Reg: Jupyterhub for multiple notebooks - Student evaluation

[suc...@successmetric.in]

Is it a good idea to share multiple notebook running on 1 ID or Running multiple IDs and separate notebooks.

Use case:

I have a class of 80 students.

Need to take their test online on Jupyter Hub.

What I can do is build a website and create 80 notebooks running on the single Hub account (Same PAM user)

Or, should I create 80 accounts for each student and ask them to access with their separate login.

I don't wish to make it complex else I have to create scripts to keep adding accounts that I will need to remove once course is done.

Same question paper but each user has notebook like Answer_StudentID.ipynb

[Doug Blank]

In my opinion, you should do it the normal, regular way: use JupyterHub, with each student having a login.

We have a "jupyterhub service" that make it easy to 1) add a bunch of user logins at once, 2) assign them easy to remember passwords, and either 3) print out the passwords/accounts or 4) email the passwords/accounts to them. You can find it here: https://github.com/BrynMawrCollege/jupyterhub/blob/master/accounts.py

There is also a "first use athenticator" that allows them to enter a password on first login, and to use that password thereafter: https://github.com/yuvipanda/jupyterhub-firstuseauthenticator 

Hope that helps!

-Doug

--
You received this message because you are subscribed to the Google Groups "Project Jupyter" group.
To unsubscribe from this group and stop receiving emails from it, send an email to jupyter+u...@googlegroups.com.
To post to this group, send email to jup...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/jupyter/8593a8c4-8693-4c66-9650-c332698aa318%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[suc...@successmetric.in]

Thank you Doug.

For a follow up question, I was just wondering if its a good idea to create a new linux user account when I just need to take say a series of 6 tests.

[Yuvi Panda]

You could also use something like http://tljh.jupyter.org/

[suc...@successmetric.in]

Thanks Yuvi. Will check this out.