Grid Computing

Building your own infrastructure for grid applications requires capital investment, facilities space, and time — and no matter how big you build it, it may still not be enough. Having to choose between overspending or underprovisioning is a classic dilemma.

By deploying your grid in the cloud, you gain on-demand access to virtually infinite computing resources without the upfront investment. No more scheduling gymnastics, users waiting in line for others' jobs to complete, or constraints on the number of servers available at any given time. In the cloud, it costs the same whether you have 100 servers working for 10 hours or 1,000 working for one hour. You gain access to compute resources to meet peak loads with on-demand scalability — and you choose how fast you want the results.

Real-world examples of grid systems built on RightScale include:

• Pharmaceutical Analysis — Researchers expected a protein analysis comparing 2.5 million compounds to take a week of processing on internal servers. Using hundreds of servers in the cloud, the job completed in one day.
• Insurance Claims Loss Control — Systems for detecting fraudulent, improper, or duplicate claims in batches of millions of claims required months of processing time to run and millions of dollars in capital outlay to build. Using the cloud, these batch runs now finish in a few days.
• Media Rendering and Transcoding — One project involving transcoding images and music to render video on demand reduced processing time from hours on internal resources to minutes in the cloud. Another business supported viral growth of user-generated movies by 7,000 percent and processing capacity in 3.5 days using cloud scalability.

Grid Computing Solution Pack

The RightScale Grid Computing Solution Pack has everything you need to run a complete, scalable grid application environment in the cloud. It includes a flexible grid framework that is automated, resilient to errors, and 100 percent auditable with professional services to help you get up to speed quickly.

Designed to leverage cloud resources, the RightScale pre-configured grid framework processes large numbers of jobs in a scalable and cost-efficient manner. It lets you easily insert your grid "worker" task and define key parameters such as worker array sizes, scaling policies, and storage locations for input and output data.

You can use our grid computing framework as-is or customize it to fit your needs. For example, grid deployments can take input from and deliver results to cloud storage or internal data center storage. Also, grids can be chained together with the results of one step providing the input to another.

Using our Grid Computing Solution Pack, organizations can enable both business users and researchers to easily deploy their own grids in the cloud, supplementing or replacing compute-intensive grid processing jobs that have traditionally run inside companies' data centers. These users won’t have to wait in line to get batch jobs processed, but can launch their own individual grid processes on demand. With virtually unlimited resources, users can get results faster and even take on projects that could not be executed otherwise. For IT, the cloud offers an on-demand infrastructure that enables service to internal clients without having to invest capital for transient peak consumption.

Download the white paper Grid Computing Applications in the Cloud.