Using feather or other to communicate between python Dash apps

Hi there,
I am wondering IF anyone has played with starting multiple Dash Python apps under different IP addresses say app1 8050 app2 8060 SO in app1 ( 8050) you select some rows from a df and create some data you want to plot BUT the screen area you have left on app1 is not enough to do what you want to do AND you don’t want to keep scrolling up and down. SO app1 writes to a feather dataframe which app 2 ( 8060) spots has changed and it has a whole screen to plot to. So you use feather in memory dataframe as the “communication” between two Dash apps. What do you think? is there a better way? @chriddyp already pointed me at BUT I still would like to consider a shared memory approach.

anyone got any comments on this please? Thank you