It's a Cloudy Web Ahead

Of late, there has been a lot of buzz around Cloud Computing around the web.

Some people say it's just marketing hype, while others say it's old wine in new

bottle and is the same as distributed computing. Either way Cloud Computing is

on the rise, with biggies like IBM, Google, Amazon, RedHat etc, all launching

Cloud based solutions. It's supposed to change the way you compute. So let's see

what all the noise is really about.

What's the buzz about



During the past few years we have seen massive growth in data and power

hungry applications. We have reached a stage where critical data needs to be

available at all times; one of the reasons why Web 2.0 apps have clicked amongst

enterprises. Most of these applications store and process data on the Cloud

these days.

Simply put a Cloud is a pool of virtualized systems which can be spread

across the world or even in a data center. The term Cloud Computing itself

consists of the platform called the Cloud and applications that are used to

compute on this Cloud. The platform prepares servers dynamically for users as

and when they are needed, and users only need to pay for resources they use.

A user on the Cloud has complete control over its resources. He can add or

delete resources as required in real time, monitor applications running on the

Cloud and perform other management tasks such as Workload balancing, Reimaging,

etc. Similar to this is cloud storage, where data is stored in the Cloud and you

pay according to the storage space you utilize. Cloud Computing can process

several workloads and also supports grid applications.

Elastic Clouds



Amazon is probably the first company to start selling Cloud based services

in the form of its Amazon Elastic Compute Cloud, popularly known as Amazon EC2,

a part of Amazon's web services platform. Currently in its beta version, it

provides computing capacity in the cloud to run applications.

In EC2, Amazon's data centers are used to run applications that require huge

processing power and pay Amazon on the basis of usage. When using EC2, you have

complete control of resources and can choose hardware configurations such as CPU

type, memory and storage required, etc. You can add or delete computing capacity

at any time within minutes, thus providing elasticity, as the name suggests.

EC2 works simultaneously with Amazon S3 service, which provides storage on

Clouds. It is simple to use as well. You just need to create an Amazon machine

image using tools provided by Amazon. The image that contains the application,

configuration settings, OS etc, needs to be uploaded to Amazon S3.

If you don't want to create an image, Amazon also provides template AMI

(Amazon Machine Image), which can be used to run applications. At the moment

Amazon only supports Linux 2.6 kernel systems.

After that you can execute the application using EC2 service and a unique AMI

ID, provided by Amazon to identify your Amazon machine image.

IBM's Blue Cloud data center

In EC2, you have to pay-by-use on an hourly basis. For its smallest instance

it charges $0.10 per hour in which it provides 1 EC2 compute unit, 160 GB of

instant storage and 1.7 GB memory on a 32-bit platform. For data transfer it

charges $0.10 per GB for data transferred onto Amazon EC2 and $0.18 per GB for

first 10 TB of data transferred per month.

Red Hat has also partnered with Amazon to provide ready to use Red Hat

Enterprise Linux with all developed and deployed, existing or new applications.

Blue Clouds



Blue Cloud is an upcoming Cloud Computing offering from IBM which is similar

to Amazon's EC2. However, IBM is targeting enterprise data centers. According to

IBM, Blue Cloud solution will be supported by Tivoli, which will ensure optimal

performance of servers. It consists of Hadoop parallel workload scheduling and

XEN, and PowerVM virtualized Linux OS images. The solution is scheduled for a

spring 2008 launch.

Where can Cloud Computing be used?



Cloud Computing can be very useful for all businesses and applications which

require a huge amount of computing power. At the backend, web applications have

to perform a large amount of data processing. Using Cloud Computing, enterprises

can create a web application and launch them using Cloud without worrying about

the hardware. Cloud Computing also gives scalability advantage, for eg if a web

app running on Cloud suddenly starts to experience a huge workload, more

resources can be immediately provisioned to the application. Similarly, if a

company requires more computing power at a particular time during the day, it

can easy schedule resources for that time.

It's been said since long that the future of operating systems is online, ie

10 or 20 years from now all your files will be stored online. This means, all

processing will be done in Cloud and most applications will be streamed to your

desktops. However, the killer requirement for this remains the bandwidth and how

much people will be willing to pay for these services. Cloud Computing is

already serving as a backend for many web operating systems, now also known as

'Cloud OS'.

Cloud Computing is already being used by university researchers to perform

tests that require huge computing power which was otherwise not possible for

them within their university campuses. IBM and Google have partnered together to

provide a Cloud based research lab to some universities. It's not known whether

or when Google will launch similar services for data centers.

Cloud Computing can be significant for virtual world apps as well. Virtual

worlds such as Second Life are huge and ever growing and require large computing

power. On the flip side, don't be surprised if you see Cloud Computing being

used for cracking encryption algorithms, etc.