This skill provides an overview of the scheduling of jobs on a supercomputer. It covers generic and widely used concepts that serve the purpose to maximize the efficiency of a supercomputer.

Batch jobs submitted to a job queue define the workloads in batch systems. A workload manager of a cluster system typically deals with:

• Job Control to provide a user interface for submitting jobs to job queues, monitoring their state during processing (e.g. to check their estimated starting time), and intervening in their execution (e.g. to abort them manually)
• Scheduling and Resource Management to select a waiting job for execution and to allocate nodes to the job meeting all its other demands for computing resources (memory, special processing elements like GPUs, etc.)
• Accounting to record historical data about how many computing resources (e.g. computing time) have been consumed by a job

• Comprehend the exclusive and shared usage model in HPC.
• Differentiate batch and interactive job submission.
• Comprehend the generic concepts and architecture of resource manager, scheduler, job and job script.
• Explain environment variables as a means to communicate.
• Comprehend accounting principles.
• Explain the generic steps to run and monitor a single job.
• Comprehend scheduling principles (first come first served, shortest job first, backfilling) to achieve objectives like minimizing the averaged elapsed program runtimes, and maximizing the utilization of the available HPC resources.
• Comprehend the differences between Batch Systems and Time-Sharing Systems.
• Explain the concepts and procedures for resource allocation and job execution in an HPC environment.
• Run interactive jobs and batch jobs.
• Comprehend and describe the expected behavior of job scripts.
• Change provided job scripts and embed them into shell scripts to run a variety of parallel applications.
• Analyze the output generated from a job scheduler and describe the cause of typically generated errors.
• Comprehend accounting principles (billing for the jobs).
• Comprehend the set of terms for performance criteria like:
  • Resource Utilization.
  • Throughput.
  • Waiting Time.
  • Execution Time.
  • Turnaround Time.
• Comprehend scheduling strategies that increase productivity.
• Comprehend that typical goals of job scheduling are:
  • Maximization of resource utilization.
  • Maximization of throughput.
  • Minimization of waiting time.
  • Minimization of turnaround time.
• Comprehend that there is a variety of scheduling algorithms from rather simple to more complex like:
  • First-Come-First-Served (FCFS).
  • Shortest-Job-First (SJF).
  • Priority-based.
  • Fair-Share.
  • Backfilling.
• Apply advanced scheduling principles (e.g. backfilling) to achieve objectives like minimizing the averaged elapsed program runtimes, and maximizing the utilization of the available HPC resources.
• Discuss sophisticated scheduling principles (e.g. fair share) to achieve objectives like treating the users fair, and maximizing the utilization of the available HPC resources.