MyQueue

MyQueue is a lightweight, frontend-agnostic task scheduler and workflow manager. It is designed to manage high-throughput calculations on local machines or HPC clusters (SLURM, PBS, etc.) with a simple command-line interface. Unlike comp…

9. FRAMEWORKS 9.2 Workflow & Job Management VERIFIED 1 paper
Back to Mind Map Official Website

Overview

MyQueue is a lightweight, frontend-agnostic task scheduler and workflow manager. It is designed to manage high-throughput calculations on local machines or HPC clusters (SLURM, PBS, etc.) with a simple command-line interface. Unlike complex workflow engines, MyQueue focuses on simplicity, using a folder-based structure where folders represent tasks.

Reference Papers (1)

Full Documentation

Official Resources

  • Homepage: https://myqueue.readthedocs.io/
  • Documentation: https://myqueue.readthedocs.io/
  • Source Repository: https://gitlab.com/schmidt-group/myqueue
  • License: GNU General Public License v3.0

Overview

MyQueue is a lightweight, frontend-agnostic task scheduler and workflow manager. It is designed to manage high-throughput calculations on local machines or HPC clusters (SLURM, PBS, etc.) with a simple command-line interface. Unlike complex workflow engines, MyQueue focuses on simplicity, using a folder-based structure where folders represent tasks.

Scientific domain: Job scheduling, simple workflows, high-throughput
Target user community: ASE/GPAW users, computational physicists

Capabilities (CRITICAL)

  • Folder-based Tasks: A folder with a script (e.g., calculate.py) is a task.
  • Dependency Tracking: Dependencies defined by folder hierarchy or explicit links.
  • Scheduler Agnostic: Unified interface for SLURM, PBS, LSF, and local execution.
  • Command Line: Simple commands (mq submit, mq list, mq kick).
  • Python Integration: Works well with ASE scripts.
  • Restart: Easy restart of failed tasks.

Sources: MyQueue documentation

Inputs & Outputs

  • Input formats: Python scripts inside directories
  • Output data types: Standard output/error logs, data files

Interfaces & Ecosystem

  • ASE: Often used in conjunction with ASE scripts
  • GPAW: Commonly used by the GPAW community

Workflow and Usage

  1. Create folder structure: simulations/run1/
  2. Place script run.py in folder.
  3. Run mq submit simulations/run1/
  4. Monitor with mq list.
  5. If failed, fix and mq resubmit.

Performance Characteristics

  • Very low overhead
  • Simple file-system based state
  • Ideal for personal high-throughput management

Application Areas

  • Managing batches of DFT calculations
  • Parameter sweeps
  • Simple dependency chains (relax -> band structure)

Community and Support

  • Developed at DTU Physics (Jakob Schiøtz and contributors)
  • Used by CAMD/Thygesen groups

Verification & Sources

Primary sources:

  1. Documentation: https://myqueue.readthedocs.io/
  2. GitLab: https://gitlab.com/schmidt-group/myqueue

Confidence: VERIFIED

Verification status: ✅ VERIFIED

  • Website: ACTIVE
  • Documentation: AVAILABLE
  • Source: OPEN (GitLab)
  • Development: ACTIVE
  • Applications: Simple job management, folder-based workflows

Related Tools in 9.2 Workflow & Job Management