Cloud Computing. What it is and why it is important
March 14, 2009 5 Comments
In a relative short period we have seen a sharp rise of online applications completely being operated through your browser. The latest developments in web technology such as Java, Ruby on Rails and AJAX make websites interactive as never before. Almost daily I discover sites which look profoundly better than an average desktopapplication, even -dare I say it- Mac applications. Many of these sites depend on a technology called cloud computing.
Now instead of trying to explain in text what cloud computing is I suggest you look to the below video.
So basically cloud computing is the flexible use of a bunch of servers and in combination with a clever idea you can create some innovative applications. For example, I am currently typing this blogpost on wordpress.com in a web application which looks suspiciously similar to a standard word processor. But I don’t have to save anything to my hard disk, it saves it automatically to the cloud. I can upload pictures to the cloud which I can embed in my blogpost, and if I exceed a certain storage amount I just pay for more storage.
Other examples of cloud computing we often use:
Obviously there are hundreds, if not thousands of other applications.
The interesting thing about cloud computing is that it is a disruptive innovation. Why? Because it tears open existing processes in application development and business processes.
With the proper cloud computing service, knowledge about web technologies and a credit card almost anybody can start a service. No need to pay for expensive computers or IT-related stuff. Just buy storage and a platform and start building. A good example of that is the app development program of Salesforce.com.
At the same time that is the biggest fear for corporate IT departments. Suddenly John Doe doesn’t need the nerds anymore. He can just start using applications without the involvement of corporate IT simply because nothing needs to be installed on the computer. Prohibiting installation of ‘strange’ software on the company computer was always the ‘coupe-de-force’ of the IT department. Not anymore. And they don’t like that, at all.
Don’t get me wrong, I think that a certain level of standardization on IT-related matters within organizations is absolutely necessary. You just don’t want people to download all kinds of stuff and consequently bothering the helpdesk with questions about obscure error messages caused by even more obscure apllications. However, I do think that an IT department should be much, much more proactive in looking for ways to improve the way we work with computers and how the can actually enrich our productivity and collaboration. And frankly, I’ve experienced in most cases that IT is more like a burden to new developments rather than an added value.
But cloud computing will survive and will change the computing landscape yet again. What we are seeing now is just the beginning. If you think logically you just know that a) the internet is getting faster and b) web technologies will get more and more features. Suddenly it’s not so strange to understand why Google decided to develop their own web browser. The OS is getting generic, the future of user interface is in the cloud and hence on your browser. It’s also not so strange to realize why the netbook is the only category growing within the hardware segment. Because people don’t need computer power. People need connectivity and portability.
To finalize I would like to blundly copy and paste an interview with James Staten, Principal Analyst at Forrester which I found on MeetTheBoss. He outlines the findings of his research into cloud computing to find out why it is such a compelling business proposition.
Can you outline the finding of your research into cloud computing?
James Staten. The research we’ve done so far into cloud computing has specifically looked at infrastructure offerings, this means they’re not a SAS offering, they are a platform or an infrastructure upon which you can build and deliver your application. Our cloud computing research found that the majority of the users of these platforms are startups, small businesses, and in some cases interactive agencies or consultants who build specific projects or test bed projects for enterprises.
In doing this analysis what we also found is that there are some people in enterprises who are looking at or using cloud computing without the knowledge of their IT department. And the reason they’re doing this is because there’s a tension point that has risen and been present for quite some time between the IT developers in the business units and central IT, and the tension point comes around the availability of resources for experimentation and for ad hoc projects.
Typically the IT department needs a plan, they need a budget, they need some timing ahead of time in which to allocate the resources, and that simply doesn’t work in the business unit developer’s mind where he needs to have new offerings available and experimenting with new things right away.
Do you think that’s going to change?
JS. Well, what’s happening is the IT departments are starting to respond to this and they’re trying to figure out how they can better accommodate these needs, as any good service-oriented organization should do. Their attempt initially is going to be an internal cloud which is an effort to try and take either their existing virtualization infrastructure, like VMware based infrastructure and make it available to those developers for these types of ad hoc projects.
The challenge in doing that is that most IT departments are under significant budget constraints. They don’t have available free resources and so it’s very hard for them to build a very sizable cloud that could accommodate a lot of these applications. As a result, what they would like to find is a way to either accommodate those needs within their own walls some way, and they’re looking to the vendor community to try and help them with this, or they want the ability to have their small internal cloud flex into a larger public cloud. Their concern with that is that they’re concerned that the large public clouds don’t have the necessary security and the necessary procedures to meet their business needs.
And why, in your opinion, is cloud computing such a compelling business proposition?
JS. Well, the reason that cloud computing has had so much attention is that it provides a dramatically lower price point to get started with an infrastructure deployment. So if you are two guys in a garage, for example, and you want to create a new business from scratch and you don’t have a lot of money, you can actually create the business, have it live worldwide on the internet and across multiple geographies with just a credit card. And that’s what’s so compelling. Not only can you do it as a business example, but if you also have just a short-term project and you need access to 100 servers for one day, you can do that for an incredibly low price. You can do that for $150, for example.
