This remote Jupyter Notebook tutorial was meant for Users at the SimCenter at UTC, but this should work on any remote Linux server.
- Get a user account with Ethan and set up a VPN so you can ssh onto the server remotely.
- Learn a little bit about managing environments with Anaconda.
Connect to GlobalProtectVPN so ssh works.
local> ssh <user>@lookout00 (could be lookout01, 02, 03 as well)
lookout> export PATH=/usr/local/anaconda2/bin:$PATH
lookout> conda create env_name
lookout> source activate env_name
If not installed already, install Jupyter (and anything else you need - like scipy, tensorflow, keras, etc.)
(env_name) lookout> conda install jupyter
Set a password for your personal Jupyter usage (make it a good but memorable one).
(env_name) lookout> jupyter notebook password
Run a headless version of Jupyter Notebook instance. This will expose Jupyter on port 9000 of the lookout instance you're sshed into.
(env_name) lookout> jupyter notebook --port=9000 --no-browser &
local> ssh -N -f -L localhost:8888:localhost:9000 <user>@lookout00 (again, could be lookout01, 02, 03)
Now, just open up localhost:8888 in your browser and you'll have have a Jupyter Notebook running from the lookout servers.