Avetec

Advanced Virtual Engine Test Cell

Avetec Case Studies

Platform Computing: Intelligent Job Schedulers

Overview

Platform Computing is a global leader in HPC infrastructure software.  The company delivers integrated software solutions that enable organizations to improve time to results and reduce computing costs. Platform Load Sharing Facility (LSF) software takes advantage of high performance interconnects and hardware-specific functionality on the supercomputers to efficiently manage multi-threaded parallel applications. LSF’s MultiCluster technology has the ability to network distributed sites to make optimum use of the compute resources based on resource availability and business priorities.  More information about Platform can be found on their website at www.platform.com.

Challenge

This project sought to address the challenge of improving management of large data sets in order to enhance job throughput and maximize use of available computing resources.

Objective

The project’s objective was to develop a prototype that extends the current capabilities within Platform’s LSF Grid technology to enable intelligent job placement across multiple LSF clusters and allow for data set location awareness along with data set transfers.

Results

Platform developed the Data Aware Scheduling (DAS) prototype solution for intelligently scheduling work across clusters. The prototype was developed in two phases using DICE’s geographically distributed environment. Platform was able to complete the design necessary for the DAS solution. Continued development is needed to produce a production-quality DAS plug-in.

Ultimate Value

A DAS solution will allow for a more efficient distribution of data sets in a time-effective and cost-effective manner. Adding the DAS plug-in will add to a solid infrastructure to support data intensive applications and improve cluster use across multiple sites when using LSF, MultiCluster and DAS.

 

Case Studies Home