Hi everyone, I have a question regarding how the processes are run on a shiny server when multiple users use the app simultaneously.
Suppose, for instance, that I have a shiny app that performs machine learning training on the user input data and takes at least 12GB of RAM to do the training and takes half an hour, by average, to be completed. If I deploy the app on my personal server with 16 cores and 32GB of RAM, what would happen if 5 users try to use the app to perform model training simultaneously? Would the shiny app try to perform the training for the most possible tasks (in this case 2 (2*12 = 24 < 32))?
Or would the app crash because of memory overload? If so, what is the best workaround to handle this issue? Is there any way to put the jobs submitted by my shiny app users in the queue?
I would really appreciate any help