09-04-2011, 10:38 AM
[attachment=11902]
INTRODUCTION
Grid computing is a term referring to the combination of computer resources from multiple administrative domains to reach a common goal. The grid can be thought of as a distributed system with non-interactive workloads that involve a large number of files. What distinguishes grid computing from conventional high performance computing systems such as cluster computing is that grids tend to be more loosely coupled, heterogeneous, and geographically dispersed. Although a grid can be dedicated to a specialized application, it is more common that a single grid will be used for a variety of different purposes. Grids are often constructed with the aid of general-purpose grid software libraries known as middleware.
Grid size can vary by a considerable amount. Grids are a form of distributed computing whereby a “super virtual computer” is composed of many networked loosely coupled computers acting together to perform very large tasks. Furthermore, “distributed” or “grid” computing, in general, is a special type of parallel computing that relies on complete computers (with onboard CPUs, storage, power supplies, network interfaces, etc.) connected to a network (private, public or the Internet) by a conventional networkinterface, such as Ethernet. This is in contrast to the traditional notion of a supercomputer, which has many processors connected by a local high-speed compute bus
Grids enable the sharing, selection, and aggregation of a wide variety of resources including supercomputers, storage systems, data sources, and specialized devices that are geographically distributed and owned by different organizations for solving large-scale computational and data intensive problems in science, engineering, and commerce. Thus creating virtual organizations and enterprises as a temporary alliance of enterprises or organizations that come together to share resources and skills, core competencies, or resources in order to better respond to business opportunities or large-scale application processing requirements, and whose cooperation is supported by computer networks.
PC GRID COMPUTING
The cocept of grid derived from electric power grid and the terms refer to an environment in which various information processing resources(computer,storage,displays,experimental and observations equipment etc) distributed across the network are used as virtual computer.Grid computing aims to provide the necessary aount of processing resource to its operator,on demand. Its potential benefits as follows:
• Collection of distributed processing resources for centralized use.
• Effective utilsation of idle resource.
• Load balancing to eliminate the need to maintain the processing capacity to meet the peak load.
• Ensured fault tolerance for improved reliability
DIVISION OF GRID ACCORDING TO CONFIGURATION
1.COMPUTING GRID
A network of distributed high performance computers(e.g supercomputers) working like asingle huge computer.
2.PC GRID COMPUTING
A concept similar to the computing grid.Collecting the idle CPU power of numerous PC’s to perform large -scale processing.
3.DATA GRID COMPUTING
Making a grid of disk devices and file system that is remotely accessible through the network and works like a large external storage devices.
4.SENSOR GRID
A group of myriad of distributed and network sensors from which data can be collected for specific purpose such as global environment monitoring system.
MECHANISM OF OPERATION
A PC Grid computing work as follows:
1.All participating PC owners download special software from the web server and install it on their PC’s.
2.The special software request to the central server the application programs and the data that each PC is to process as part of the grid.
3.The central server transmits the parallel processing programs and the data to the PC’s divided into packages of appropriate size.
4.The PC’s run the received programs and data during their idle CPU time as their lowest priority task.
5.When the processing is complete ,the special software returns the results to the central server and request new data.(step 3 to 5 are pepat until the entire project is finished)
6.The central server collects and compile the results returned from participating PC’s into the final results.
CLASSIFICATION BY STRUCTURE
• OPEN STRUCTURE
This is the most common type of PC grid computing. This type of PC grid computing is connected to the internet and comprised of Pc owned by individuals who are willing to offer their Pc’s idle processing power. Such project sometimes involves a great many PC’s from a broad range of individuals. Since participation is essentially on voluntary basis, providing an incentive is the key to success. Recognizing that sending goods or real money to individual participants is not practical because of huge delivery cost, project operators are finding other cost-effective ways of rewarding the participants, such as sending electronic money, electronic mileage points or other incentive points over the network, or lottery systems.
Instead of giving such financial incentives ,other grid project operators choose to appeal to people’s volunteer spirit by emphazing the contribution to social welfare, the search for truth and contribution to human advancement. Such programs are sometimes called volunteer computing because participants offer their PCs extra power for free. A famous example is SETI@home, a project operated by university of California, Berkeley. It aims to search for extraterrestrial intelligence based on data collected with a Radio Telescope. More than five millions PC’s voluntarily participate in this project from around the world.The computing power is said to reach 100 TFlops which is almost comparable to the performance of IBM’s Blue Gene/L(140 TFlpos) the worlds largest supercomputer. There are many other open grid project as shown in table.
