This article is part - 2 of the Cloud Computing Introduction series. To follow this article and understand the scenario based discussion, you are advised to read Cloud Computing Service: A Basic Introduction – 1.
At the end of the previous article I had pointed out something very important, briefly discussed below.
The organization/entity discussed will have a spike in traffic on some particular day or days, i.e., the spike is a bit seasonal in nature. They need a large number of servers to meet that kind of traffic, perhaps 1000 times more than what they would need during normal days, yet it’s not advisable to buy and amass a large number of servers and put them on standby to be used only on those heavy traffic days. This would be a sheer waste of valuable resources. Moreover, only an astrologer can predict when your business will pick up and you’ll need 1000 times more infrastructure/server than you needed in your last quarter! The reverse can also happen, i.e., a recession can strike again and you may need to reduce your infrastructure drastically, because you don’t need it until the government announces a billion $$$ bailout!
In technical terms the competition and economics has lead to a scenario where a business needs the following when it comes to computing as a whole:
● Resource sharing
Note: Curious to know how the above concepts are practically implemented? Read:
Dynamism: It’s quite simple, something like the way you use your mobile phone connection. If you want to talk more, you’ll buy a top-up card (if you are a pre-paid customer like me). If you are a post-paid customer you’ll change your plan to meet your requirement. Your need is dynamic, so should be your infrastructure to support the changing needs.
Abstraction: From an end user’s perspective, they don’t need to care for the OS, the plug-ins, web security or the software platform. Everything should be in place without any worry. The business/consumer should focus more on its core competency rather than worrying about the OS and Software.
Resource Sharing: The whole architecture should be implemented in such a way that provides you the flexibility to share applications as well as other network resources (hardware etc). This will lead to a need based flexible architecture where the resources will expand or contract without any major configuration changes.
There’s one model or style of computing (Gartner's phrase) which satisfies the three requirements mentioned above, and is becoming the technology trend of future. It’s known as Cloud Computing. Let me ask you a simple question. Have you ever used Cloud Computing? Most of you will answer in the negative. Maybe you’ve been hearing the Cloud Computing buzz from the last few months, but you don’t think it has anything to do with you.
But if you are reading this page I can assume that you are web savvy enough to have an e-mail account. That’s it. You are already on the cloud. An e-mail (Not all e-mail services are cloud based; a few viz. Gmail, Yahoo, Hotmail are cloud based) can be the simplest example of an SaaS (Software as a Service). SaaS is a subset of Cloud Computing.
There are various definitions of Cloud computing floating on the web. I found the following to the point and simple:
Common, Location-independent, Online Utility that is available on Demand
--- (Chan, 2009)
The following self explanatory figure describes the fundamental elements of Cloud Computing.
Image Credit, Based on Rayport and Heyward (2009, P4).
Need Based, Real Time Scale Up or Scale Down
So, how is Cloud Computing going to help the entities mentioned in our last article’s examples? The simple solution is IaaS (Infrastructure as a Service). The entities (website owners, in our scenario) can simply use the services of a specialist IaaS cloud provider viz.: Amazon, Rackspace, GoGrid. This model works just like an electricity subscription, in the old days, or a mobile phone or internet usage subscription in modern times. So it’s subscription based, or Pay-as-you-go. If your demand increases, simply ask your cloud provider to add more infrastructure. Pay-as-you-go ensures you’ll never pay anything extra. This results in a happy customer, or, perhaps more appropriately, a customer who is delighted to be saving money.
Your cloud service provider will provide you cloud servers, cloud storage, reliable network and load-balancers on-demand to build and scale up or down your cloud computing infrastructure in real time. Moreover, I scanned through a few SLAs (Service Level Agreements) and I must say they are simply outstanding.
- GoGrid SLA offers 100% uptime, free 24x7x365 support (at no additional cost)
- Amazon S3 SLA offers at least 99.9% Uptime during any monthly billing cycle
Many of them will also offer free Anti-Virus, managed DNS, Hardware Load-balancer and DoS protection. What else do you want? The cloud war is getting more intense every day, with the entry of tech-giants viz. Microsoft and IBM. Google is already there.
Simply focus on your core business, rest assured and leave everything else to the specialist who knows how to manage it best.
Stay tuned, the journey of Cloud Computing has just begun…
Happy Cloud Computing.
Cloud Computing Articles at Techno-Pulse
- Infrastructure as a Service – IaaS: Cloud Computing
- Cloud Computing Platform Introduction – PaaS
- SaaS Introduction with Example – Cloud Service