Amazon Web Services’ Outage Teaches Crucial Lesson

April 21, 2011

As you’ve probably heard, Amazon Web Services’ outage has wreaked havoc on some pretty big online names like Foursquare, Reddit, Quora and lots of small ones too. It’s tough to predict the fallout this will have for these companies, but considering the impatience of today’s consumer, it could be severe. Just as with the recent disappearing Gmail accounts, this incident no doubt leaves many business decision-makers wondering how reliable the cloud really is. After all, if Amazon and Google are fallible, who isn’t?

As the largest Salesforce implementation partner, we’ve learned a thing or two about reliability in the cloud over the course of the past 10 years and 3000+ implementations. Perhaps the most important lesson to be taken from this is that the cloud is no different from any other technology infrastructure in that it requires a redundant process and system to back it up. When your company’s lifeblood is in the “cloud,” it’s not in some nebulous mass hovering above the ozone layer. It’s on a server, just as it was before – just not your server. Therefore, the very same safety rules apply. In this case, redundancy is king.

Keeping this in perspective, the fact is that most of the major cloud providers take redundancy seriously, and have world-class facilities in disparate locations across the globe. In most cases, they provide much more reliability than any organization’s in-house IT team could offer. However, Amazon’s outage teaches us that even the giants are subject to failures from time to time. The cloud is neither magic nor perfect.

Does this mean that embracing the cloud and its inestimable benefits means accepting the occasional outage? Certainly not! We work with companies on a daily basis to plan and build a fail-free, multiple risk redundant recovery systems in order to avoid this very scenario. It is absolutely feasible to keep your systems running smoothly, even in the face of this kind of catastrophe – it just takes some thought, planning, and a partnership with a cloud-integration vendor that knows what they’re doing.

See More