CLOSED STRUCTURE
PC grids in closed structure are constructed by business enterprises and other organization, based on their existing PC’s. Organisation can have high computing power at low cost, while effective using existing resources. The benefits of creating this type of PC grid include following. Once the organisation decide to launch a project, there is no need to consider incentives for participants: the state of participating PC’s can be monitored and managed with relative ease; and since the each participants ID is known, security risk are risks are better controlled than open structure. To construct a grid using PCs within a single building, an organization can purchase a software package that easily integrates PCs connected through a LAN into a grid.
SEMI-OPEN STRUCTURE
Where multiple organizations that may own many PCs,these organisation can build a single network that extends beyond their boundaries to achieve high computing power. Such a grids would have a semi-open structure and allow public organizations (municipal offices, schools etc.) and local business to jointly provide the local community with shared computing resources. This is a PC grid that makes computing with shared computing resources. This is a PC grid that makes computing resources available at low cost to local small- to- medium companies that hardly afford to use supercomputers .Even large companies can benefit from such a grid because owning expensive super-computers is not always an option. Universities and research institutes in the region also enjoy this benefit. When many regions are making a variety of efforts to enhance their information infrastructure, regionally based grid computing scheme would strengthen these efforts. This is the scheme of ‘of the region, by the region , for the region’.
An example of this type of grid project is a field experiment conducted in Gifu Prefecture in February 2005.Led by Gifu National college of Technology, universities, high schools, education boards, research institutes and other organizations in the prefecture participated in the project, offering over 1000 PCs. The experiment was designed to solve” the traveling salesman problem for 80 muncipalities in Gifu by using by using parallel by using parallel genetic algorithms. After the experiment the institutions involved expressed their expectations for the future if anundant computing resource were to become easily available, including reaserch project that would otherwise not be feasible, such as highly complex simulation .On the other hand, the experiment exposed social issues, such as whether each organisations rules permits its PCs time to be used for the purpose of other than the original intent and how to compensate for the difference in security policy among participating organizations.
INTRODUCTION
Grid computing is a term referring to the combination of computer resources from multiple administrative domains to reach a common goal. The grid can be thought of as a distributed system with non-interactive workloads that involve a large number of files. What distinguishes grid computing from conventional high performance computing systems such as cluster computing is that grids tend to be more loosely coupled, heterogeneous, and geographically dispersed. Although a grid can be dedicated to a specialized application, it is more common that a single grid will be used for a variety of different purposes. Grids are often constructed with the aid of general-purpose grid software libraries known as middleware.
Grid size can vary by a considerable amount. Grids are a form of distributed computing whereby a “super virtual computer” is composed of many networked loosely coupled computers acting together to perform very large tasks. Furthermore, “distributed” or “grid” computing, in general, is a special type of parallel computing that relies on complete computers (with onboard CPUs, storage, power supplies, network interfaces, etc.) connected to a network (private, public or the Internet) by a conventional networkinterface, such as Ethernet. This is in contrast to the traditional notion of a supercomputer, which has many processors connected by a local high-speed compute bus
Grids enable the sharing, selection, and aggregation of a wide variety of resources including supercomputers, storage systems, data sources, and specialized devices that are geographically distributed and owned by different organizations for solving large-scale computational and data intensive problems in science, engineering, and commerce. Thus creating virtual organizations and enterprises as a temporary alliance of enterprises or organizations that come together to share resources and skills, core competencies, or resources in order to better respond to business opportunities or large-scale application processing requirements, and whose cooperation is supported by computer networks.
PC GRID COMPUTING
The cocept of grid derived from electric power grid and the terms refer to an environment in which various information processing resources(computer,storage,displays,experimental and observations equipment etc) distributed across the network are used as virtual computer.Grid computing aims to provide the necessary aount of processing resource to its operator,on demand. Its potential benefits as follows:
• Collection of distributed processing resources for centralized use.
• Effective utilsation of idle resource.
• Load balancing to eliminate the need to maintain the processing capacity to meet the peak load.
• Ensured fault tolerance for improved reliability
DIVISION OF GRID ACCORDING TO CONFIGURATION
1.COMPUTING GRID
A network of distributed high performance computers(e.g supercomputers) working like asingle huge computer.
