Virtual Computing Cluster
26 November 2008
Platform Computing's Virtual Computing Cluster is a dynamic resource management solution combining Platform's VMO and VMware infrastructure. This solution allows enterprises to reduce the cost of their compute infrastructure by optimising application performance and dramatically increasing resource utilisation. Virtual Computing Cluster rapidly changes the operating system (OS) running on a compute node, manages how jobs are prioritised and controls which applications are instantiated on a node.
Traditional job scheduling is dependent upon the OS image and application version residing on a given compute node. This limits flexibility to schedule jobs on any available compute node as jobs enter the queue. By adding VMware Infrastructure, Platform's Virtual Computing Cluster solution enables users to run applications with different revision levels and various OS instances simultaneously on the same compute node, and it automatically checkpoints when changes are made to virtual machines which enables increased reliability and priority job queuing.
Virtual Computing Cluster removes the dependency on the installed application version and OS image by leveraging the power of VMware virtualisation. This solution provides guest image and the dynamic resource management capabilities of Platform VMO and Platform LSF to manage the prioritised scheduling of simultaneous jobs over thousands of compute nodes.