Undoubtedly phones were ringing and there emails were flying this week in the wake of an Amazon Web Services (AWS) outage that adversely affected organizations all across the Internet. But now that the initial hysteria has passed there’s an opportunity for IT service providers to have one of those teachable-moment conversations with their customers.
Line-of-business executives embraced public clouds because it allows them to both treat IT as an operating expense and, frankly, deploy applications faster. But in the rush to the public cloud many of the fundamentals of data protection have been ignored.
Even the public cloud isn't perfect
There is no such thing as the perfect platform, and there is always going to be some unanticipated single point of failure. However, in the immortal words of Oscar Wilde, organizations need to expect the unexpected. In the case of a public cloud, an IT organization should be able to fail over their applications to either a local data center or another public cloud in a matter of minutes. Thanks to rise of virtualization, that is more feasible than ever.
The trouble is most organizations haven't really gotten around to implementing that level of business continuity assurance. Most of them assume the public cloud has built-in redundancy. In fact, because of that assumption many of them don’t even back up data residing on a public cloud onto another site.
The truth of that matter is that in terms of maintaining availability AWS still does much better than just about any internal IT organization could do on premise. But given the fact that so many organizations experience the same downtime problem at the same time, there’s now a unique window of time where business continuity is on the minds of every business leader. Every one of them, after all, reports to a board of directors that is specifically charged with minimizing risks to the business. Many business leaders are about to have some uncomfortable conversations about risk management with the board of directors.
Spotlight on business continuity planning
Given the millions of dollars in lost revenue racked up this week, the cost of implementing a robust business continuity plan appears comparatively trivial. Many of the members of those corporate boards assumed the organization had some form of IT contingency plan. The hue and cry being generated this week makes it clear that many of those organizations did not have anything approaching a contingency plan.
In time, hybrid cloud computing will advance to the point where the outage of one cloud won’t have nearly as much of an impact as it does today. In the meantime, however, the AWS outage highlights once again just how fragile IT environments really are. Each new application service creates a tangled web of interdependencies that make it difficult to predict how the sudden disappearance of one piece of application or IT infrastructure will impact the overall environment. The hope is that there’s enough redundancy built into the system that it will never completely fail. But as everyone learned this week, hope is a poor substitute for an actual business continuity plan.