2.PC GRID COMPUTING
A concept similar to the computing grid.Collecting the idle CPU power of numerous PC’s to perform large -scale processing.
3.DATA GRID COMPUTING
Making a grid of disk devices and file system that is remotely accessible through the network and works like a large external storage devices.
4.SENSOR GRID
A group of myriad of distributed and network sensors from which data can be collected for specific purpose such as global environment monitoring system.
MECHANISM OF OPERATION
A PC Grid computing work as follows:
1.All participating PC owners download special software from the web server and install it on their PC’s.
2.The special software request to the central server the application programs and the data that each PC is to process as part of the grid.
3.The central server transmits the parallel processing programs and the data to the PC’s divided into packages of appropriate size.
4.The PC’s run the received programs and data during their idle CPU time as their lowest priority task.
5.When the processing is complete ,the special software returns the results to the central server and request new data.(step 3 to 5 are pepat until the entire project is finished)
6.The central server collects and compile the results returned from participating PC’s into the final results.
CLASSIFICATION BY STRUCTURE
• OPEN STRUCTURE
This is the most common type of PC grid computing. This type of PC grid computing is connected to the internet and comprised of Pc owned by individuals who are willing to offer their Pc’s idle processing power. Such project sometimes involves a great many PC’s from a broad range of individuals. Since participation is essentially on voluntary basis, providing an incentive is the key to success. Recognizing that sending goods or real money to individual participants is not practical because of huge delivery cost, project operators are finding other cost-effective ways of rewarding the participants, such as sending electronic money, electronic mileage points or other incentive points over the network, or lottery systems.
Instead of giving such financial incentives ,other grid project operators choose to appeal to people’s volunteer spirit by emphazing the contribution to social welfare, the search for truth and contribution to human advancement. Such programs are sometimes called volunteer computing because participants offer their PCs extra power for free. A famous example is SETI@home, a project operated by university of California, Berkeley. It aims to search for extraterrestrial intelligence based on data collected with a Radio Telescope. More than five millions PC’s voluntarily participate in this project from around the world.The computing power is said to reach 100 TFlops which is almost comparable to the performance of IBM’s Blue Gene/L(140 TFlpos) the worlds largest supercomputer. There are many other open grid project as shown in table.
CLOSED STRUCTURE
PC grids in closed structure are constructed by business enterprises and other organization, based on their existing PC’s. Organisation can have high computing power at low cost, while effective using existing resources. The benefits of creating this type of PC grid include following. Once the organisation decide to launch a project, there is no need to consider incentives for participants: the state of participating PC’s can be monitored and managed with relative ease; and since the each participants ID is known, security risk are risks are better controlled than open structure. To construct a grid using PCs within a single building, an organization can purchase a software package that easily integrates PCs connected through a LAN into a grid.
SEMI-OPEN STRUCTURE
Where multiple organizations that may own many PCs,these organisation can build a single network that extends beyond their boundaries to achieve high computing power. Such a grids would have a semi-open structure and allow public organizations (municipal offices, schools etc.) and local business to jointly provide the local community with shared computing resources. This is a PC grid that makes computing with shared computing resources. This is a PC grid that makes computing resources available at low cost to local small- to- medium companies that hardly afford to use supercomputers .Even large companies can benefit from such a grid because owning expensive super-computers is not always an option. Universities and research institutes in the region also enjoy this benefit. When many regions are making a variety of efforts to enhance their information infrastructure, regionally based grid computing scheme would strengthen these efforts. This is the scheme of ‘of the region, by the region , for the region’.
An example of this type of grid project is a field experiment conducted in Gifu Prefecture in February 2005.Led by Gifu National college of Technology, universities, high schools, education boards, research institutes and other organizations in the prefecture participated in the project, offering over 1000 PCs. The experiment was designed to solve” the traveling salesman problem for 80 muncipalities in Gifu by using by using parallel by using parallel genetic algorithms. After the experiment the institutions involved expressed their expectations for the future if anundant computing resource were to become easily available, including reaserch project that would otherwise not be feasible, such as highly complex simulation .On the other hand, the experiment exposed social issues, such as whether each organisations rules permits its PCs time to be used for the purpose of other than the original intent and how to compensate for the difference in security policy among participating organizations.
