CEO Reflections
Cloud OutageLightning struck on August 8th knocking down Amazon's Europe data center. On the same day, Amazon EC2 was down on the East Coast. Amazon has been down before, in April 2011.Does this imply that Cloud is inherently unstable? My answer is no. Outages happen in company data centers as well, but don't get publicized simply because they don't affect more than one customer. What a business should do to avoid being affected by these occurrences is probably more important.One option is to have cloud redundancy with multiple data centers, wherein your data is resident across two geographically disparate data centers, and the application has smart routing built in to drive access from one to the other. This will certainly increase the cost since now the data has to be replicated near real-time across to nodes. When selecting a cloud service, as a customer you should check if this option is offered, even if at a higher price.Another option is to adopt a hybrid solution wherein the service enables "off-line" access for your team. This can be accomplished with an optional on-premise component that automatically replicates data locally and across the cloud. This Hybrid Cloud solution is possible for some applications but not for others. In the case of Egnyte, we offer a Hybrid Cloud solution that provides ubiquitous file access.The advantages of cloud services far outweigh the vulnerabilities, and as technologies mature I think the world will be looking at more and more "hybrid" solutions, than pure cloud or pure on-premise solutions.