As an equalizer, the cloud is breaking down traditional IT barriers and enabling rapid change. Yet last week's widespread Amazon Web Services (AWS) EC2 outage brings to light that cloud computing isn't infallible. In short, Amazon's four-day service interruption was caused by a networking glitch that subsequently compounded in complexity. You can read specifics from Gartner's Lydia Leong here.
Looking at this from our own business perspective, we live entirely in the cloud at Ajilitee. Our sister company Discovery Health Partners lost some productivity in the initial hours of the failure, but overall our disaster recovery plan worked pretty well in last week's high impact outage. Still, we have some learnings to apply to improve our responsiveness and availability the next time this happens (and there will be a next time regardless of where your applications are hosted; downtime happens in all environments).
It's probable that this event will scare some companies away from cloud adoption. But I believe we can use these instances to build confidence in business continuity planning. This example of cloud failure is a reason to assess what worked and what didn't, and apply a course correction if needed. My latest article, "On the Reliability of Cloud Computing" provides additional thoughts on cloud architecture planning in the wake of Amazon's outage, which might create a stronger and more confident cloud community vs. a fearful one.
Posted April 27, 2011 10:22 AM
Permalink | No Comments |