Definitions and Concepts

We have always been storing the programs and data that we need onto our computer’s hard disk and accessing it whenever required. This is computing. But now technology has taken over and the need to store everything on your physical hard disk is no longer there. Here, Cloud Computing comes into picture. Cloud Computing is the method of computing in which the data and programs are stored over the Internet and not on your hard disk. The Internet is referred to as the Cloud in ‘Cloud Computing’. Similar to yours, others’ and different organizations’ data and programs are also stored over the Internet. When it comes to an organization’s need, the requirements are even more. Servers, applications, Storage, and so on. Creating a physical infrastructure to store and install all these would be time consuming and expensive. Instead, if you get a share of what is already installed over the Internet for use, it is cost effective and meets your needs as well. So Cloud computing is mainly based on sharing the computing resources.

Brief History of Cloud Computing

The evolution of Cloud Computing took a number of years, in different phases. The concept of a global network emerged in the 1960’s. The idea was that everyone be interconnected and would be able to use data and programs from any site located anywhere on the globe. Though the concept was already available, the Internet, in a good bandwidth was available only by the 1990’s. Another key factor which contributed to the quick evolution of Cloud Computing was the emergence of the Virtualization technology. Some important milestones in the evolution of cloud computing were as follows:

1. Salesforce.com, which offered there services on the SaaS platform.
2. The Amazon suite of cloud-based services and Elastic Compute Cloud.
3. Arrival of Web 2.0.

Cloud Computing Terminology

The Cloud Computing terminology is vast and increasing with the advancement of technology. Some of the most popular terminologies are listed in the following table.

5 Main Characteristics of Cloud Computing

The five major characteristics of the Cloud Computing are as follows:

• On-demand self-service - The service of the cloud is available round the clock and provides computing capabilities on-demand of the user automatically.
• Broad network access - Users can access the services via different modes as the heterogeneous thin and thick client platforms.
• Resource pooling - The feature of multi-tenancy where users are assigned resources dynamically, based on demands.
• Rapid elasticity– The service is flexible and can be scaled up or down to suit the business requirements. Resources and programs can be used based on requirement and the user is billed only for the usage.
• Measured service – Usage metering is available and you pay only for what you use. You need not pay for any infrastructure that you do not use.

3 Service Models

There are different service models in Cloud computing. With the advancement of technology, newer service models are evolving. However there are 3 major service models for Cloud Computing. These are Iaas, Paas, and SaaS.

Iaas – In Infrastructe as a Service, users can access the infrastructure required to run their applications, storage, operating systems, etc. The infrastructure can be servers, storage, and virtualization infrastructure and so on. Buying and installing these infrastructure can be very expensive process for the organizations. And more expense would be there on the maintenance part. So the IaaS model is a reliable and cost effective method for the users to run their businesses. For example, if in an organization, virtualization infrastructure is required for a new project with less funding, acquiring the expensive infrastructure will not be possible. Here IaaS is the best option.

Paas – In Platform as a Service, users are provided the platform to develop and run their applications. Changing the platform, upgrading to a newer version, or maintaining uniformity in the platform used by all employees of a company is possible at a less expenditure with PaaS. For example, if you are working on an experimental project where you wouldn’t want to spend more on the development, you can opt for PaaS. This way, you save money and effort by not having the need to buy the platforms and hire people to maintain them. There are platform operators who will help you when in need of help. The user need not manage any of the underlying cloud infrastructure, but can manage the applications required for the work.

SaaS – In Software as a Service, user can access software applications from the cloud. You no longer would need to buy the expensive software. You can access the software through a thin client interface, like a web browser. Similar to the other service models, you cannot control the underlying infrastructure. You can run the applications in the cloud and the cloud administrator manages the portions.

Some of the newer service models are XaaS (Anything as a service), CaaS (Communication as a service), and MaaS (Monitoring as a service).

A user can choose the required model or models, based on business requirements and costs involved.

4 Deployment Models

The cloud services can be deployed in different methods. The deployment model is based on the service model, organizational structure, location, user base, and so on. The four most commonly used deployment models are as follows:

Public Cloud

In this model, the infrastructure is accessible to the public and it is owned by a vendor, who offers the services of the cloud to the users. The cloud vendor shares the cloud resources with the end users. The resource pool is huge and the services are shared by lots of users. The services of this cloud model can be free or available for nominal charges. Google uses a public cloud deployment model. With this model, users need not purchase any infrastructure but can use that of the vendor. A drawback of the public cloud model is that it poses a security threat. If you have very confidential data running in your network, it is not safe to use the public cloud model.

Private Cloud

As the name suggests, this would be a privately owned cloud. Here, the user or organization owns the cloud and only the user or employees of the company have access to the cloud, thereby making data and transactions secure. There is more control over resources when compared to the Public Cloud model. The Private cloud model uses the Virtualization solution and the data centers belong to the company. The major advantage of this model is the security and the control that the users have over the resources and application. However, the drawback is that more financial investment is required and the offering is not as big scale as that of a public cloud model.

Community Cloud

In the Community Cloud model, the infrastructure is owned jointly by different organizations. The organizations may have a similar set of requirements, policies, and customer base. So, they can combine the offerings and make the customer base even bigger. Duplication of same or similar applications and resources are avoided. This model helps reduce the costs, which would otherwise be higher if the organization deploys the Private Cloud model. This is again a classification of the Private Cloud, as it is available to only a certain group of users.

Hybrid Cloud

The Hybrid Cloud deployment model comprises of two or more clouds. This can be a combination of the other three cloud types – public, private, or community. The hybrid deployment is complex compared to the other three owing to the execution and management tasks involved. An example scenario of this model can be where an organization is on the private cloud but there are load spikes which the private cloud cannot handle. For this the organization depends on the public cloud to support the load. The shift from the private to the public cloud and back will be seamless to the end user.