One of the poster children for cloud computing is Amazon. The online retailer is one of the biggest Cloud providers on the Web today and supposedly also the one with the most experience and understanding of the business.
The Amazon Elastic Cloud Service is provided from at least 6 datacenters worldwide – 2 of them in the United States. One of their customers is coincidentally a provider of cloud services to my home: the PlayStation Network runs in great parts on several Amazon Cloud Services provided from their North Virginia datacenter.
So – over the Easter weekend I had to experience first hand what it means if a cloud service is down for more than a couple of hours – in this case several days:
Of course I survived a couple of days without online gaming (mainly because of the best weather we had for a long long time). But what does this mean to more serious business applications running based on cloud services ? I think it just means the same as for any internal run service – computers can fail, complex systems can fail – and they will fail at some point. The question is how to cope with it and to be prepared for these scenarios.
Me personally – I played Halo 3 on my XBOX 360 on Saturday night for the first time in at least a year. Redundancy on the process level will let you survive any disaster on the technical level.