The sbatch command is a crucial tool for users of the Slurm Workload Manager, a highly scalable cluster management and job scheduling system for Linux clusters. This command allows users to submit batch jobs to the queue for execution. Understanding how to use sbatch effectively is essential for anyone working in a high-performance computing (HPC) environment, as it enables efficient resource management and job scheduling.

Examples:

1. Basic sbatch Script:
Create a simple batch script named myjob.sh:

   #!/bin/bash

   #SBATCH --job-name=myjob

   #SBATCH --output=output.txt

   #SBATCH --error=error.txt

   #SBATCH --time=01:00:00

   #SBATCH --partition=compute

   #SBATCH --ntasks=1

   #SBATCH --mem=4G



   echo "Running my job on $(hostname)"

Submit the job using the sbatch command:

   sbatch myjob.sh

2. Specifying Resources:
You can specify various resources and constraints directly in your script or via command line options. For example, to request 2 CPUs and 8GB of memory:

   #!/bin/bash

   #SBATCH --job-name=resource_job

   #SBATCH --output=resource_output.txt

   #SBATCH --ntasks=2

   #SBATCH --mem=8G



   echo "Running resource-intensive job on $(hostname)"

Submit the job:

   sbatch resource_job.sh

3. Using Environment Modules:
If your job requires specific software or environment modules, you can load them within your script. For instance:

   #!/bin/bash

   #SBATCH --job-name=module_job

   #SBATCH --output=module_output.txt



   module load python/3.8

   python myscript.py

Submit the job:

   sbatch module_job.sh

4. Job Dependencies:
Sometimes jobs need to be run in a specific order. You can set dependencies between jobs using the --dependency option. For example, to run job2\.sh only after job1\.sh completes successfully:

   sbatch job1\.sh

Note the job ID returned by sbatch, then use it to set the dependency:

   sbatch --dependency=afterok:<job1_id> job2\.sh