Elastic Computing: the cloud term for Stretch Armstrong.
One of the bestselling terms, in my opinion, for Cloud computing is Elastic Computing. Elastic Computing is a pay-as- you go server model. Traditionally, when a physical server was purchased for a datacenter it was spec’d out to be able to support what you predict entire peak load will be for whatever application it is supporting. For instance, you could purchase a server that has the following generic spec’s: Quad-Core processor, 500GB HDD, 12GB Ram. But what happens if over the lifetime of that server you actually only use very little CPU, 200GB of the 500GB, and 6GB of the 12GB? You find that the money and resources you put into your static Physical server where actually very much over provisioned and lead to resources and monies wasted!!!
What elastic computing allows you to do is “rent” a VM in the cloud with what you know you need now, and then you can grow or shrink on demand. This is the pay-as-you go philosophy. Scaling can be manual but typically for larger or more critical applications it can be setup to scale automatically and will grow or shrink depending on the factors you provide. Elastic Computing is similar to the original stretch Armstrong doll from the 70’s. It could be pulled to grow four or five feet from its original form and shrink back.
So if you are a startup company and you are expecting your inventory or content to grow dramatically then I would suggest elastic computing to be your best solution. Not only does it provide the pay-as-you go model but allows for other things such as disaster recovery, content distribution, load-balancing, etc….In the cloud there are endless possibilities for computing, but they do all come with a price tag!!!