Longbow

**Longbow** is a tool for automating simulations on remote HPC machines. It is designed to mimic the normal way an application is run locally but sends simulations to powerful machines, supporting LAMMPS, GROMACS, NAMD, AMBER, and other…

9. FRAMEWORKS 9.2 Workflow Job Management VERIFIED
Back to Directory Official Website

Overview

**Longbow** is a tool for automating simulations on remote HPC machines. It is designed to mimic the normal way an application is run locally but sends simulations to powerful machines, supporting LAMMPS, GROMACS, NAMD, AMBER, and other simulation packages.

Reference Papers

Reference papers are not yet linked for this code.

Full Documentation

Official Resources

  • Source Repository: https://github.com/CCPBioSim/Longbow
  • Documentation: https://longbow.readthedocs.io/
  • PyPI: https://pypi.org/project/longbow/
  • License: Open source (BSD-3)

Overview

Longbow is a tool for automating simulations on remote HPC machines. It is designed to mimic the normal way an application is run locally but sends simulations to powerful machines, supporting LAMMPS, GROMACS, NAMD, AMBER, and other simulation packages.

Scientific domain: HPC job automation, remote simulation management
Target user community: Researchers running MD/DFT simulations on remote HPC clusters

Theoretical Methods

  • Remote HPC job submission
  • Local-like execution model
  • File staging and retrieval
  • Multi-scheduler support (Slurm, SGE, Torque)
  • Job monitoring and recovery

Capabilities (CRITICAL)

  • Remote HPC job submission
  • Local-like execution interface
  • File staging (upload/download)
  • Multiple scheduler support
  • Job monitoring
  • Error recovery and resubmission

Sources: GitHub repository, ReadTheDocs

Key Strengths

Remote Execution:

  • Local-like interface
  • Automatic file staging
  • Multiple schedulers
  • Job monitoring

Multi-Code:

  • LAMMPS, GROMACS, NAMD, AMBER
  • Any command-line application
  • Flexible configuration
  • Plugin architecture

Easy to Use:

  • Simple configuration
  • One-command execution
  • Automatic recovery
  • Logging and debugging

Inputs & Outputs

  • Input formats: Local simulation files, HPC configuration
  • Output data types: Retrieved results, log files

Interfaces & Ecosystem

  • Slurm: Job scheduler
  • SGE/Torque: Alternative schedulers
  • SSH: Remote access
  • Python: Core language

Performance Characteristics

  • Speed: Job management (fast)
  • System size: Any
  • Automation: Full

Computational Cost

  • Framework: Negligible
  • Simulations: Hours (on HPC)

Limitations & Known Constraints

  • SSH required: Need HPC access
  • Configuration: Per-cluster setup
  • No provenance: No data tracking
  • Not DFT-specific: General simulation tool

Comparison with Other Codes

  • vs AiiDA: Longbow is simpler, AiiDA has provenance
  • vs FireWorks: Longbow is remote-focused, FireWorks is database-backed
  • vs Parsl: Longbow is local-like, Parsl is Python-native
  • Unique strength: Local-like remote HPC execution with automatic file staging and multi-scheduler support

Application Areas

HPC Simulations:

  • Remote MD simulations
  • DFT calculations on clusters
  • Batch job management
  • Multi-node execution

Multi-Code:

  • LAMMPS on HPC
  • GROMACS on HPC
  • Any CLI application remotely

Best Practices

Setup:

  • Configure HPC connection
  • Set up scheduler template
  • Test with simple job
  • Use recovery mode

Community and Support

  • Open source (BSD-3)
  • PyPI installable
  • CCP BioSim maintained
  • ReadTheDocs documentation

Verification & Sources

Primary sources:

  1. GitHub: https://github.com/CCPBioSim/Longbow

Confidence: VERIFIED

Verification status: ✅ VERIFIED

  • Source code: ACCESSIBLE (GitHub)
  • PyPI: AVAILABLE
  • Specialized strength: Local-like remote HPC execution with automatic file staging and multi-scheduler support

Related Tools in 9.2 Workflow Job Management