by Jason F Bennett
22. April 2011 10:21
photo by Origamidon on Flickr
Early yesterday, Amazon's cloud offering, Web Services (AWS) had widespread failures and latency issues, effectively blocking Amazon's clients from serving up online services. This effectively blocked companies like Reddit and Hootsuite users from their main services. Hootsuite was completely shuttered for the day, and Reddit blocked logins to the site. With this reminder of the risks associated with heavy investment in the cloud, it's worth surfacing a couple of terms to think about when considering a cloud offering.
High Availability. This is the idea that moves beyond mere uptime for all of your servers, and focuses resources to make sure high-business-impact components are not just up, but have multiple systems of redundancy. No failure allowed.
Points of failure. Again, server uptime isn't sufficient for discussing problems that arrive 1 or 2 hops away from your customers. Diagramming the network points between your services and your customers can identify weak links that won't surface in mere platform uptime analysis.
The irony in Microsoft's recent "all in" cloud messaging is that the for businesses focusing on online services, supplying "brick and mortar" customer service argues the vice versa of traditional disaster recovery. The message is still the same - hedge your bets on platform and network, investing in solutions that deliver 24-7 global services that customers demand.
Company Blog
Subscribe
Log in