A robust, user-friendly SLURM job monitoring utility that handles job name truncation, detects job completion status, and provides intelligent feedback for HPC environments.

pip install --user jobmon-slurm

No admin privileges required! Works on any HPC system with SLURM and Python.

• Handles SLURM job name truncation automatically
• jm dual_env_audit works even when SLURM shows dual_env
• Progressive fallback strategies (exact → shortened → partial)

• Uses sacct to check job history when queue is empty
• Detects quickly-completed jobs that finish before monitoring starts
• Shows job exit codes and completion timestamps
• Distinguishes between completion types

• 🎯 COMPLETED: Job finished successfully
• 🛑 CANCELLED: Job cancelled via scancel
• ❌ FAILED: Job failed during execution
• ⏰ TIMEOUT: Job exceeded time limit
• 🔄 RUNNING: Job currently executing

• Real-time transitions: Automatically moves from queue monitoring to output viewing
• Helpful suggestions: When searches fail, provides actionable alternatives
• Debug mode: jm -d pattern shows detailed search process

jm                      # Monitor all your jobs
jm foldx               # Monitor jobs matching "foldx"
jm 41005547            # Monitor specific job ID

jm dual_env_audit      # Works despite SLURM name truncation!
jm -d pattern          # Debug mode shows search process
jm -q pattern          # Quiet mode (less verbose)
jq pattern             # Short alias for quiet mode

jm -u username         # Monitor another user's jobs
jm -i 5                # 5-second refresh interval
jm --help              # Full help and options

$ jm my_job
# Start monitoring...
# User runs: scancel 41005547
# Monitor automatically detects:
==========================================
🛑 JOB CANCELLED at Wed Jul 23 10:15:42 2025
==========================================
💡 This job was cancelled (likely via 'scancel' command)
   If this was intentional, no further action needed
   If accidental, you may want to resubmit the job

$ jm dual_environment_audit_script
# Automatically tries:
# 1. "dual_environment_audit_script" (exact)
# 2. "dual_env" (SLURM truncated)
# 3. "dual" (shorter pattern)
# Shows which pattern worked in debug mode

• SLURM environment (uses squeue, sacct, scancel)
• Python 3.6+ (for package management only)
• Standard Unix utilities (bash, tail, grep, etc.)

pip install --user jobmon-slurm

git clone https://github.com/yourusername/jobmon-slurm.git
cd jobmon-slurm
pip install --user -e .

# Download latest release
wget https://github.com/yourusername/jobmon-slurm/releases/latest/download/jobmon-slurm.tar.gz
pip install --user jobmon-slurm.tar.gz

We welcome contributions! Here's how to help:

• Use GitHub Issues for bug reports
• Include SLURM version, Python version, and example commands
• Paste error messages and job output

git clone https://github.com/yourusername/jobmon-slurm.git
cd jobmon-slurm
pip install --user -e .[dev]

1. Edit the main script: jobmon_pkg/scripts/jobmon.sh
2. Test locally: pip install --user -e . && jm --help
3. Submit PR: Include test cases and documentation updates

make update-patch    # Bug fixes
make update-minor    # New features  
make update-major    # Breaking changes
make upload         # Publish to PyPI

• National Labs: NERSC, ORNL, ANL, LLNL
• Universities: Stanford, MIT, UC Berkeley, and many more
• Industry: Research teams worldwide

Add your organization! Submit a PR or create an issue.

MIT License - see LICENSE file for details.

• 🐛 Issues: GitHub Issues
• 💬 Discussions: GitHub Discussions
• 📧 Email: Open an issue instead of email for faster response

• SLURM developers for the amazing workload manager
• HPC community for feedback and feature requests
• Python packaging team for making distribution easy

⭐ If this tool helps your research, please star the repository!

jobmon-slurm: Making SLURM job monitoring actually enjoyable 🚀