5. Job execution

This chapter describes jobs that execute MPI programs.

• 5.1. Overview
  • 5.1.1. About the jobs
  • 5.1.2. Types of job
  • 5.1.3. Create job script
  • 5.1.4. The command for creating template of job script
  • 5.1.5. Job allocation operation
• 5.2. Normal job
  • 5.2.1. Submitting a job
  • 5.2.2. Refer to job execution result
  • 5.2.3. Statistical information
• 5.3. Step job
  • 5.3.1. Job submitting
  • 5.3.2. Step job option
  • 5.3.3. Sub job ID
  • 5.3.4. Referencing job execution results
  • 5.3.5. Job submission example with dependency
• 5.4. Bulk Job
  • 5.4.1. Job submission
  • 5.4.2. Bulk job script
  • 5.4.3. Sub job ID
  • 5.4.4. Referencing job execution results
• 5.5. Interactive job
  • 5.5.1. Job submission
  • 5.5.2. Command format
  • 5.5.3. Command option block
  • 5.5.4. Script file block
• 5.6. Workflow job
  • 5.6.1. Job submission
• 5.7. Master-worker type job
  • 5.7.1. Job Submission
  • 5.7.2. Dynamic generation of worker processes
  • 5.7.3. Worker process creation request to Agent prosess
  • 5.7.4. Worker process generation by pjaexe command
  • 5.7.5. Notes on job creation
  • 5.7.6. Impact when system failure
• 5.8. Environment setting
  • 5.8.1. Tools to reduce search load on dynamic libraries(sort_libp)
• 5.9. Job execution command options
  • 5.9.1. Basic option
  • 5.9.2. Resource specification
  • 5.9.3. MPI option
  • 5.9.4. Job statistical information
  • 5.9.5. Environment variables in job scripts
  • 5.9.6. Environment variables in MPI processes
  • 5.9.7. Specifying an execution start time
• 5.10. Large page space allocation
  • 5.10.1. Overview
  • 5.10.2. Environment variable for large page library setting
  • 5.10.3. Large page library linking
  • 5.10.4. Mitigating Memory Fragmentation
• 5.11. CPU core and memory binding
• 5.12. Displaying a job status
  • 5.12.1. Normal job
  • 5.12.2. Step job
  • 5.12.3. Bulk job
  • 5.12.4. Computer resources
  • 5.12.5. The job statistic information
  • 5.12.6. Job status display command options
• 5.13. Deleting a job
  • 5.13.1. Normal job
  • 5.13.2. Step job
  • 5.13.3. Bulk job
  • 5.13.4. Job delete command options
• 5.14. Notes on job operations
  • 5.14.1. Delete job (pjdel)
  • 5.14.2. Job hold (pjhold)
  • 5.14.3. Unpin a job (pjrls)
  • 5.14.4. Wait for job (pjwait)
  • 5.14.5. Send signal to job (pjsig)
  • 5.14.6. Job parameter change (pjalter)
• 5.15. Use of custom resource
  • 5.15.1. Function overview
  • 5.15.2. How to check
  • 5.15.3. Assign method
• 5.16. Checking when a job ends abnormally
  • 5.16.1. Job manager exit code
  • 5.16.2. Message output during job execution
  • 5.16.3. PJM 0079 ERROR REASON list
  • 5.16.4. CPU clock change (Power capping)
• 5.17. Sample Scripts
  • 5.17.1. Normal job (sequential program)
  • 5.17.2. Normal job (MPI / 1D shape)
  • 5.17.3. Normal job (MPI / 2D shape)
  • 5.17.4. Normal job (MPI / 3D shape)
  • 5.17.5. Normal job (thread parallel (automatic parallel))
  • 5.17.6. Normal job (thread parallel (OpenMP))
  • 5.17.7. Normal job (Flat MPI / 1D shape)
  • 5.17.8. Normal job (hybrid parallel)
• 5.18. Obtaining TofuD TNR statistics
• 5.19. Job list display command affected by the failure
  • 5.19.1. job_events
  • 5.19.2. show_evict_node
• 5.20. Note
  • 5.20.1. How to use pjrsh command
  • 5.20.2. Job Execution in Startup Project (trial)