Data protection using NetApp® SnapCenter® leverages the Volume Shadow Copy Service (VSS) framework in Microsoft Windows Server to coordinate interaction between SnapCenter (the requester), Exchange server, and the NetApp ONTAP® storage system hosting the Exchange data. This interaction supports the creation and management of application-consistent backup NetApp Snapshot™ copies of the Exchange resource.

 

The workflow in the following figure illustrates data protection using SnapCenter.

  1. An on-demand or scheduled backup is kicked off from the SnapCenter Server based on the backup policy assigned to the resource group that contains the Exchange resource (which can be a database or database availability group [DAG]). This backup policy includes a set of rules that governs the SnapCenter backup, such as backup type, DAG settings, schedule frequency, retention, and prescript and postscript with arguments. Through the SnapCenter Plug-in for Windows (which contains the SnapCenter Plug-in for Microsoft Exchange Server and SnapCenter Plug-in for Windows Server) installed on the Exchange host, SnapCenter sends a command to the Microsoft VSS on the Windows server to:
    1. Enumerate the writers (Microsoft Exchange Writer in this case)
    2. Gather the writer metadata (like EDB file, logs, and checkpoint file for each database selected for backup)
    3. Enable making snapshots of the Exchange mailbox database
  2. VSS signals Microsoft Exchange Writer to temporarily freeze the Exchange database and put it in read-only mode; all data in server memory is flushed to the mailbox database. This freeze holds for only a couple of seconds, during which all writes to the databases on disk are stopped and all the write I/O is queued. In the case of passive replica backup, writing of logs copied from the active server to the local disk is stopped during this freeze time. The freeze ensures that data is written in a consistent order and that SnapCenter makes a consistent Snapshot copy, with no significant impact on the performance of the production system.
  3. VSS creates a point-in-time Snapshot copy of the volumes hosting the Exchange database, using the Data ONTAP® VSS Hardware Provider (VSS provider) available on the ONTAP storage system.
  4. When the NetApp Snapshot operation is complete, VSS acknowledges Snapshot completion. The logs are truncated and database headers are updated on the active node.
  5. To resume regular operations, VSS “thaws” the Exchange database and allows it to release the Microsoft Exchange Writer; all queued write I/O held in RAM is allowed to go to disk again. In the case of a passive replica, the replication service resumes log copy and replay.
  6. VSS acknowledges to SnapCenter a successful backup operation. The SnapCenter server updates the backup metadata with this status. If a full backup is being taken, Microsoft Exchange Server truncates the transaction logs (if the database is part of a DAG, log truncation is replicated among all the copies) and records the time of the last backup for the database.

This is the workflow of a backup of standalone active and passive replicas. The only difference in the backup operation between active and passive replicas is where the Microsoft Exchange Writer runs:

Exchange Server always truncates the transaction logs as a result of performing a full backup. In case of a database availability group,  this means that a full backup on one of the nodes causes log replication to other nodes in the DAG, because the log truncation is propagated from the node where the backup is performed to all other nodes in the DAG. This means that you can perform up-to-the-minute restore only on the node where the backup is taken. The challenge is that you lose the up-to-the-minute restore ability on other DAG nodes. Essentially, this implies a transaction log chain gap, which causes data loss. You can overcome this loss by simply backing up all the nodes, but space and management can become a constraint, considering the number of backups to manage.

 

SnapCenter introduces portable backup of Microsoft Exchange resources to replace the SnapManager for Microsoft Exchange  gapless backup feature, where the backup of an Exchange DAG can be offloaded to one or more replica copies instead of backing up all (active and passive) replicas. This portable backup of the replica copy can be used to perform up-to-the-minute or point-in-time restore of the active copy on any failed DAG node. This mechanism helps conserve storage space and lessens the backup Snapshot copy management overhead, because not all of the copies need to be backed up.

 

To learn more about Microsoft Exchange backups, watch this video.

 

Restore Workflow

The SnapCenter Server, which is the requester, initiates the restore operation, using either the SnapCenter Server GUI or PowerShell.

 

Note: At restore time, the active copy for the database availability group could be in a different node than the node from where the backup was taken. In that case, the restore happens to the node that owns the active copy

 

During a restore, SnapCenter instructs the Exchange Writer to coordinate with the Exchange store to mount the database and log disks from the backup Snapshot copy onto the active node, and to perform VSS restore. SnapCenter dismounts the database, restores the database files, mounts the database, and then replays the transaction logs, either up-to-the-minute or point-in-time restoration, as required. This restored database can be mounted as a regular active database.

 

When the restore is successful, the status is updated on the SnapCenter Server.

 

Note: The target node for restore should have access to the controllers hosting DB and LOG LUNs mounted in the node where the backup was taken.

 

To learn more, watch this video, demonstrating the fast recovery of a Microsoft Exchange resource from a backup Snapshot copy taken using SnapCenter.

 

NetApp Single Mailbox Recovery (SMBR) enables data availability for the enterprise through high-speed recovery of mailbox items or public folders that were accidentally deleted. SMBR leverages existing SnapCenter backup Snapshot copies of Microsoft Exchange Server databases to recover individual mailboxes, folders, messages, attachments, and even calendars, notes, and tasks, directly to your production Microsoft Exchange Server or any PST file. This saves time, money, and resources for granular recovery and eliminates time-consuming and expensive single mailbox (brick-level) backups or the need to set up a recovery server to which you can restore a full database backup to explicitly recover mailbox items from. SMBR also lets you search and create a copy of all archived email that matches a given keyword or criterion.

Benefits of Using Single Mailbox Recovery

Minimize the time to restore an individual mailbox.

SMBR can slash restore time, making it possible to restore mail items from SnapCenter backup Snapshot copies, directly into your production Microsoft Exchange Server, or directly into a new or existing PST file. This eliminates the need for a recovery server and the extra steps required to separately import mail back into Microsoft Exchange Server or Microsoft Office Outlook.

Eliminate backups of an individual mailbox.

SMBR eliminates the need for expensive, cumbersome, and disk-intensive daily brick-level backups because you can restore individual mail items from SnapCenter backup Snapshot copies.

Minimize the time to locate all email messages that match specific criteria.

With the SMBR Advanced Find feature, you can search across all mailboxes in selected SnapCenter Snapshot copies, rather than searching one mailbox at a time or bringing an old backup back online for analysis. And you can search by a variety of criteria, including keywords, subject, date, and specific users. It also provides mailbox recovery permission access control and audit trailing.

Minimize the storage space and cost to store and archive your backups.

Because you don’t need to back up mailboxes individually, you eliminate the backup space and cost associated with performing brick-level backups.

Minimize the time required to back up all mailboxes.

SMBR eliminates the need to back up mailboxes individually. Normally, companies do a full Microsoft Exchange Server backup and then run a second process to back up Very Important Mailboxes (VIMs) individually as well. Single Mailbox Recovery eliminates this second process.

For more information about Single Mailbox Recovery, watch this video. You can also download the technical report Best Practices Guide for Microsoft Exchange Server Using NetApp SnapCenter.

 

Cheryl George

Cheryl George is a Technical Marketing Engineer with over 15 years of experience in everything Microsoft and its data protection. Being in Product Management, she helps build state-of-the-art products and enterprise solutions which help solve real-world customer and business problems. Customer engagements is what she enjoys the most.

Cheryl is music enthusiast. She maintains her Zen through yoga and mindful walking. She loves to travel but the irony is that it mostly happens in the head.