The cloud, the cloud! The amount of media hype around cloud is at an all time high. No matter where one goes now you see cloud being pushed on you. The little TV screen in the back of New York cabs, high above us in the United terminal at SFO… it’s all over the place. Why? Well, some would argue that using a cloud service is lower cost than traditional computing, and I would agree with that. Another common cloud value proposition is that it enables a more rapid transition to mobile computing. However, there’s one use case for cloud that often flies under the radar: it enables better disaster recovery — something that many organizations, large and small struggle with.
As a former IT leader myself, I was often heavily involved in disaster recovery planning so I know how challenging it can be to execute on a DR plan. In fact, I often joke about that fact that every company is an expert in backing up information, but restoring it is where the challenge is. DR challenges have no prejudice. Small companies, large companies, government organizations, overseas, domestic, you name it — they all struggle to achieve a recovery point objective of 24 hours. Think about that statement. 24 hours, or one day of outage before things are brought back up. This can cost companies millions of dollars.
One of the benefits of the cloud is that the infrastructure is located off premise, so there’s no local infrastructure to deal with. This shifts the problem to the someone else, but this doesn’t mean we can’t still all learn a lesson from the people it gets shifted to. Some cloud providers are trying to build resiliency into their offerings to dramatically improve uptime. Towards the end of the year, I moderated a luncheon, sponsored by one of the mainstream cloud providers, at the IT Roadmap show in DC. One of the goals of this provider was to create a service with a recovery point objective of zero — that is, instantaneous fail-over so any customer using the service will never have any downtime. The network may fail, but downtime will not be from a failure with the cloud service.
Organizations that want to use multiple cloud providers, or build their own cloud environment, should strive for a similar recovery time. The first step in toward this would be to embrace the cloud for mission-critical applications. I know there are a handful of companies that have already done this, but the majority have not. The next step would be to tear down the walls that separate the IT silos. And the final step would be leveraging a Software-Defined Network (SDN) to make the network a dynamic resource.
SDNs have obviously become one of the hottest trends in networking over the past year or so and will eventually become a key part of enterprise cloud and virtualization strategies. Why? Well, an SDN can enable the seamless movement of IT resources across a data center, or even between data centers and cloud providers. Traditionally, the network has had very little in the way of agility and flexibility, but SDNs can be used to make the network as agile as virtual compute infrastructure.
One of the keys to successfully implementing SDNs is to embrace the concept of automation, and automate as much as you can. This includes obvious tasks like coordinating the movement of security, storage, and network policies with VM migrations, but also includes jobs such as automating the movement of WAN optimizers and application delivery controllers. Additionally, find tools with robust GUI front ends and track all of your changes — manual changes done through command line interfaces are slow, error prone, and can be hard to record and replicate. These types of tasks are near impossible with legacy networks, but SDNs make true IT coordination possible.
So if you’re looking for a reason to go cloud, considering making disaster recovery part of the justification. Just make sure that you’ve at least considered using an SDN to make the network part of the overall cloud and DR strategy — it could be the difference between a DR plan that works or one that fails.