euler-cluster-guide

Complete RSL Euler Cluster Guide

This guide provides comprehensive documentation for using the Euler HPC cluster at ETH Zurich as a member of the Robotics Systems Lab (RSL).

📚 Guide Sections

1. Getting Started

• Access requirements and approval process
• SSH configuration and key setup
• Verifying RSL group membership
• Initial setup script

2. Data Management

• Understanding storage locations (home, scratch, project, work)
• Storage quotas and limits
• Best practices for data organization
• Using local scratch ($TMPDIR) effectively

3. Python Environments & ML Training

• Installing Miniconda in the correct location
• Creating and managing conda environments
• Complete ML training workflow example
• Performance optimization tips

4. Computing on Euler

• Interactive sessions for development
• Writing SLURM batch job scripts
• GPU selection and memory management
• Job monitoring and debugging

5. Container Workflow

• Building Docker containers locally
• Converting to Singularity format
• Deploying and running on Euler
• Performance considerations

🔧 Additional Resources

• Scripts Library - Ready-to-use SLURM scripts and examples
• Troubleshooting - Common issues and solutions

🚀 Quick Start Checklist

1. ✅ Get access approval (fill out form or contact Manthan Patel)
2. ✅ Set up SSH keys and config
3. ✅ Run the setup script to verify access and create directories
4. ✅ Install Miniconda in /cluster/project/rsl/$USER/
5. ✅ Test with an interactive session
6. ✅ Submit your first batch job

📞 Support

• RSL-specific issues: Contact your supervisor or Manthan Patel
• General Euler support: ETH IT ServiceDesk
• Guide improvements: GitHub Issues

---

This guide is maintained by the Robotics Systems Lab (RSL) at ETH Zurich

This site is open source. Improve this page.