分类：
2009-04-23 12:20:23
"Cloud Computing," to put it simply, means "Internet Computing." The Internet is commonly visualized as clouds; hence the term “cloud computing” for computation done through the Internet. With Cloud Computing users can access database resources via the Internet from anywhere, for as long as they need, without worrying about any maintenance or management of actual resources. Besides, databases in cloud are very dynamic and scalable.
Cloud computing is unlike grid computing, utility computing, or autonomic computing. In fact, it is a very independent platform in terms of computing. The best example of cloud computing is Google Apps where any application can be accessed using a browser and it can be deployed on thousands of computer through the Internet.
Cloud computing is cost-effective. Here, cost is greatly reduced as initial expense and recurring expenses are much lower than traditional computing. Maintenance cost is reduced as a third party maintains everything from running the cloud to storing data. Cloud is characterized by features such as platform, location and device independency, which make it easily adoptable for all sizes of businesses, in particular small and mid-sized. However, owing to redundancy of computer system networks and storage system cloud may not be reliable for data, but it scores well as far as security is concerned. In cloud computing, security is tremendously improved because of a superior technology security system, which is now easily available and affordable. Yet another important characteristic of cloud is scalability, which is achieved through server virtualization.
In a nutshell, cloud computing means getting the best performing system with the best value for money.
Cloud computing architecture, just like any other system, is categorized into two main sections: Front End and Back End. Front End can be end user or client or any application (i.e. web browser etc.) which is using cloud services. Back End is the network of servers with any computer program and data 
Cloud has centralized server administration system. Centralized server administers the system, balances client supply, adjusts demands, monitors traffic and avoids congestion. This server follows protocols, commonly known as middleware. Middleware controls the communication of cloud network among them.
Cloud Architecture runs on a very important assumption, which is mostly true. The assumption is that the demand for resources is not always consistent from client to cloud. Because of this reason the servers of cloud are unable to run at their full capacity. To avoid this scenario, server virtualization technique is applied. In sever virtualization, all physical servers are virtualized and they run multiple servers with either same or different application. As one physical server acts as multiple physical servers, it curtails the need for more physical machines.
As a matter of fact, data is the most important part of cloud computing; thus, data security is the top most priority in all the data operations of cloud. Here, all the data are backed up at multiple locations. This astoundingly increases the data storage to multiple times in cloud compared with a regular system. Redundancy of data is crucial, which is a must-have attribute of cloud computing.
Google Apps., Salesforce.com, Zoho Office and various other online applications use cloud computing as Software-As-Service (SAAS) model. These applications are delivered through browser, and multiple customers can access it from various locations. This model has become the most common form of cloud computing because it is beneficial and practical for both the customers and the services providers. For customers, there is no upfront investment and they can Pay-As-They-Go and Pay-As-They-Grow. On the other hand, the service providers, can grow easily as their customer base grows.
Aamzon.com, Sun and IBM offer on-demand storage and computing resources. Web service and APIs enable developers to use all the cloud from internet and allow them to create large-scale, full-featured application. Cloud is not simply limited to providing data storage or computing resources, it can also provide managed services or specific application services through web.
Security of confidential data (e.g., SSN or Credit Card Numbers) is a very important area of concern as it can make way for very big problems if unauthorized users get access to it. Misuse of data can create big issues; hence, in cloud computing it is very important to be aware of data administrators and their extent of data access rights. Large organizations dealing with sensitive data often have well laid out regulatory compliance policies. However, these polices should be verified prior to engaging them in cloud computing. There is a possibility that in cloud computing network, sometimes the network utilizes resources from another country or they might not be fully protected; hence, the need arises for appropriate regulatory compliance policies.
In cloud computing, it is very common to store data of multiple customers at one common location. Cloud computing should have proper techniques where data is segregated properly for data security and confidentiality. Care must be taken to ensure that one customer’s data does not affect another customer’s data. In addition, Cloud computing providers must be equipped with proper disaster recovery policies to deal with any unfortunate event.
A good service provider is the key to good service. So, it is imperative to select the right service provider. One must make sure that the provider is reliable, well-reputed for their customer service and should have a proven track record in IT- related ventures. The Cloud Computing Incidents Database (CCID) records and monitors verifiable, noteworthy events that impact cloud computing providers. Visit the following Wikipedia link to obtain the list all such events.
Comparison has often been drawn between Relational Database and Cloud Computing. They are related for sure but they should not be confused for being the same thing. In actual fact, they are not really competing with each other. There are some unique requirements of applications when they do not call for any advance query techniques but rather need fast access to database. In such scenarios cloud computing should be used. In cloud, data is stored across myriad geographic locations, and processing data from different geographic database leads to delay in receiving data. In case of applications where there is the need of processing huge database using complex queries, traditional relational database is best suited. Cloud has its limitations. As for now, it only supports distributed computing; transactional operations are not currently supported in cloud computing.
SummaryCloud Computing is the next big thing in the arena of computing and storage. There are some concerns about security and its availability. However, different service providers are coming up with various solutions and suggestions in response to customers’ concerns. In any case, cloud is getting bigger and better, and as long as they are available through web services, without capital infrastructure investment at reasonable price, it is for sure going to proliferate and create robust demand in times to come.
While writing this article, I really enjoyed reading Cloud Computing Manifesto ( ) - public declaration of principles and intentions for cloud computing. Manifesto suggests 10 principles of cloud computing, namely User centric, Philanthropic, Openness, Transparency, Interoperability, Representation, Discrimination, Evolution, Balance, and Security.
