Offsite backup: a necessary evil that can be unreliable, hard to track and expensive to support. We know; it’s not what you got into technology for. Nonetheless, the business depends on putting mission-critical data offsite for those just-in-case moments of disaster.

The problem is, most “disasters” tend to be a single file that someone accidentally deleted or that got corrupted somewhere along the email chain.

Granted, if the primary business location becomes a smoking hole in the ground (perish the thought), we would be thankful to have all of that data available in a remote location somewhere so at least we know it could be restored “some day” instead of “never.” But for the normal course of doing business, what a hassle!

Option A: Recover your offsite backup file from tape

Let’s say accounting needs one file back from offsite tape backup. OK, here we go:

  1. figure out where the file was located,
  2. identify the right tape,
  3. recall the right tape from the vault,
  4. retrieve the tape from the offsite location,
  5. halt any currently running backup processes,
  6. swap the tape back into the library,
  7. attempt to restore the file (50% – 77% fail rate depending on the cited source),
  8. copy it back to the original location,
  9. notify and validate that the restored file is accurate,
  10. pull the tape back out of the library (make sure you have the right one),
  11. prepare for sending it back offsite,
  12. return tape to offsite location and verify its arrival.

When you add up all of the time spent, it could easily reach four hours for an experienced system admin to reliably restore one file. That’s four hours of time NOT spent on the mission-critical projects that are tied to helping the business achieve current goals, one of which is not spending inordinate time on the restoration of single backup files.

Option B: Recover your offsite backup file from snapshot

Maybe you’ve convinced your organization to make the transition from tape to a digital alternative like regular snapshots, meaning taking and storing images of the entire server on a daily basis. Now the restoration of that single file looks like this:

  1. figure out where the file was located,
  2. identify the right snapshot,
  3. download the entire server snapshot,
  4. restore the entire server on local hardware,
  5. find the file,
  6. attempt to restore the file (better odds here over tape),
  7. copy it back to the original location,
  8. notify end-user and validate that the restored file is accurate,
  9. delete the temporary server and downloaded snapshot.

From the initial request to getting the file restored and cleaning up the restored snapshot server, you might have cut the four hours of time to two. Nonetheless, it’s still the time of a resource with a well-paid skillset whose time is undoubtedly better spent elsewhere. Let’s hope you only get these requests occasionally.

Option C: Recover your single file with self-serve file-level restoration

When you have the opportunity to guide your organization to consider an offsite backup and recovery that makes the most effective use of your resources, consider the next two options when factoring in the real costs of your staff.

This is where your blood pressure starts decreasing and you get to allocate more resources on your critical business projects instead of recovering errant files.

If you can leverage a recovery technology that affords file-level restoration, you can use point-and-click with Windows, or console access with Linux, to do the following:

  1. login into your backup server,
  2. browse (Windows) or change directory (Linux) to the location of your file,
  3. select the version you want to restore (yesterday’s? last weeks?),
  4. specify target destination,
  5. click file (windows) or enter console command (Linux),
  6. wait for file to download,
  7. notify end-user it has been restored.

Our clients report it takes 5-15 minutes, depending on file size. Obviously this will depend on how much data you’re restoring and your connection speed. It’s also fair to say that you probably don’t need your most senior system administrator performing these tasks. They don’t have to know how to spin up a new server, be familiar with tape libraries, or unwind a snapshot. In addition to spending less time, you may be spending less for the time spent.

Did you notice the omission of verifying the file restored correctly? Not necessary if you have integrity validation at the time the backup is performed. More about that in a future post.

Option D: Managed file-level restoration

Some providers are now offering a managed flavor along the lines of Backup-and-Recovery-as-a-Service (everything needs to be available “-as-a-Service” these days, right?) In this case, file restoration might be as simple as:

  1. call or submit a ticket with the name and destination of the file to be restored,
  2. notify end-user when it has been restored.

Will you pay more for file-level restoration and managed recovery services? Probably, but don’t forget to factor in the time, headaches, and human resources of managing the “less-managed” solutions.

Cheers to fewer, faster recoveries!

Credit to Nick Lumsden for insights into the pains of backup and recovery.


RELATED CONTENT

7 questions to keep your offsite backup and recovery out of the outhouse

7 business drivers for your backup and recovery strategy

White paper: Disaster Recovery