IBM’s Blue Cloud Distributes Load

The term cloud computing describes a system in which IT services within an enterprise have a similar structure to the Internet and are used in the same way. In line with this model, data centers should be organized as distributed and globally accessible resources. Based on IBM’s model, large-scale IT systems will be connected and provide a pool of IT services that can respond flexibly to changing load and demand.

According to an IBM statement, Blue Cloud is based on Open Source software and uses open standards. It will use Xen and PowerVM-virtualized Linux operating system images for virtualization. IBM’s own Tivoli software management product will handle server management. An additional, integrated software solution will help to distribute load dynamically over multiple servers, and automate resource assignments. The infrastructure model for Blue Cloud was created at IBM’s Almaden Research center and is the result of many years of experience with projects in highly-scalable environments.

IBM has the support of the Vietnamese Ministry of Science and Technology (MoST), where a cloud computing infrastructure has been deployed in the scope of a pilot project. The first Blue Cloud offerings will become available in spring 2008, initially for IBM Blade Centers with Power and x86 architectures. An offering for a System Z Mainframe will become available in 2008; according to IBM it will support a “very large” number of virtual systems. IBM is convinced that there is substantial demand for cloud computing, based on the increasing use of service-oriented architectures (SOA) and Web 2.0 applications.

Linux distributor Red Hat and Online traders Amazon are currently beta testing a similar model dubbed "Red Hat on Amazon Elastic Computing Cloud". Enterprises will be able to rent computing capacity and software as a service by end of year. Projects such as SETI@home are prime examples of informal computer networks.