In our previous blog, we set the stage for why traditional backup and DR scenarios don’t work in the modern threat landscape. We discussed how we have gone from hardware failures and natural disasters to an ever-evolving landscape of zero-day exploits and intelligent threat actors. This blog will start talking about air-gapped backups, methodologies, and some of their limitations.
PART 2
Air-Gapped Backups
For starters, what is an air gap? An air gap is an old term that referred to the fact that there was no direct path for electricity to cross between two things. A literal gap of air was created to insulate one thing from another to prevent electricity from crossing over.
In IT, it basically means ‘not connected to the network’. Putting that context around it, how do we have air-gapped backups? Well, essentially it means that once backups are copied to a medium, it is taken off the network, so it’s not air gapped when data is populated, but it is then removed from the network soon thereafter, creating the gap. Common examples are tape backups or rotating USB hard drives.
Tape drives were our original backup methodology that allowed for offsite copies. Once data is on the tape and the tape is carried away, no one on-network can access it. Tape drives, however, have a really bad name from the early 2000’s when their reliability was questionable, and the software that put data on them was often poorly designed, creating random reads from a sequential medium.
The more “modern” solution to tape drives was the USB hard drive. This fixed the random access problems we had with tapes and the tape drive maintenance issues. Furthermore, USB hard drives (when properly cared for) were often more reliable.
Both tapes and hard drives gave us a truly offline copy of data that we could take offsite, but they had similar constraints that made them problematic. The most important and constant issue was that both required a human to do something for backups to keep running properly. If the old tape/HDD was left connected, then the data would either be overwritten, or backups would fail. This routinely caused backups to not run, or there to not be historical versions.
Other constraints related to humans also included simply unplugging the old drive, but not taking it offsite – which meant when there was a disaster, both copies of the data would be lost. Finally, when backups WERE rotated successfully, the offsite storage location was often not optimal. For example, drives were just left on a desk at an IT person’s home, which opens a whole host of issues around compliance and data management.
So, What’s Next?
So, due to this, everyone moved to non-air-gapped solutions, Mirazon included. In the age where the main things we were protecting ourselves from were physical disasters (fires, tornados, meteors) or virus outbreaks, having a segregated but online system, ready and waiting to restore from was considered way more useful than having to reconnect media and wait for restores.
This has lead us to our current dilemma, we want to be able to restore from something online, that’s immediate, but we also want the survivability of an air-gapped backups solution. What are we to do?
This is where immutable backup copies come into play. Check our next blog on why this hasn’t been as viable as we would like it to be, and what’s changed recently.
If you’d like to learn more about air-gapped backups or immutable storage solutions, please call us at 502-240-0404 or email us at info@mirazon.com.