Use Jupyter notebooks over SSH on Hera, Jet
*If you already know your Local port, you can skip this step
When you successfully log into Hera/Jet, you will see the following text:
Local port 12345 forwarded to remote host.
Remote port 67890 forwarded to local host.
Mark down your Local port number which will be needed in the following steps.
ssh -Y -L12345:localhost:12345 [email protected]
replace 12345 with your actual local port and First.Last with your actual user name.
The RDASApp EVA Python environment provides all the needed modules for pyDAmonitor.
If you don't have a RDASApp copy, you can use the following one:
Hera
module use /scratch1/BMC/wrfruc/gge/RDASApp/modulefiles
module load EVA/hera
Jet
module use /lfs5/BMC/wrfruc/gge/RDASApp/modulefiles
module load EVA/jet
jupyter notebook --no-browser --port=8800
NOTE:
a) 8800 is the Jupyter Notebook server port and it is different from the Local port in steps 1 and 2.
b) Pick up a port number between 8800-8900.
c) When this command completes successfully, it will output a URL address similar as follows which will be needed in step 5.
Or copy and paste one of these URLs:
http://localhost:8806/?token=95fa2d9543d9acd01e8c3c9c82ff6b2cd8df3cd341c53ad6
or http://127.0.0.1:8806/?token=95fa2d9543d9acd01e8c3c9c82ff6b2cd8df3cd341c53ad6
Open another terminal, and run the following command:
ssh -N -f -p 12345 -L 8800:localhost:8800 First.Last@localhost
NOTE:
a. replace 12345 with your actual local port, 8800 with your actual Jupyter Notebook server port, and First.Last with your actual user name.
b. When asked for a password, use your RDHPCS RSA token password.
b. When the command succeeds, it shows no output and gets back to the command line prompt.
It should show something like as follows and you can now run Jupiter Notebooks:
