As individual departments deploy High Performance Computing (HPC) solutions on commodity platforms, incompatible HPC workload managers often begin to proliferate in many large enterprises. To build truly global grids and get the most out of software and infrastructure assets, these discrete workload management solutions need to work as a unified system.

Platform MetaScheduler allows organizations to develop enterprise-wide standards for HPC data centers while still providing individual workgroups and departments the flexibility to deploy the workload management solutions that best suits their needs. Platform MetaScheduler is a practical approach to integrating multiple workload managers. It unifies Platform LSF with other workload management systems, to provide a single job submission, management, monitoring, and reporting interface. This improves asset utilization, while still providing the flexibility of running multiple workload management solutions.

How you can benefit from Platform MetaScheduler:

• Unify incompatible grid systems (Platform LSF™, Sun Grid Engine™, OpenPBS, Torque, and Altair PBSPro™, Microsoft Windows HPC Server and IBM® Tivoli® Workload Scheduler LoadLeveler®) with one interface for management, monitoring, reporting and job submission
• Maximize utilization of assets
• Reduce maintenance & operational costs as well as overall management costs
• Provide unified reporting metrics
• Preserve freedom of choice for departments
• Avoid the cost and complexity of traditional metascheduling approaches

Selected facts about Platform MetaScheduler:

• Allows the Platform LSF scheduler to submit work to foreign schedulers and query and control jobs transparently and securely through a single management interface
• Jobs submitted from Platform MetaScheduler and running on remote clusters are made to look like Platform LSF jobs in all respects from the perspective of both Platform LSF command line utilities and web-based tools interacting with the Platform LSF API