-
Notifications
You must be signed in to change notification settings - Fork 3
Home
I teach 1-3 day workshops with 10-40 people and getting all the software working is often a stumbling block and hinders students learning and progress. I am experimenting with using a JupyterHub on AWS to provide a virtual coding platform.
I was inspired by this post by someone who uses a JupyterHub for similar kinds of workshops. https://outpw.github.io/tutorials/JupyterHubSetup.html Follow the instructions in the link above to set up your JupyterHub on AWS. My notes give my experience following the instructions and how I customized my hub. Note, in take 2 I tried to install TLJH on a t2.small instance (not in the Free Tier). This didn't work. I haven't figured out what I need to change yet to get it to work.
Cost: The free tier, t2.micro, should be fine for set-up. I think you can leave it on without cost since you get 750 hours per month free. For a recent workshop, where we used t2.small and t2.medium, the cost came out to $100 per day for 65 users. But that was on 2i2c.org where each user got their own instance. In the TLJH, it is on one server and the users share an instance.
- How to install TLJH directly on an instance bigger than t2.micro. I was able to install on t2.micro and then once installed, I could upgrade my instance to something bigger.
- Why won't
xarray-spatialinstall? I figured out that this needs more than 1G RAM (t2.micro) to install. I was able to install after I upgraded my t2.micro instance to t2.small.