Skip to main content

IBM Systems  >  Clusters  >  Software  > 
 

Tivoli Workload Scheduler LoadLeveler


Get the most out of your computing resources
Tivoli Workload Scheduler LoadLeveler® is a parallel job scheduling system that allows users to run more jobs in less time by matching each job's processing needs and priority with the available resources, thereby maximizing resource utilization. LoadLeveler also provides a single point of control for effective workload management, offers detailed accounting of system utilization for tracking or chargeback and supports high availability configurations.

When jobs are submitted to LoadLeveler, they are not necessarily executed in the order of submission. Instead, LoadLeveler dispatches jobs based on their priority, resource requirements and special instructions; for example, administrators can specify that long-running jobs run only on off-hours, that short-running jobs be scheduled around long-running jobs or that certain users or groups get priority. In addition, the resources themselves can be tightly controlled: use of individual machines can be limited to specific times, users or job classes or LoadLeveler can use machines only when the keyboard and mouse are inactive.

LoadLeveler tracks the total resources used by each serial or parallel job and offers several reporting options to track jobs and utilization by user, group, account or type over a specified time period. To support chargeback for resource use, LoadLeveler can incorporate machine speed to adjust chargeback rates and be configured to require an account for each job.

LoadLeveler offers both a command line interface and a graphical interface in addition to an API enabling user-written applications to control it. LoadLeveler also supports high availability configurations to ensure reliable operation and automatically monitors the available compute resources to ensure no jobs are scheduled to failed machines.

Advanced LoadLeveler capabilities and features

  • Easily scalable to thousands of processing nodes and thousands of parallel jobs in the job queue, LoadLeveler is an industrial-strength product that has been available on parallel computing platforms for 15 years.
  • LoadLeveler incorporates the latest technologies in parallel batch scheduling research. An early adopter of the backfill scheduling algorithm, IBM has continued to improve the speed, scalability and performance of this algorithm in LoadLeveler.
  • The first batch scheduler to offer a complete scheduling API, LoadLeveler currently offers unprecedented flexibility with a complete array of APIs allowing a high level of site-specific customization.
  • LoadLeveler is a completely distributed program with extensive failover and self-repair capabilities to survive even severe system events, usually without administrator intervention.
  • LoadLeveler offers job checkpointing and suspension with optional job cancellation, hold and re-queue). These capabilities provide great flexibility in defining real-time job and resource priority control.
  • LoadLeveler integrates with AIX Workload Manager (WLM) to provide both resource specification on job start and resource utilization controls to prevent resource overuse by errant applications.
  • A new LoadLeveler algorithm is available for the Blue Gene machine BG/L that extends LoadLeveler's reliable performance to unprecedented scalability levels (on the order of tens of thousands of processors) within the Blue Gene family of machines.
  • LoadLeveler integrates easily with the IBM High Performance Switch (HPS) for IBM System p hardware, enabling immediate use of the scalable bandwidth and additional multitasking capabilities of the HPS.