I am setting up my first course with jupyterhub+nbgrader.
I have managed to install it in a AWS EC2 instance, and everything seems fine. This is also my first time using AWS.
I am paying it with my scarce funds so I am planning to run it with the instance type t2.micro (1 vCPU + 1Gb RAM) when not lecturing and burst it to a t3a.large (2 vCPU + 8Gb RAM) around lecture time. This should be a sufficient start for a 15-20-student class. I will burst it up more when I need.
I saw it is possible to automate the process of stoping the instance, change its type, and start it again.
My question is how could I warn/remind the users that are logged in with an active kernel that the system is going down shortly. Of course I can tell them at the beginning of the course that this will happen every lecture, but it would be nice to send them an alert.
Can this be done within JupyterhHub? Maybe set up a cron job or so?
Thanks for any help,