That’s really what’s so compellingly different about cloud computing. In most of the service provider offerings that were available prior to this you had to sign up for a minimum of a six month contract, if not a year contract, you had to know what amount of infrastructure you were going to need to consume. There really wasn’t a play area or an experimentation area, and that’s really what cloud computing has provided.
So we’ve touched on a few, but what are the main obstacles for the enterprise adoption of cloud computing?
JS. Well, right now the biggest obstacle to the adoption of cloud computing is your level of knowledge as an engineer or developer. If you’re going to use the low-end platforms like Amazon compute cloud, you can’t just know your application. You need to know middle ware, you need to know low balancing, you need to know scalability and availability services, and you need to know how to configure and set up the OS the way you want it. That’s a lot of work for a developer who really just wants to concentrate on his cloud computing application, so to fill that gap we’re starting to see some integrators that are coming in that have sort of a fast on ramp for these types of developers where they take care of all the infrastructure for them.
If you’re a developer that’s using one of the higher end platforms, then all of that’s taken away for you. If you just know Ruby on Rails, for example, you can use Google App Engine right now. You don’t have to have any knowledge beyond Ruby on Rails to accommodate that. For Force.com it’s mostly the same, but it’s more open to different development platforms. So whether you’re building a .NET or Java or Ruby or whatever you can run it, build an application and run it on Force.com today and you don’t have to have a lot of the knowledge of the things below.
Where you start getting into trouble is if you’re not all that experienced you may realize that, yeah, you can build your application and take advantage of the services they have, but if you decide you want to move off of their platform you might find that very, very difficult.
I’ve read that there’s been some resistance to cloud computing; why do you think this is?
JS. Well, that’s because a lot of the clouds don’t let you choose, nor do they let you understand what the infrastructure is that they’re basing you upon. So a lot of these clouds, they provide you with a virtual machine or they provide you with this abstracted platform, but if you have questions about well, exactly what hardware is it running on? Exactly what version of the operating system is underneath? What’s the exact hypervisor? What’s your security process? What do you do to back things up?
Often the answer is we don’t disclose that and for a lot of businesses that’s a very uncomfortable answer. So we’re seeing more and more that they’re starting to open this up and they’re starting to tell companies, either under nondisclosure or in private agreements, about what the infrastructure looks like so that they can get more comfortable with it, and in turn use cloud computing. But there’s actually a really good reason for why a lot of these service providers don’t want to tell the world what infrastructure they’re based upon. The reason is because they want the opportunity to change it.
They’re providing you a layer of abstraction, which at that layer they will make very clear APIs, they will make a standard platform, and that’s what you care about. But everything below that they want the right to change out.
So where do you see cloud computing heading in the next six to 12 months?
JS. In the next six to 12 months we expect to see significantly more XSPs offering cloud computing offerings. We expect to see two or three more super sized vendors provide a cloud computing offering based upon their large infrastructure they’ve built for their own web services. So that could be a Yahoo, a Microsoft, an eBay, one of those class of companies could jump into this cloud computing space. We expect to see the hypervisor platform vendors get into this space as well. In fact two of the major ones have already made announcements in this regard recently: VMware and Citrix both announced cloud computing initiatives. They don’t have a platform yet for the service provider to buy, but they’re starting to put those together.
We’ll see those platforms come to market and I suspect that we will also see a fair number of enterprises who build out their internal clouds. It’ll be, like I said, experimental. They’ll be short-term, but they won’t have the flex capability yet, that flex capability to tap into a public cloud and what they would like to call a virtual private cloud probably won’t come until beyond that six to 12 month horizon.
And beyond the 12 months?
JS. Beyond the 12 months I fully expect that we will see a big menu of services from your preferred service provider that taps into software and services offerings, persistent hosting, and cloud computing offerings. And they’ll probably have two variations of cloud computing offerings: they’ll probably have a low end do-it-yourself one that they offer themselves, which will be the highest margin for the service provider, and then they’ll have a relationship with one of these large super clouds that is available to you, that they resell to you.
Now of course you can of course bypass your local service provider and do this yourself. A lot of companies will do that who have very strong IT expertise internally, but those that use a lot of outsourced resources will rely on these service providers to provide them that service plus other services.
One other thing that we’re probably going to see in this, I suspect this will happen in the next 12 months is I think that all of the hype around cloud computing is actually going to diminish. The reason it’s going to diminish is because right now the expectations of change that the market believes cloud computing will drive are getting out of whack. So I suspect in the next 12 months we’ll see a lot of people say, “You know I see cloud computing is here, but where are the profits? Where are the big customers? Where are the big wins? Where is the proof that productivity is revolutionized by this?”
None of those questions are realistic, yet they’re out there and so when we come back in six to 12 months with the answer being no to all of those questions, the bubble’s going to burst and a lot of people are going to start saying, “Oh, cloud computing was a flash in the pan, it wasn’t as big of a deal as we all thought,” and that’s when it’s going to get really interesting because at that point is when you’re going to see people that don’t care about the hype, but are investing in cloud computing, are going to start turning on big productivity gains. But it’s going to take six to 12 months beyond that point to be visible to the market.