---
title: Launching Jupyter Notebook in a Remote Server
tags: data-engineer
---

:::info
Remote server: On-prem Linux Redhat 8
:::

Method: Use SSH, port forwarding mechanism

# Start jupyter lab without opening a browser, 
# run as a background process, using nohup
# dump the output to a specified logfile
nohub jupyter lab --no-browser --port=8889 > ./jupyterlab.log &
# &: allows to get a new prompt?????

# Open log file and get the token
vim ~/tmp/jupyterlab.log

# Check if jupyter lab is still running
ps aux | grep jupyter-lab
# Shutdown the jupyter lab
kill (pgrep jupyter-lab) # Get the Process ID of jupyter lab then kill the process

# SSH tunnel creation in local machine using port forwarding
ssh -L 8000:localhost:8888 username@server_IP

Explanation:
- `ssh -L`: specify given port on the localhost (your local machine) will be forwarded to a given host (your remote server) and port on the remote host.

Two ways accessing jupyter notebook from local machine:
- Using token key given after running `jupyter notebook` in remote server
- Using password setup before running `jupyter notebook` in remote server

## Using SSH and self-signed certificate for HTTPS

# Create a folder to put the certificate related files
mkdir ~/ssl_cert && cd ~/ssl_cert
# Generate a new private key
openssl genrsa -out example.key 2048
# Createa a signed certificate
openssl req -new -key example.key -out example.csr
# This then prompts you to fill out the certificate
# with the country, city, state, common name, and other details

# Create a self-signed certificate
openssl x509 -req -days 365 -in example.csr -signkey example.key -out example.pem

**Setup Jupyter Notebook with Password**

# Create a notebook configuration file
jupyter notebook --generate-config
# Create a password for your notebook
jupyter notebook password

**Find the config file then open it**

vim ~/.jupyter/jupyter_notebook_config.py

Apply the following updates into the config file:

c.NotebookApp.open_browser = False
c.NotebookApp.ip='*'
c.NotebookApp.port=8888

# Add location of the certificate file and private key to the config file
c.NotebookApp.certfile = u'/home/[user]/ssl_cert/example.pem'
c.NotebookApp.keyfile = u'/home/[user]/ssl_cert/example.key'

:::warning
Do we need change settings of Firewall to enable outside access for Jupyter Notebook on the server?

Setup a Firewall rule under VPC network in the Google Cloud Console.
:::

**Start jupyter notebook server**

nohub jupyter notebook --no-browser --port=8888 > ./jupyterlab.log &

**Accessing jupyter notebook**
(More here: [Running jupyter notebooks on remote server](https://towardsdatascience.com/running-jupyter-notebooks-on-remote-servers-603fbcc256b3))

nohup ssh -N -f -L localhost:8888:localhost:8888 username:password@remote_server_ip

Explanation: The `-N` flag tells ssh that no remote commands will be executed. At this point we do not need to execute any remote commands. The `-f` flag pushes the ssh process to the background as mentioned previously. Finally, the `-L` flag specifies the port forwarding configuration using the syntax `local_server:local_port:remote_server:remote_port`. The configuration specifies that all requests sent to port 8889 on the local machine, e.g., your laptop, to port `8889` on the remote machine at `username:password@remote_server_ip`. The `nohup` command has been prepended to silence the output.

**Smoothening Workflow**

Open `~/.bashrc` file:

alias port_forward='nohup ssh -N -f -L localhost:8889:localhost:8889 username:password@remote_server_ip'
alias remote_notebook_start='nohup ssh -f username:password@remote_server_ip "cd rne; . virtual_environment/bin/activate; jupyter notebook --no-browser --port=8889"; port_forward'
alias remote_notebook_stop='ssh username:password@remote_server_ip "pkill -u username jupyter"'

Reload:
`source ~/.bashrc`

Localhost Forwarding: https://phoenixnap.com/kb/ssh-port-forwarding