User Guide

This guide explains how to use Autosubmit to create, configure, run and manage experiments. It is intended for users who want to set up their own workflows, whether on a local machine or on remote HPC platforms.

The sections follow the typical life cycle of an experiment, from creating and configuring it, through defining and running its workflow, to monitoring and managing it. Advanced topics such as provenance, metrics and debugging are covered at the end.

Setting up an experiment

• Create an Experiment
• Configure Experiments
• Experiment IDs

Defining the workflow

• Defining the workflow
• Script templates
• Variables reference
• Wrappers

Running and managing

• Running Experiments
• Job Retries
• Restarting the experiment
• Manage Experiments

Monitoring and output

• Monitor and Check Experiments
• Inspect & Debug Tools
• Log operations
• User-defined metrics
• Performance

Configuration and provenance

• Configuration details, setup and sharing
• Developing a project
• User Mapping
• Provenance
• Traceability

Tutorials (How to)

Step-by-step guides for common tasks:

• Create an Experiment

• Configure Experiments

• Fine-grained dependencies

• Restarting the experiment

• Monitoring job statistics

• Archiving an experiment

• Advanced Configuration