As I read the news and see images of the recent carnage caused by spring tornados all over the south-central United States, I can’t help but be reminded that June 1st, the official start of the 2014 hurricane season, is now only 14 days away.
If you are unlucky enough to have your building hit by a tornado, that certainly qualifies as a disaster. But, as destructive as tornados can be, rarely will one completely destroy an entire city leaving the electrical and communications infrastructure unusable. Unfortunately, hurricanes have the real potential for that kind of widespread damage as anyone who was living and/or working in Miami, FL or New Orleans, LA when Hurricanes Andrew and Katrina came to visit will confirm.
Here in South Florida, although we tend to think of hurricanes as the major threat from a disaster recovery standpoint, it is actually far more likely that lightning, fire, theft, or flood will cause a data loss as opposed to a major storm. In the end, when a customer has a data loss and needs to recover data, the cause becomes completely unimportant. The only issue is whether the data can be recovered and how long it is going to take.
The vast majority of data recovery work that we do for our customers is almost always single files or directories of files that need to be recovered because of accidental deletion. This type of recovery is simple, fast, and can easily be recovered from a local backup device or cloud based solution. A failure to recover the data requires re-creating the data, which is time consuming and inconvenient, but generally possible. True disaster recovery conditions, either at the site level (your building) or the regional level (your city) will cause catastrophic data loss if the data is not being replicated somewhere else. It only takes a moment to consider the difficulty of trying to recreate all of the existing data that a business needs to just sell and accept payments. And in a real disaster, you would have to purchase, deliver, and stand-up all new hardware, install and configure all of your software over again, and then begin recreating everything necessary to conduct business. The statistics on business survival after a disaster are very sobering.
Fortunately, there are some affordable options for offsite data replication depending on the amount of data you need to replicate, and how fast you would need access to it. For files only, cloud backup services like Carbonite and Mozy Pro work very well, and are very inexpensive. This type of solution does not protect server configurations, active databases, or operating systems. If you have a site disaster, you at least have all of your documents which can be restored once you have running servers again, but you will essentially be starting from scratch on your servers, software, and configurations. If you have products that use active databases (like Microsoft Exchange or SQL Server), this solution is not even an option, as it only protects individual files.
A much better solution is to co-locate a NAS (Network Attached Storage) device at an offsite data center that is used to replicate backups from the primary site. This can be done for as little as the cost of the NAS device (less than $500.00) and a recurring cost of $250.00/month. This provides bare-metal (able to restore to any piece of hardware) restorable backups of whole servers. New hardware would still need to be ordered, but once the new hardware arrived, the restore process would be simple, and relatively fast. Once the restores were completed (which would normally take 1-2 hours per server), your environment would be absolutely identical to what it was when the backups ran. This is by far the easiest and least expensive option for recovering from a site disaster where all of your equipment and data is destroyed. The only downside to this solution is that you would still need servers to restore your backups too, which could take weeks to purchase and deliver. There are many of these types of solutions in the marketplace, but all of them have the same limitation. This is disaster recovery, but not business continuity.
If your business requires that your IT infrastructure is up and running in 24 hours or less after a disaster, then you really need a business continuity solution. There are many vendors in the market place offering business continuity solutions, but the cost places most of them out of the budgets of the SMB market. ShadowProtect, a software backup program widely used to back-up servers, has a cloud add-on that can be purchased to replicate the backups to the cloud. In addition, as part of the cloud service, the customer can use their cloud server infrastructure for 30 days a year to “spin up” the replicated backups in running servers that are accessible to the internet. However, the monthly recurring cost is based on how much storage space is needed for the backups. For 1TB of data, the recurring price to store the backups in the cloud is $870.00/month. The vast majority of our clients are utilizing at least 1TB of space, so although this is a fantastic solution for being able to get your installation back up and running very quickly after a disaster, it is quite a bit more expensive.
No matter what option you choose, now is the time to start thinking about disaster recovery options, deciding how valuable your data is to the operation of your business, and choosing a solution. If you are like the large majority of small businesses in the US, your business can’t live without it.