Hi everyone! I have a question about JupyterLab’s capabilities. I would like to use JupyterLab for students at Title 1 schools to run some python code for an engineering class we are offering. Might any of you know if there is a maximum number of users that try-jupyter will allow from a single IP address?
Thank you in advance for your guidance!
If you are actually just using the current offering of JupyterLab put forth on the Try Jupyter page (and the tile to the right for Jupyter Notebook), the kernels do not run on remote computers. The code is actually running in a virtual machine inside the browser of the connecting computer using Web Assembly (WASM). You can tell because in the upper right corner you’ll see ‘Python (Pyodide)’. Pyodide is one of the Python-supporting WASM-based kernels available.
So my understanding is there is no limitation because you are just fetching the files needed to run the Jupyter product. It is just looking at a web page essentially and those pages get a lot of hits on a daily basis and so I don’t think it would be limited in any way.
You should though make sure this is suitable for your needs though. Please consult the paragraph right above the tiles on the Try Jupyter Page that warns the following:
"
Experimental
You can read more about JupyterLite at the first link in that and be sure to note the ’ 
Status ’ section highlighted with the lightning bolts for good reason.
It is more likely though you want to have things set up with packages pre-installed and so you are instead you may be looking at using MyBinder with custom configuration files in your own repo specifying what environment to launch with in the session on a remote virtual machine or a JupyterHub such as the Littlest JupyterHub.
MyBinder has limitations on concurrent connections covered at 'Maximum Concurrent Users for a Repository’and that may have been what you wanted to ask about?
There may still be an option if you want to have packages pre-installed and JupyterLite is enough…
If you are lucky and the packages you need can run in WASM, you can pre-install them and still use JupyterLite using Xeus Python. See here and here for more about it. That latter link is to the demo that includes a guide how to deploy it yourself from GitHub. I did my own deployment here a couple of years ago. You can try it by clicking ‘try lite now’ there. (Bear in mind that was based on a two-year old version of the demo; I tested things and they seem to work at this time though.)
thank you so much for your response! this is so helpful!