Quite often, the team at Hedgehog Hosting is asked about disaster recovery. Sometimes, the question is posed as, "What does Hedgehog Hosting do for disaster recovery?" Then again, and more often, clients ask, "What kind of disaster recovery do we (the client) have in place?"
Both of these are excellent questions and the answers, while sometimes similar, are usually very different. That being said, disaster recovery discussions should always start with an assessment of one's individual tolerance for a situation rather than an assessment of what others are doing.
In assessing potential disaster options that may arise for Hedgehog, one of our initial decisions when we started our company was to have multiple carriers for our cellualar service. This decision was made mainly because our cellular service is our lifeline to our clients. We use our web-enabled devices in most every aspect of what we do to support our clients. That being said, we felt that it was essential for us to have redundant carriers when it comes to our contactability via cellular lines. We also use our cellular devices to ensure we can be contacted by our monitoring services at all times. Knowing that this was such an important facet of our business we chose to diversify our carriers. This decison, while costing a bit more and requiring a bit more management, was the right decision for us.
Today, AT&T experienced issuses with their mobile messaging service that prevented and/or delayed mobile messages from being sent via email. Had this been our only provider of cellular service, we would have been without the use of our internal and external monitoring systems during that time. Fortunately, our diversity allowed us to operate normally with no real interuption of service.
In the end, disaster recovery should be looked at in two distinct ways. The first is, "What are the real potential issues that could impact one's business?" Stay away from decisions that involve issues that are least likely to happen. Don't make planning for X situation a priority if the chance of X situation happening is slim to none. Too often we are planning for a major disaster rather than assessing the immediate and more real threats. Second. Plan to support the mission critical needs first and then establish a solution that supports those needs. From there you can, establish a different set of criteria for the non-mission critical systems and handle them in a way that supports their needs.
Separating different pieces of one's architecture for the purpose of disaster recovery is an idea that more groups should entertain.
No comments:
Post a Comment