Thoughts and recommendation on sharing data across session?

[KoderKow]

Hello friends! I have a question that I am having a hard time finding resources for.

• When a user loads up the app, run_app() is executed
• On RStudio Connect you can have multiple users connected to a "session"
• If you only want to load the initial static data once per session, how would go about that?
  • This is beneficial for larger datasets so only the first person who loads up the app has to wait
  • The following users will get instant access due to the previous load on the session
• Generally, I do all my data loading and process in the server, but that is per person
• If I load the data in the run_app() function and globally assign it (example" my_data <<- read_data()), it will make it available in other sessions
  • After this step, I believe run_app() is not executed

I am very interested in your thoughts on this. Finding any discussion about this topic for Shiny, in general, is difficult!

Thanks, appreciate your work!

[ColinFay]

hey @KoderKow,

Sorry for the late reply.

I just pushed this repo that you might find interesting : https://github.com/ColinFay/golemexamples/tree/master/golemloadorder

The idea is to log all the parts of the app, so that we can see how often a piece is executed.

As far as I can tell, if you take the app.R and launch it inside multiple windows, the following parts are only executed once :

• app.R
• run_app.R

> shiny::runApp()
Warning in loadSupport(appDir, renv = sharedEnv, globalrenv = NULL) :
  Loading R/ subdirectory for Shiny application, but this directory appears to contain an R package. Sourcing files in R/ may cause unexpected behavior.
── app.R ───────────────────────────────────────────────────────────────────────────────────
ℹ Loading golemloadorder
── run_app() ───────────────────────────────────────────────────────────────────────────────

Listening on http://127.0.0.1:6827
── app_ui() ────────────────────────────────────────────────────────────────────────────────
── golem_add_external_resources() ──────────────────────────────────────────────────────────
── mod_one_ui() ────────────────────────────────────────────────────────────────────────────
── app_server.R ────────────────────────────────────────────────────────────────────────────
── mod_one_server() ────────────────────────────────────────────────────────────────────────
── mod_one_server.moduleServer() ───────────────────────────────────────────────────────────
── mod_one_server.renderPlot() ─────────────────────────────────────────────────────────────
# NOTE : THIS IS THE MOMENT I OPENED A NEW WINDOW
── app_ui() ────────────────────────────────────────────────────────────────────────────────
── golem_add_external_resources() ──────────────────────────────────────────────────────────
── mod_one_ui() ────────────────────────────────────────────────────────────────────────────
── app_server.R ────────────────────────────────────────────────────────────────────────────
── mod_one_server() ────────────────────────────────────────────────────────────────────────
── mod_one_server.moduleServer() ───────────────────────────────────────────────────────────
── mod_one_server.renderPlot() ─────────────────────────────────────────────────────────────

Also, the rnorm() from app.R & run_app.R are the same for all sessions, the one from the module is different for every session.

So basically, as far as I can tell, everything inside app.R & run_app.R is executed one time by R session, meaning that everything defined here will be shared across all shiny sessions (as long as they happen inside the same R session).

Let me know if that answers your question.

Colin