Advertisement

Solving the Snapshot Migration Dilemma

By on
Read more about author Stefan Hirschberg.

Detailed planning and thoughtful consideration are critical for IT leaders moving their unstructured data to a new platform, whether it be on-premises or in the cloud. If they aren’t cautious, enterprises may encounter cost overruns, project delays, and data risk.

Migrating snapshots from one network-attached storage (NAS) to another ranks high on the list of important project considerations.

But the words “if and how” can create headaches for IT leaders regarding snapshot migration. Implementation of snapshots on a NAS platform is incompatible between different platforms and specific just to that platform. Many technological differences also exist in how vendors use snapshot functionality. Some do copy on write (COW), others do redirect on write (ROW), and the list just goes on and on.

This leads us to the question: What do IT leaders need to consider when facing the unavoidable snapshot migration dilemma?

First, a snapshot is a point-in-time image of a file system, or a part of a file system, which consists of pointers to blocks or files that haven’t been edited since the snapshot was made. It only physically stores later changes (new, deleted, or changed files).

Although snapshots (or images) of files look like logical full copies, they only consume the spaces for the changes since the snapshot was originally taken. They don’t consume much physical space on the storage system, they’re created in just seconds, and even so, they represent a (virtual) copy of a file system.

Often, snapshots are used as complementary backups to restore from logical errors. This can happen if files on the original system copy get corrupted, deleted, or encrypted either by accident or through cyberattacks. Snapshots can also be used for both short-term (limited) retention kept only for hours or days, as well as long-term (extended) retention to provide full backup history depending on the agreed SLA for backups.

Migrating Between NAS

Now, what do you do with snapshot history when you’re planning to migrate from one vendor’s NAS to a different vendor’s platform?

Just copying over some pointers and blocks (this is basically what a snapshot represents on a system) won’t be enough. The new vendor’s system won’t know how to use the pieces of information because it has its own way of creating and maintaining snapshots.

The recommendation, when dealing with a NAS system with a limited snapshot history of days or even hours, is to migrate only active data and recreate the snapshot policies on the new platform. The original, old system can be kept online as new snapshots are created and the migration runs, and it serves as a safety net until all snapshots are aged out.

Make Migration Easy

Unstructured data management solutions available provide the ability to individually select snapshots and migrate them to the new target. In many migrations, it is preferable to avoid the content of snapshots and exclude them from your migration streams. These particular solutions can make it so that you only need the capacity of the biggest snapshot plus the changes between each snapshot.

Imagine having a filesystem that keeps yearly snapshots over 10 years (2011 – 2021) and monthly snapshots over the last 12 months and you need to migrate these all to a new system.

Following these slightly simplified steps, you can end up with a similar space efficiency you had on the source system:

  • Rename the oldest snapshot (2011) on the source to a universal name (for example, ‘migration’)
  • Then, migrate the path (‘migration’) over to the target path
  • Take a snapshot of the target system and name it ‘2011’
  • Rename the source snapshot back to its original name (2011)
  • Rename the second oldest snapshot (2012) on the source to the same universal name (‘migration’)
  • Migrate that to the same target path
  • Take a snapshot and name it ‘2012’
  • The solution will only copy the files that changed between the 2011 and 2012 snapshots
  • Repeat this process for all yearly snapshots and the monthly snapshots from oldest to newest

If you need to verify to an auditor that you’ve perfectly migrated every snapshot, perform a switchover in the software after every migration step and let the software compile the chain of custody report for each migration.

In Summary 

A solution with powerful migration functionality can give businesses total control of their unstructured data management projects. The right platform provides quick, efficient migrations with massive scalability, easy-to-follow wizards, and strong protection and security whether you’re dealing with snapshot or live filesystem data. Additionally, a chain of custody document provides proof that the content of written files matches the content of files from the source system.

Leave a Reply