, you can always rely on an advanced Exchange recovery tool, such as Stellar Repair for Exchange. The software can help directly restore the mailboxes from the failed DAG member server databases to the newly set up DAG or any live Exchange Server. It auto-maps the source and destination mailboxes, allows mailbox user creation directly from the software interface on live Exchange AD, edits mailbox mapping, and much more in a few clicks. It’s also 4x faster—capable of simultaneously exporting up to 4 mailboxes to destination Exchange Server. You can also save the mailboxes as PST files or export them directly to Office 365 (Microsoft 365).?

Mike Williams

Operating Systems Analyst @ Covenant Health | Exchange On Prem and 365 Administration

1 年

When full backups are run on passive db's are the logs truncated after the backup?

回复

要查看或添加评论,请登录

社区洞察