What to do when the cloud goes down

Despite all of its beneficial components and quirks, cloud computing is still subject to the same kinds of outages that a traditional computing/networking setup is. That's right, cloud computing is not immutable or invincible; however, it might surprise most people to learn that it does provide users with a sophisticated series of contingent resources to 'pick up the slack' so to speak. In other words, you might not have to do very much at all in the way of taking specific action when your cloud service goes down, as reputable and experienced service providers usually have one or several contingencies in place to deal with any problems.

Multiple availability zones
Cloud service providers are only able to offer uninterrupted service if they have an infrastructure of remotely positioned resources on hand. So, in the continental U.S. for example, an affluent provider might have up to 10 or more data centers (availability zones) positioned across the country. These availability zones are essentially designed to carry the workload of other downed systems. In fact, many cloud computing providers will often copy and spread client data across multiple systems (availability zones) in order to both increase data transfer rate and ensure complete continuity of service. It should be noted however that most service providers will charge extra for use of these availability zones; these charges may be included in the established package price or may be incurred on a pay-per-use basis.

Hybrid setups
If you're one of the many companies that might be slowly making a transition toward cloud computing, but still employ the use of on-site servers, routers or other hardware then you are especially at risk. While it's safe to say that any data or virtualized components that run on a cloud computing infrastructure will be relatively safe from a catastrophic event, the same cannot be said of your localized hardware. This is why you need to consider fully transitioning to the cloud, or perhaps enacting your own hardware backup / contingency system. At the very least this should include some form of daily/weekly backup or data caching. So for example, in a lax scenario, you might want to create a hard copy of everything on your entire system at least twice a week; that way, even if you lose everything you will at most, only lose around half a week's worth of work. If however, you've made the move to the cloud already, you should be good to go.

Some companies are even moving toward creating specific hardware backup solutions for those in the cloud. These systems are often used by those organizations that are still in the process of changing over to the cloud, and are essentially added to one's local setup to collect and backup crucial information on a continual basis. Based on traditional data center designs, these backup systems may also function as another set of hardware resources that can be tapped in need be.

Some are actually using the cloud as a backup
Still, some companies haven't given up on the traditional grid computing model. Some organizations are actually using their cloud service(s) as a backup for their local system, and nothing more. While this method completely forgoes all of the technology of cloud computing, including its limitless resources, it is still a viable use of the services offered by a cloud provider. Deciding to relegate a cloud computing service to the job of backing up a traditional network, is a bit like using a small economy car to tow a tractor trailer. However, if a cloud service is setup to perform this type of task and is priced low enough, it is a great opportunity for those that aren't ready for mainstream cloud computing. In fact, the price of setting up and maintaining a cloud computing backup solution is miniscule when compared to the costs associated with a more traditional setup (which would also include a sizable workforce, each member requiring a salary).

In most cases (given that you have signed with a reputable cloud service provider) you won't have to do that much at all if a cloud outage occurs. The simple truth is that the big cloud computing providers have already thought about how to best deal with outages and that's why they put a multiple region-based system into use. While most people probably cringe at the idea of having their data spread around to multiple sites or seem at odds with having to rely on remotely positioned hardware, you must keep in mind; this is the best system that's ever been created for dealing with loss of data issues. In the traditional model, failures meant not only loss of critical data, but also loads of downtime as well; cloud computing has elegantly solved both of these problems in one fell swoop.