In the past several years, grid computing has emerged as a way to harness and take advantage of computing resources across geographies and organizations on a network (usually Internet) to solve large scale computational problems.Computing becomes pervasive and individual users gain access to computing resources as needed with little or no knowledge of where those resources are located or what the underlying technologies, hardware, operating system, and so on are. Grid computing offers a model for solving massive computational problems. intelligence. Using the Internet, SETI has brought together the processing power of more than 3 million PCs from around the world, and has already used the equivalentof more than 600,000 years of PC processing power. Grid Computing – How It works To enter into a grid, one has to egister and then renew the registration every year with a virtual organization which works on a common interest and goals. On registration, a certificate is issued (from a Certifying Authority) which will be installed on the machine and will be used to prove the identity of individual machine. On login, a temporary session will start which remain active for 12 hours for work and to execute a job, it will be defined in Job Description Language (JDL) describing application to use, resources required, data inputs etc. Once a job is defined it is submitted to a Resource Broker (RB) and the RB queries the Information Service & Replica Catalogue (IR&RC) to check resources. The RB uses this information to match the job of a suitable computing element. The RB makes its choice and when it finds the suitable Computing and storage elements with necessary data, it informs the Logging and Book Keeping service of its decision. The RB submits the job to the selected computing element service of its decision. The computing element gets necessary data from storage element. The job is eventually completed in chosen computing element. Evolution of Grid Computing Grid computing has evolved through stages as mentioned below : Meta Computing Meta Computing was a name coined for a particular type of distributed computing, very popular in the early 90’s, which involved linking up supercomputer centers with what was, at that time, high speed networks. Cluster Computing In this type of computing, the PCs communicate with each other in a cluster which can have different sizes. Advantage is of “scalability”: as cluster can grow simply by adding new PCs to it though with certain limitations. Peer to Peer computing ‘Limewire’ is an example of peer to peer computing which helps music lovers to share music files all over the world by downloading piece of software on local machine & helps computers communicate with each other without going through a central server. Internet computing SETI@home project is a virtual “supercomputer” which analyses the data of the Arecibo radio telescope in Puerto Rico, searching for signs of extraterrestrial Who Need Grid Computing Governments & Scientific Organizations for bet- ter disaster management, urban planning, eco- nomic and scientific modeling and similar ac- tivities. Military needs Grid Computing for bet- ter strategic planning and military applications to enhance performance. In the medical world, the coupling of resources (such as administra- tive databases and medical image archives) and specialized instruments (such as MRI machines, CAT scanners, and cardio angiography devices) could enable a wide range of new, computationally enhanced diagnosis procedures based on rapid analysis of complex medical images and automatic comparison with dis- tributed archives to find similar cases. Training & Education E-libraries and e- learning centers could benefit from Grid-based tools for accessing distributed data and creating virtual classrooms with distributed students, resources and tutors. This sort of thing is already in place with the Web, and the Grid will enhance the potential even further. Enterprises and Corporations Large corporations the world over have sites, data, people and resources distributed all over the world. A Grid approach allows the creation of facilities to carry out large scale modeling or computing such as Computer Aided Design (CAD) using resources at many sites simultaneously. Benefits of Grid Computing • Resource Utilization Balancing An organization may have occasional unexpected peak times of activity that demand more resources and if applications are grid enabled, these can be moved to under utilized machines during such times and can thus balance the load for any type of hardware & software resources. · Parallel CPU Usage With scalable application (such as in biomedical, financial modeling, oil exploration, motion picture animation) in which one job is broken into number of sub-jobs, parallel CPU utilization can be achieved. • Virtual resources and virtual organizations for collaboration The users of the grid can be organized dynamically into a number of virtual organizations, each with different policy requirements. These virtual organizations can share their resources collectively as a larger grid where heterogeneous systems work together to a common goal. • Better Resources Utilization & Management Some machines may have expensive licensed software installed that users require. Users’ jobs can be sent to such machines, fully exploiting the software licenses beside better management of precious and sometimes unutilized resources. For further information, mail to ralli.mukesh@nic.in