Build Your Own Grid And Distributed Computing Networks

Perhaps the best and most quoted example of grid computing is the Seti@Home project, aimed to search for extra-terrestrial life. This project involves using the idle time of as many nodes as possible that are connected to the Internet. If a simple program like this can handle such a mammoth task, then imagine what would happen if a similar concept were applied to different areas. Earlier it may have sounded like something that could only be implemented on supercomputers. But now, things seem to be changing. Different variants of the concept have started appearing, not only for the scientific and engineering community, but also for areas such as enterprise storage and multimedia. These don’t need supercomputers to be implemented. There are both open source and commercial software available for the same. 

	Build A Computer Grid Using Globus This open-source software can distribute jobs from compute-intensive application to clients on the network or the internet through an application server. Storage Virtualization  Free space from computers on your network and aggregate them into a single storage pool Rendering video on a cluster Using Cinelerra’s renderfarm capabilities to render the result of the video being rendered

The Globus project, for instance, is an alliance that aims at enhancing grid computing. It has produced open-source software, which is focused at the scientific and engineering community. You can use the software to build new types of compute-intensive applications. 

Another example is storage virtualization, which involves consolidating all the storage on your network into a single entity. The storage could be on any machine on the network, independent of the platform. The software would dynamically allocate or reallocate storage space to any client that needs it from this pool. This may not be a direct example of distributed computing, but it uses multiple nodes on a network to perform a very useful function in an organization. 

Finally, there’s the multimedia community, which has a never-ending thirst for computing resources, whether it’s for graphics designing, video editing or audio mixing. It can really benefit from using distributed computing, and one classic example of the same is Cinelerra, the open-source video editing/rendering software. It can create a renderfarm, which is a set of ordinary diskless nodes that can take parts of a movie’s rendering operations and assist the mastering machine. 

In this story, we have demonstrated the power of distributed computing using the examples we’ve mentioned above. The software is easily available, and the hardware is ordinary. We’ve given some sample implementations, and leave the rest to you to try out.

Anindya Roy, Geetaj Channana and Sanjay Majumder