I have been dabbling into random technical stuff for a while now.(Guess this is what internship does to you when you take it seriously).And what use is dabbling without sweet babbling.Soooo, AUTOBOTS LET'S ROLL..
Cloud is the new buzzword of the IT industry.Everything big happens 'in the cloud'.To get things started, cloud computing is treating computing power as a commodity rather than a computer itself. Lease and release computing resources and pay as you use,somewhat like electricity. So, instead of installing a dozen new computers,you talk of getting the computing ability of a dozen computers.
Cloud Computing is not much of a technological breakthrough.It is more of a paradigm, a model. The concept is pretty much obvious and has been lingering around for a while.It is just another idea whose time has (finally) come.(The word "cloud" became a catchphrase after Google CEO Eric Schimdt used it) In my understanding, it is taking division of labor to the extreme.Just like you let the pro guys design your OS, let them do your computing as well.Simple economy of scale and specialization argument at the first glance.
The beauty and power of the model is obvious once you look at it from the consumer side..You have a new, BIG idea(or such is your conceited belief).You do the obvious and a start-up is born.But you never know whether the users will crown you or crucify you.You may be the next Twitter.(Twitter had just 50 odd employees when it stormed the net. Now it has 140 and isn't taking any more.These guys do have a thing for 140!!(If you don't 'tweet',140 is the word limit for a 'tweet')) You may be just one of the daily casualties of the internet - gone without a trace.Success and failure- both grow exponentially in cyberspace.So,here's your first headache - how many computers do you buy?? Go for the overkill ?? Play it safe and stunt your adventure?? Enter cloud computing as a knight in shining armor.Take as much computing power as you want and pay as you go. And you are really on cloud nine.And even when you survive and stabilize,no maintenance and operational overheads, no fear of crashes and failures.The IT equivalent of 'single window policy'.
The next obvious question - what does the service provider gain? A little bit of history will reveal the answer.(In CS, 90s is history and 80s is Stone Age) No one valued cloud computing in the 90s.Simple reason - the 'dot-com bubble'-the Midas touch of the .com prefix. People were too busy creating Google and Yahoo(It's not that straight.There were a couple of other factors too) When the bubble burst, there were a lot of computers to spare.Computers rarely slog at their peak ability. Average figures may be as dismal as 10% of the peak.(Finally,something in common with humans) So, a lot of computing capacity goes untapped.This led to the basic mantra of cloud computing.
Lot of lazy,underused computers + good business acumen = lease their computing ability + mint quick money
This is roughly what happened with Amazon.So, they started EC2(Elastic Compute Cloud).And it created shock waves.A test user for the trial version refused to sell his opportunity for $10,000. That's some fanatic frenzy man!! And since then, Google,IBM,Microsoft and almost all leading universities have joined the bandwagon.
The whole tale is a bit too long. So, let me call it a day. Some parting words---Why wasn't it there before if it is so obvious?? There are a few enabling factors like virtualization and pervasiveness of the net. More on it in the next post. Also look out for description of various architecture levels, security issues, criticisms and the power of platforms and frameworks in the next post.(It will be a bit technical.Can't help it)
In the technical world, only time and/or geniuses can tell a paradigm from a fad.Once it was believed that a computer will overtake a man's intelligence by the 70s and nuclear powered home appliances would see the dawn of the day in the 20th century.Too fired up by science fiction,eh?? Will Cloud Computing see clear skies??Fingers crossed..
PS: If you have read attentively,there will be a sequel post soon...If you got this far,you must be really interested in Cloud Computing.If you want to read more- go for Above The Clouds:A Berkeley View of Cloud Computing and Cloud computing: state-of-the-art and research challenges by some Waterloo guys.
No comments:
Post a Comment