Skip to main content

Workspaces and Persistence

Where your files live and what survives container removal.

The Golden Rule​

Save everything important to /workspace/

Everything else in the container is temporary.

Two Filesystems​

When you work in a container, there are two places files can be:

1. Container filesystem (temporary)​

/root/              ← Home directory (temporary!)
/tmp/               ← Temp files (temporary!)
/usr/, /opt/        ← From image (reset on recreate)

Lost when: Container is removed (container retire)

2. Workspace (permanent)​

/workspace/               ← Your project files (permanent!)
├── code/
├── data/
├── models/
└── Dockerfile

Maps to: ~/workspace/<project>/ on the host

Survives: Container removal, system reboots, everything

How It Works​

Your project directory is "mounted" into the container as /workspace:

Host Machine                    Container
────────────                    ─────────
~/workspace/my-project/    ←→   /workspace/

Same files, different path. Changes in one appear in the other.

Common Scenarios​

Saving model checkpoints​

Right (files persist):

torch.save(model, '/workspace/models/checkpoint.pt')

Downloading datasets​

Right (download once):

cd /workspace/data
wget https://example.com/data.tar.gz

Running Jupyter​

Always start from workspace:

cd /workspace
jupyter lab

Quick Reference​

LocationPermanent?Use For
/workspace/YesAll important work
/root/ (home)NoTemporary config
/tmp/NoScratch space

Common Questions​

"Where did my files go?"

Check if they were in /workspace/. If not, they're gone with the container.

"Can I access workspace outside the container?"

Yes. It's at ~/workspace/<project>/ on the host.

"How much space do I have?"

Check with du -sh ~/workspace/*

Want Comprehensive Understanding?​

This is a brief guide covering the essentials. For deeper exploration of:

  • Backup strategies and data safety
  • Disk quota management
  • Advanced mounting options
  • Troubleshooting persistence issues
  • Industry best practices

See Workspaces & Persistence in Educational Computing Context (20 min read).

Next Steps​