Success stories detail how moving to the cloud cuts operational costs, while also allowing businesses to respond to unexpected demand. Instead of trying to guess the number of servers required to run an application or service, businesses can start small and then slowly – or quickly – grow as the demands increase. It’s all very flexible. What’s interesting, however, is that the flexibility doesn’t stop with the number of servers; the cloud data center can now be flexible as well.
And that is an important factor when developing contingency plans for your websites and applications.
Natural disasters threaten traditional data centers
In the old data center model, you’re constrained by the sturdy walls of your provider. That can open you up to greater risk in the event of a devastating natural disaster.
On the East coast, data centers are built to withstand hurricanes. In the Midwest, data centers have tornadoes to contend with. And in the West, we deal with earthquakes. However, no matter how much time and engineering we put into a structure, history has prodven time and again that there is nothing we can build which nature cannot destroy. Despite all our planning, a perfectly timed event will bring a data center to its knees, along with everything in it.
And when the data center goes down, so does your application.
Cloud data center portability mitigates downtime
With a cloud data center, the repercussions of a disaster can be dodged, and sometimes seamlessly.
By now, everyone should be familiar with the concept of virtual machines. Because virtual machines are really nothing more than software, they can be moved around to different hosts fairly quickly and easily. There is, however, still a dependency on local resources needed to run these servers, and these resources are still hosted in the traditional data center. With cloud computing, these virtual machines can be reduced to nothing more than a template – a menu, if you will – calling out the needs for a particular server. Using this template for your web, application, and database servers, you have the power to move your servers anywhere.
Let’s take the most recent disaster, Hurricane Sandy, as an example. That storm was devastating. Approximately 253 fatalities and $65.6 billion in damage were reported.
If you hosted your servers in a data center on the eastern seaboard, you had a right to be worried. However, if you were using the cloud, you could take advantage of the portability it offers by standing up servers far away from the hurricane. In a matter of what could be minutes, you could have a duplicate environment running on the west coast, using the templates for your servers to ensure consistency. Once the servers were running, you could restore your data to the databases using backups (or whatever means your business prefers) and then perform a cutover of your application, directing traffic from the east coast to the west coast.
Data crisis averted, and your drivers and clients would continue using your application none the wiser.
Of course, this is an overly simplified high-level overview of what is needed to move an entire application from one cloud data center to another. There is a significant investment in time and planning (and testing) in order to accomplish this feat, but it is doable. Further, a hurricane offers the advance warning it’s coming and allows for the execution of an organized plan. In the case of tornadoes and earthquakes, some (if not all) of your cloud infrastructure needs to be built in advance in order to survive the specific event. Nonetheless, the concepts are the same.
As cloud computing continues to mature, more and more businesses will no longer feel the constraint of a fixed geographical location.
And that can help your business weather the storm.