Heath Brownsworth

Perspective

On the morning of April 21, Amazon Web Services, the global leader in cloud computing, experienced a significant issue in their U.S. East (Virginia) region. The technical post-mortem has yet to determine the root cause, but whatever the finding one thing is already perfectly clear. The future of the cloud should not be in question.

With a visible outage that caused major cloud-based applications like Foursquare, Heroku, Quora and Reddit to be impacted or knocked completely offline, it will bring a healthy debate about the viability of cloud for mission critical applications back to front and center. Public cloud in particular will fall under the heaviest scrutiny. But is cloud viability, or even the technical expertise of Amazon Web Services, the real issue? Many companies and organizations have looked to the cloud for low cost computing power and that is exactly what Amazon Web Services, rackspace and GoGrid have provided.

Even though these infrastructure providers have achieved economies of scale and data center efficiencies that very few organizations could hope to obtain, there is a lesson to be learned from this event.  Cloud computing power and advanced technologies are not a substitute for a well developed business continuity plan and a rigorously tested disaster recovery architecture. The onus is not on Amazon, or the other cloud infrastructure providers, but on users of these services.  Simply utilizing cloud infrastructure or “as a service” solutions does not absolve us of this responsibility. We may no longer be responsible for the design, implementation or ongoing monitoring and testing, but we should definitely be asking questions and ensuring appropriate safeguards are in place.

Power failures, hardware failures, server failures and network failures all occurred before the cloud. They just occurred in the cloud and it’s likely they will occur in whatever comes after the cloud.  How you prepare and recover from a failure is where the debate should focus, not on whether the cloud is viable or if its benefits are real.

Without question we learned a lesson today. Some of us learned it for the first time, others were reminded of a lesson learned long ago. We’ll go forward and apply what we learned and begin architecting even more redundant and resilient solutions because we remain fully confident not only in Amazon Web Services but in offering Lawson applications in the cloud.

Heath-