Getting started

• Logging into Keeling
  • Connecting via ssh on MacOS of Linux with terminal
  • Connecting via ssh on Windows
    • PuTTY with X11 forwarding
  • Connecting via Visual Code Studio
  • Graphical display
• Installed software and packages
  • Compilers
  • GPU (CUDA) Environment
• Editting files on Keeling
  • Traditional text editors
  • Visual Studio Code
  • Connect Python Jupyter Notebook in VS Code
• Computing on Keeling
  • Available compute resources
  • Interactive computing
    • Accessing GPU nodes
  • Batch computing
    • Example scripts
    • Helpful SLURM command line options
  • Access compute nodes directly using SSH
• VS Code Remote Tunnels
  • One-Time Setup
    • 1. Install VS Code Locally (Mac or PC)
    • 2. Install the Remote - Tunnels Extension in VS Code
    • 3. Install the VS Code CLI on Keeling (Linux)
    • 4. Authenticate the Tunnel (First Time Only)
  • Starting a Session (Do This Each Time)
    • Step 1: SSH into Keeling
    • Step 2: Start a Persistent screen Session
    • Step 3: Request a Compute Node via Slurm
    • Step 4: Start the VS Code Tunnel
    • Step 5: Detach from Screen
    • Step 6: Connect from VS Code Locally
  • Reconnecting During the 8-Hour Window
    • From VS Code (Easiest)
    • If You Need to Check on the Tunnel via SSH
  • When the Session Expires
  • Quick Reference: Setup vs. Routine
  • Troubleshooting
• File system
  • Home directory
  • Group file systems
• Managing file transfers
  • sftp
  • scp
  • rsync
  • Globus
    • Globus Connect Personal