ACADEMIA

TORONTO, CANADA—To help commercial enterprises automate and better manage complex job scheduling in distributed, heterogeneous environments, Platform Computing Inc., the leader in distributed computing software solutions, today announced Platform JobScheduler 5, a next-generation job scheduler engineered to meet the rigorous demands of distributed computing. Platform JobScheduler 5 is based on an open, Grid-enabled architecture that leverages the strengths of Platform LSF, the industry’s premier workload management solution, to automate, schedule, and manage the job flows and process automation across distributed Linux, Unix and Windows clusters. According to Milind Govekar of Gartner, Inc., “Today’s heterogeneous and complex computing environment demands highly scalable, cross-platform job scheduling products. In addition to the normal job scheduler functionality, there is a need for job schedulers to fill the gap of batch application integration and process automation needs in this ever growing and ever changing mission critical computing environment. Job schedulers also provide an opportunity to consolidate and reduce multiple scheduling points across the enterprise and to effectively manage dependencies and events.” As IT environments become more distributed and diverse, it is becoming increasingly challenging for enterprises to automate and manage workload, particularly mission-critical business applications that rely on high volumes of compute intensive jobs for financial analysis, business intelligence, scientific and business modeling and analytics. A sophisticated, event-driven scheduler, JobScheduler 5 reliably automates the execution of job flows across the enterprise to deliver predictable results. Platform JobScheduler ensures that all scheduled jobs within job flows are distributed across an Enterprise Grid, and reliably processed on available compute resources. “Platform is in the unique position of having an Enterprise Grid computing solution that incorporates a job scheduler,” said Paul Hill, vice president, marketing and business development, Platform. “Unlike other job scheduling solutions, JobScheduler 5 provides sophisticated resource virtualization, optimal resource sharing, enterprise scalability and seamless manageability through clustering. This helps organizations execute hundreds or thousands of jobs across their distributed enterprise resources, while delivering operational efficiencies in business process reliability and manageability.” Key features of JobScheduler 5 include: • Extensible Open Architecture. JobScheduler 5 is built on a robust, open architecture. It supports XML based flow definitions, and through published interfaces, including CLI and Java API, ensures application integration and adaptation with other enterprise software. • Enterprise Event-Based Job Scheduling: JobScheduler 5 provides a flexible scheduling environment where job execution can be contingent upon a number of events. Jobs can be triggered based on job or subflow dependencies; time, calendar and file events; or arbitrary, custom-defined events. • Intuitive Flow Definition and Runtime Management: JobScheduler 5’s graphical user interface makes it easy to define job flows and event dependencies, monitor job flow execution, and create custom calendars. JobScheduler 5 also provides built in exception handlers that automatically correct errors and send alarm notifications in response to failures. • Resource Virtualization. As the first Grid-enabled job scheduler, JobScheduler 5 provides resource virtualization through clustering. It executes job flows through an execution infrastructure, which eliminates the need to designate the specific servers or hosts that each job within a job flow must run on. The benefit of Platform JobScheduler in enterprise environments is clearly illustrated through its integration with SAS/Warehouse Administrator. According to Mark Neille, General Manager IT, SA Eagle, “The ETL jobs used to maintain SA Eagle’s data warehouse necessarily involve a large number of dependencies. Installing Platform JobScheduler has assisted the SAS/Warehouse Administrator functionality, and has allowed us to fully automate the use of SAS/Warehouse Administrator to run dependent jobs. The end result is more efficient job scheduling, and the ability to ‘set-and-forget’ 200 dependent jobs in a single batch window. The only other alternative would have been to write painstaking amounts of custom code; extending the scope of the project, delaying the implementation of the data warehouse, and making strategic data analysis extremely inefficient for SA Eagle.” In addition to JobScheduler, Platform’s workload management solutions include LSF, MultiCluster for Enterprise Grid computing across multiple clusters and geographic locations, Parallel for managing parallel applications, FTA for fast file transfer, reliability and security, and ActiveCluster for Windows desktop computers. These solutions are supported by a global professional services team that provides industry domain experts, customer support, and training courses to ensure that organizations achieve maximum value from Platform’s comprehensive distributed computing software solutions.