07-03-14 | Blog Post
As a sailor, the notion of being stranded is really, really scary. As remote as the possibility (and location) may be, you have to develop a contingency plan for the rare event that you might find yourself in that predicament. The plan has to contemplate extreme isolation for long periods of time and/or risky transit back to society. But the plans can be so onerous to imagine, you put it away once they are developed and don’t bother to remind yourself of them, except to know they are there. Then, if the unexpected happens, you pull out your backup plan. If done right, it can save your life. It’s why sailors say every trip should be two way, there AND back (unless you go all the way around).
Your data’s life is much like that sailor. Your data needs to be backed up from a production environment AND it needs to be able to get back to a production environment in the case of disaster. But most offsite backup solutions leave your data in isolated locations with many barriers creating risky or unreliable transit back to your production environment. That sounds like a stranded backup plan to this sailor.
Traditional Backup-as-a-Service providers backup data from your server or laptop to an undisclosed data center or storage location. Most often, you don’t know the physical location where your backup data resides. If you have no idea where your data lives, how would you know if it was in the path of a pending hurricane or tornado? How could you prepare or react in advance? If you don’t know if physical location of your backup data, you might as well consider it stranded.
Let’s say that you DO know where your data lives, and you have significant volumes of mission-critical data. How do you get all of that data back to a ready-to-go production environment so you can restore to normal operations before the health of your business systems are severely compromised?
One class of backup solutions includes cloud-backup providers like Mozy, Carbonite, Duva, VaultLogix etc. There are 100s of companies that allow you to replicate data from your servers (or laptops etc.) to their storage environment. While this provides “point-and-click” backup, I’ve always been struck by what must go through if you actually experience data loss and have to recover your backup data. Consider the following scenario.
Let’s say you have 500 GB of data from 4 servers backed at a backup service provider. Your recovery begins with a search for servers, network, and other infrastructure to which you can restore your data. That can be a very time consuming if you haven’t invested in redundant infrastructure and already have it standing by. Whether you are buying it after experiencing a disaster and need everything expedited, or purchasing it in advance, it will be expensive. Next, you have to get the data from that backup silo, wherever that is, and onto that new infrastructure, wherever that is. If you have funded redundant equipment, housed somewhere, in various states of readiness to power-up, you have a backup service that has your data and separately an infrastructure stack sitting somewhere to take that data. This model has even worse IT resource utilization than most physical servers; you’re paying for it, but can’t use any of those resources in the meantime.
This approach ignores the fact that there’s already a ton of infrastructure stood up and available, thanks to the advent of cloud and hosting business models. These business models depend on the ability to turn up new infrastructure quickly with appropriate incentives on fast deployments; the sooner they can deploy, the sooner they can start billing. This means they already have “at-the-ready” infrastructure. Better yet, you don’t pay for until you need it. Imagine if your data was backed up to a data center with a complete spectrum of on-demand infrastructure at your disposal that you could contract and pay for only when you need it.
Bottom line? Don’t strand your data in some unknown place that can only send it back to you – at a really slow rate. Backup your data to a location that can immediately begin helping with your restoration and provides local access to your backup data from on-demand infrastructure.