What's installed on your Jupyter server?
- JupyterLab 4.6 in a Python virtual environment at /home/jupyter/venv
- A /home/jupyter/not-defterleri folder for your notebooks
- Password sign-in, with only an argon2 hash in the config file
- HTTPS through Caddy on an sslip.io name based on your IP, with the service on 127.0.0.1:8888
- An unprivileged jupyter user and a systemd service called jupyter
- Python 3 and git
- Firewall with only ports 22, 80 and 443 open, so 8888 can't be reached from outside
What size do you need for Jupyter?
JupyterLab's docs don't give a RAM figure, because your data decides what you need: your dataset has to fit in memory. As a rule of thumb, look at 1 GB for learning and small tables, 2 GB for medium data with pandas and 4 GB for larger tables. If memory runs out, the kernel dies and the notebook's variables are lost. For a heavy job, you can raise RAM from your panel and lower it again at the end of the period.
| Use | Size | Price | Choose |
|---|---|---|---|
| Starter | 1 GB RAM · 1 vCore · 25 GB NVMe | $2.58/mo | |
| Medium | 2 GB RAM · 1 vCore · 40 GB NVMe | $4.01/mo | |
| Powerful | 4 GB RAM · 2 vCore · 60 GB NVMe | $6.89/mo |
Who is it for?
- People doing data analysis, visualization and data cleaning
- Anyone who wants long calculations to run without leaving their computer on
- Learners who want to work on Python from a browser
- Developers trying small models on a CPU
What happens after you order?
- Set the size and complete your payment.
- Your server starts and JupyterLab installs itself. This took about 3.5 minutes in our real-server tests.
- On the Overview tab, open the Panel address link and enter the password from the Password row.
- Open a Python 3 notebook from the Launcher screen.
- To install a package, type %pip install pandas into a cell and run it.
With every server
- Daily backups included in the price
- Unlimited traffic
- Full root access
- Scale RAM and CPU up or down later, and grow the disk
- Location: Istanbul · IST-1.
Good to know
- Don't share the password: anyone who signs in can run code in a notebook and commands in the terminal.
- This is a single-user install. The official docs warn that when several people share one server, their commands and files can collide.
- Packages such as pandas and NumPy aren't part of the install, so add them with %pip. They go into the virtual environment and stay after a restart.
- Seturack servers have no graphics card (GPU), so your code runs on the CPU.
- Updates to JupyterLab and your packages are your responsibility. After upgrading, run systemctl restart jupyter as root.
Jupyter questions and answers
How much RAM does a Jupyter server need?
JupyterLab has no official RAM figure, because your data size decides what you need. As a rule of thumb, 1 GB is a good start for learning and 2 to 4 GB for medium data with pandas. If your data doesn't fit in memory, you scale up from your panel.
How do I install a Python package in JupyterLab?
Type %pip install pandas into a notebook cell and run it. Packages go into the /home/jupyter/venv virtual environment and stay after a restart. If you need a system package, connect over SSH as root and use apt.
Does the Jupyter server have a GPU?
No. Seturack servers have no graphics card (GPU), so the code in your notebooks runs on the CPU. Data analysis and small models work on a CPU, while large model training needs an environment with a GPU.
Where are my notebooks, and how do I download them?
Notebooks are in the /home/jupyter/not-defterleri folder on the server. You can upload and download files from JupyterLab's file panel. If you prefer SFTP, you can connect to the server as root.
Can my team share one JupyterLab?
This is a single-user install, so everyone signs in with the same password as the same jupyter user. The official docs warn that commands and files can collide when several people work at once. For a team, a separate server per person is cleaner.
How do I change the Jupyter password?
As the jupyter user, run ~/venv/bin/python -c 'from jupyter_server.auth import passwd; print(passwd())'. Put the resulting hash on both hashed_password lines in ~/.jupyter/jupyter_server_config.py, then run systemctl restart jupyter as root.
How do I connect my own domain to JupyterLab?
Point the A record at your server, replace the old name everywhere in /etc/caddy/Caddyfile and run systemctl reload caddy. Also put the new address in allow_origin in jupyter_server_config.py, then run systemctl restart jupyter.