Summary

This article explains how to protect data originally backed up to a SnapVault Secondary device on tertiary storage (in most cases tape) using SnapMirror to Tape. This article assumes that you are familiar with the basic concepts of SnapVault, Open System SnapVault and NDMP

Step By Step

The data selected for a SnapVault job is stored in QTrees in volumes on the secondary device. At the conclusion of the SnapVault job, a snapshot is taken of the volume that contains these QTrees. These snapshots are now “point-in-time” backups or recovery points of the primary data, and are used to recover data at later times.these snapshots can be backed up to tertiary storage (tape in most cases) for archiving purposes. The recovery of the data from these data dumps however yields only the one snapshot that was backed up to tape. So using this method to store data on tape cannot be used to reconstruct a lost volume on a filer, since only a few snapshots are normally archived. In order to recover the contents of an entire lost volume from tertiary media, the volume needs to be backed up using the snapmirror-to-tape technology, which is supported by DPX. This article describes how to backup and recover volumes using this technology 

Differences between Dump and SnapMirror-to-Tape for Volume Data Protection

The key differences between the Dump and the Snapmirror-to-Tape technologies for protecting data to tertiary media are:
 

 

Storing Secondary Data to Tertiary Media using SnapMirror-to-Tape (smtape) for D/R Purposes

To allow DPX to backup a volume from a filer to tape using smtape, one has to configure the software first. From within the DPX GUI, click on Configure|Enterprise, and click on the filer, and select smtape from the Backup Type field as shown below:
 

 
 
After completing this step, define a Backup|NDMP job selecting the appropriate volume(s) for smtape backup.
 

Recovering Data from Tertiary Media

To recover a volume that was backed up to tape using the smtape functionality through Syncsort, a new volume needs to be created on the new destination filer first. The size of this volume needs to be equal or larger than the size of the backed up volume. Tip: if the size of the backed up volume is unknown, create a large (e.g. 10 TB) volume with Volume Guarantee set to None, which means that the volume size will not be taken out of the containing aggregate. The restore process will then resize the volume at completion. After the volume is created, take the volume offline (using FilerView). The volume needs to be offline before the restore job is run.
 
With the new volume created and offline, add the filer that contains this volume to the DPX Enterprise (from within the Enterprise window in DPX). Make sure that the logical name of this filer is different from the filer that originally contained the volume. Then define the restore job by navigating to the Restore|NDMP window, and selecting the snapshot of the volume to be restored (source = left pane) to the newly created volume on the new filer (destination = right pane) (as shown below), and run the restore.
 

 
 
After the restore job completed, the volume will be online. The size of this volume (along with all the other characteristics) will be identical to the backed up volume. This volume will also now contain all the snapshots that was part of the backed up volume. In order to restore data from this volume, the volume needs to be made writable. By default, this volume is left in a read-only state (SnapMirror destination). This is accomplished by issuing the following command on the filer:
vol options <volume name> snapmirrored off
with <volume name> the name of the restored volume. A message will confirm the action (Flexible volume 'Test' is now writable).
 
At this point the data stored in the snapshots on this volume can be used for recovery. In the case of a server recovery using the BMR operation, one can simply provide the new filer’s ip address and volume name to the BMR software when prompted. For Instant Availability (AI) and file/folder level SnapVault restore, one will have to use the Alternate Secondary dialog box to point the software to the new filer and volume before executing the AI or file/folder recovery. This is done by right-clicking on the appropriate snapshot from the Restore|SnapVault window, and selecting Alternate Secondary from the drop down menu. From the menu select the new filer as the alternate secondary, and then select the newly restored volume as the volume on this filer to use for the recover. With this step completed, proceed as normal with the IA or file/folder level restore.
 

As noted above, this method should be used for protecting the SnapVault volumes on the secondary filer to tape. Make sure that the data contained in the DPX Catalog is also well protected to allow for the recoveries explained above. If the catalog is backed up to the filer using OSSV and the master server and filer is lost, one will not have access to the catalog to do the smtape restore. So it is good practice to backup the catalog to tape or other disk (outside of the filer) that can be used to recover the catalog in case of a disaster.