Upgraded Grid Tool Improves Scalability, Performance

Platform Computing bundles the latest version of its LSF workload management technology with its EGO resource allocation platform.

Platform Computing is looking to ease the management of grid computing environments while at the same time making them more scalable.

The company, of Markham, Ontario, on Nov. 9 unveiled the latest generation of its LSF workload management software, which is now bundled with Platforms EGO (Enterprise Grid Orchestrator) offering.

The move—which decoupled LSF 7 from its resource allocation capabilities and combined it with EGOs ability to virtualize hardware and share IT resources for applications—enables users to increase the scalability of their environments and improve the performance of their applications, said Ashar Baig, product marketing manager for Platforms high-performance computing/LSF business.

Now LSF users will be able to scale from thousands to tens of thousands of nodes, and will be able to complete up to 10 million jobs per day, Baig said. The enhanced software also enables up to 90 percent utilization of the grid.

Another advantage to bundling LSF with EGO is the ability to schedule jobs in a single grid from across multiple remote sites, an important move in situations such as enterprises with remote office or retail businesses, he said. Resources can be allocated dynamically, and LSF 7 ensures accurate service-level agreements.

In addition, the software enables departments or groups within an enterprise to voluntarily and temporarily donate compute resources to the grid if needed, and ensures that those resources can be returned on short notice if needed. Such capabilities are important for businesses that experience periodic spikes in demand.

LSF 7 also introduced the Platform Management Console, a Web-based interface that gives users a central place for monitoring the resources on the grid, including data on the health of those resources, utilization and service levels.

"This way you can have mixed clusters without having a mixed scheduler," Baig said.

He said capabilities like those found in LSF 7 will be key to pushing the adoption of grid computing. Currently research institutions and larger enterprises tend to be the leading users of grid technology, but the adoption in midsize businesses has lagged, Baig said.

"These types of companies will be a target because of the ease of use [of the Platform technology]," he said.


Check out eWEEK.coms for the latest news, views and analysis on servers, switches and networking protocols for the enterprise and small businesses.