This space is reserved for the TACC Wrangler jobs (to be run in Hadoop or Local mode). Wranlger employs the Simple Linux Utility for Resource Management (SLURM) job scheduler which leverages the tools and commands to perform the most important operations in batch processing: job submission, job monitoring, and job control (cancel, resource request modification, etc.)

Please refer to the sample SLURM scripts for details on the headers and format. Here are few things to keep in mind.

• It is a good practice to use .slurm extension for the job file.
• sinfo -o "%20P %5a %.10l %16F" - Lists the availability and status of queues. The output contains a header (A/I/O/T) for the NODES which means:
  • A: Allocated Nodes
  • I: Idle Nodes
  • O: Other Status for Nodes
  • T: Total Nodes in the partition
• squeue -u - Helps to monitor the jobs. Displays job status and other basic information.
• showq -u - Helps to moniter the jobs. Displays job status and other basic information.
• scontrol show job - Provide more details about the job, its dependencies, working directory, etc.
• sbatch - Schedules the job into SLURM scheduler. From here, the job scheduler takes the lead, look for available nodes and executes the job. If there is no available node, the job show "PD" (Pending) status.
• scancel - Cancels the job for the given job id.