Grid Computing

     Grid Computing

Introduction

 Grid computing is a type of networking.  As opposed to  regular networking that focuses on communication between devices, grid networking focuses on tapping unused processor cycles, storage and other resources between many computers to solve complex problems, that can't be solved by single computers.  These networks can contain millions of computers.

Definition: " A computational grid is a hardware and software infrastructure that provides dependable, consistent, pervasive, and inexpensive access to high-end computational capabilities." -- Dr Ian Foster (Argonne National Laboratory & University of Chicago) explains in his article "What is the Grid? A Three Point Checklist".

Grid networks  usually require special software to connect the many computers in the network.  Grid networks are used for projects that involve huge amounts of data and processor cycles. They harnesses unused processor cycles of member PC such as "screensaver time".  The software then assigns a task for the individual PC to perform.

Member PC's naturally are tied together by a network. The internet makes it possible for  millions of PCs to be part of a grid.

The various grid networks handle different types of computing and usually run on different platforms.  Some areas where grid networks are used are life sciences, engineering, energy, manufacturing,  pharmaceuticals, chemicals, energy, engineering materials science and financial services. Grid computing enables the sharing and aggregation of a vast number of geographically diverse computational resources( such as supercomputers, computer clusters,  instruments and people.) 

Uses, Descriptions and Benefits 

Grid computing has many advantages because it lets people share computing power, databases, instruments and other online tools securely over geographical, organizational and institutional boundaries without sacrificing local networks.

Computational tasks are handled from administratively distant locations.  This provides very flexible computing beyond the local domain.  Grid computing can be extended beyond just cpu cycles to data storage, applications, network etc.  An organization can benefit from grid-computing by linking their networks, local and remote, into one huge network that allows it's computational resources to be optimized.  It can also help slow down obsoleteness of individual PC's by encouraging the use of entirely different types of computers into a large virtual computer.

It can save money by providing an organization to a way to use resources that they already have as opposed to constantly buying newer, faster PCs.

Dr Ian Foster further proposes a 3 point checklist for grid networks: 

1) coordinates resources that are not subject to centralized control …(A Grid integrates and coordinates resources and users that live within different control domains—for example, the user’s desktop vs. central computing; different administrative units of the same company; or different companies; and addresses the issues of security, policy, payment, membership, and so forth that arise in these settings. Otherwise, we are dealing with a local management system.)

2) … using standard, open, general-purpose protocols and interfaces … (A Grid is built from multi-purpose protocols and interfaces that address such fundamental issues as authentication, authorization, resource discovery, and resource access. As I discuss further below, it is important that these protocols and interfaces be standard and open. Otherwise, we are dealing with an application specific system.)

  

3) … to deliver nontrivial qualities of service. (A Grid allows its constituent resources to be used in a coordinated fashion to deliver various qualities of service, relating for example to response time, throughput, availability, and security, and/or co-allocation of multiple resource types to meet complex user demands, so that the utility of the combined system is significantly greater than that of the sum of its parts.) 

A grid network can be compared to an electricity grid where power is generated at different locations but users are able to access electricity without worrying about the source of the power

The need for grid networks is actually very obvious because there are many computing problems that are simply too complex to solve by individual computers or networks even with the advances of computing technologies.

Advances in communications technologies and mainly the internet provide one of the main backbones of grid computing.  Internet computing makes it possible to for grid networks to handle many teraflops (floating-point operations) that can be used to analyze petabytes of data.  For instance, enormous calculations can produce many petabytes of information and would need as much as 20 or more teraflops for basic analysis.  A typical supercomputer can handle about 3 teraflops and would be wholly inadequate to complete the task hence the need for more powerful computing solutions like grid networks.

Computing technologies have advanced to the point where most modern PC are as powerful as supercomputers from the early 90's.  Your typical supercomputer today with about 8000 microprocessors is still inadequate for some computations but with 400 millions PC worldwide grid computing is an incredible resource.

The problem of idle un-harnessed CPU cycle has been address as far back as 1951.  In 1985 it was proved  proved that most computers are often idle, and a suggested system to use  those idle cycles for useful work was formulated.  Grid networks can produce more than 400 CPU days of mostly "free"  computing and up to 1 teraflop per second  for use within the network.  

Advantages for your Organization 

So think of it this way.  Got a huge monstrous computation that  involves copious amounts of data ? No problem just set up a Grid network and put to work.  Grid computing networks are already credited with solving complex computations like the the highest prime number 

The main trick in grid computing is by dividing a large computation into many small data sets and then distributing them to individual PC's with a divide and conquer approach. 

Grand challenge problems like climate/weather modeling,  financial modeling, earthquake simulation and protein folding can potentially be solved by grid - computing.  It also offers a way to maximize resource usage within an organization.  Secure authorization of the grid network is vital because it allows a remote user to access the computing resources. 

Here are some categories that you can classify grid computing 

Equipment grids which have a primary piece of equipment e.g. a telescope, and where the surrounding grid is used to control the equipment remotely and to analyze the data produced. 

Data grids or the controlled sharing and management of large amounts of distributed data 

Computational grids (including CPU scavenging grids) which focuses primarily on computationally-intensive operations 

Grid computing can be thought of as a conceptual framework rather than a physical resource.

Grid network are run by one big sophisticated Operating system which divides the processes into datasets that can be solved by the individual PC in the Network, it would help to save time an money for large organizations. Some more advantages of  a Grid computers over supercomputers is that a grid computer is made up a mostly independent computers that can stand alone and the network can be very geographical diverse. 

Interest in grid computing comes not just from the computer sciences but also form scientists businesses, journalists and you too.  You could think of a grid computer as a sort of low cost supercomputer for your organization. 

References

1) Dr Ian Foster (Argonne National Laboratory & University of Chicago) "What is the Grid? A Three Point Checklist"