Back in the old days, businesses would all keep their IT infrastructure “on premise” (in the building where they operate) and would manage it internally. Seems like the most natural thing in the world, but you will soon understand how inconvenient this actually is
Buying computing power is a long-term and expensive investment, that takes a long time to change or to replace, so you want to do it right. But doing it right is actually not as easy as it may sound: it is usually very hard to predict how much computing power your company needs, and even harder to predict what it will need in a year or so. This is especially a problem for startups who have no clue where their enterprise will be in the next few years.
Let me illustrate this. Just picture yourself in the skin of a young aspiring entrepreneur about to launch a new super-funky online business. Now of course, you have no idea how your e-business is actually going to take off, and how much traffic you will generate in the next few months, or years, and at which rate. So your best strategy is (I hope) to invest in A LOT of computing power to make sure your servers never crash. And just like most entreprises, you end up hosting your initially small e-business of three and a half daily customers on over-the-top, NASA-style web servers, and pay for way more than you actually need.
Does not sound like the most convenient thing in the world, does it? Well, it did not take very long for Amazon to target the IT infrastructure industry as its next victim.
In 2006, Amazon completely revolutionized the IT infrastructure industry by inventing what is commonly known as Infrastructure as a Service (IaaS), or Cloud Computing. The concept consists of offering online access to distant servers via their web platform, Amazon Web Services (AWS).
AWS allows businesses to rent computing power for as long and they need, and only charges them “only for what they use”. This means that enterprises could rent a server via the AWS platform for a day, and would only be charged for that day. No more long-term investment is necessary.
This is revolutionary in many different ways.
Firstly, no more need to guess what computing power you will need. You can scale your infrastructure up and down anytime, and have your computing power updated within minutes. Your website is suddenly making a buzz and experiences an unexpected amount of visits? Within minutes, your IT infrastructure can be scaled up to handle the extra traffic.
Secondly, IaaS allows flexibility. Companies can now test different IT infrastructures, different servers, different databases, etc. Since they are only charged by the hour, they can cancel their rent and have their infrastructure updated in no time, anytime they want.
Thirdly, and most importantly, it has made computing power easily accessible to everyone. IT is no more a constraint for startups and young entreprises. Within minutes, I can have the servers necessary to launch a new website and scale up as demand increases. Everyone is playing in the court, Startups and Multinationals. Everyone now has access to nearly unlimited IT resources. If I want, from my laptop in my student apartment, I can have access the same computing power as Facebook or the NASA. This was not imaginable 15 years ago.
And the list of advantages of IaaS goes on and on: reliability, fault tolerance, security, ease of use, and cost efficiency.
To this day, I do not see any disadvantage that Cloud Computing has brought to the IT infrastructure industry. Can you?
