I have been writing about Cloud Computing since quite some time. Even before since I first used a Public Cloud myself. That was courtesy http://pcquest.com where I actually deployed a few Private Cloud Middleware and evaluated it. That was way back in early 2009.
Since then I have been using Cloud Computing in some or the other form and have felt and tried making others feel the benefit of it. But this time I actually have got some hard figures which had made me believe that my feeling about cloud computing was never wrong.
This feeling came from a project which I was working currently. It was essentially an e-com website for which I was responsible of designing the entire infrastructure. Being a consumer focused or as we call it B2C ecom site, the expected hit rate was pretty high. So in any normal scenario if I would have opted for a traditional deployment in the existing datacentre which the company already had, I would have required a minimum investment of 15 to 30 Lakhs of Rupees to build a setup which can work for at least next 3 to 5 years without any further investment.
But I opted to go with AWS or as we also know it as Amazon Cloud. That was the good part. But the bad part was, the project didn’t work out due to some business reasons and the project died after 3 months of work.
Now, as the entire setup was on cloud, I actual till now have provisioned only the bare minimum capacity of cloud instances, as the site was not yet live. And it was only used for development and testing purposes. I had the luxury of doing so just because I knew that scaling this site will be a matter of minutes when we actually go live, as my entire setup was sitting on the cloud.
We were even doing occasional stress testing on the site. And at that time, I used to scale the infrastructure to the capacity which I was planning to use when we go live. And once the stress testing used to be over, I used to scale it down back to the minimum possible capacity to just keep the site and the backend running.
Now as the project had failed, I had the luxury to bring down the infra cost to ZERO by turning all the servers down. And not just that, when I sat and calculated what is the money I had spent on this projects for the infrastructure. I myself was amazed to see that it’s less than 20,000INR.
So if I go by the most pessimistic calculation as well, I have only lost 20k instead of 1500k for infrastructure in the event of a failed business!!
So the bottom line is. If you are stable in your business and running it pretty well, then there is a point in sitting and doing a calculation on should you move to cloud or not. But if you are a start-up and have a chance to fail or a need to adapt significantly in the short run, then Only Go with Cloud at least till the time you are successful or stable!!
P.S: I actually had worked for a company long back, which started a Job Portal in India and spent a few Crores on buying the infra. And then final the business died leaving those servers to be used/misused by the rest of the business run by the same group. I don’t blame them as at that point of time, there was nothing called Cloud Computing!