shinyproxy icon indicating copy to clipboard operation
shinyproxy copied to clipboard

Dash app crashes silently while redeploying ShinyProxy

Open dylancis opened this issue 1 year ago • 3 comments

When redeploying Shinyproxy while a Dash app is up and running already, the app will die (as expected) but silently, e.g. there would be no obvious way for any user to realise this unless they open up DevTools and see the error there - which us unliekly. err dash gran

With RShiny app, this wors better since the app will greay out and preventing the user to even try - you can catch generically the error using https://cran.r-project.org/web/packages/shinydisconnect/index.html

Is there any way Dash hosted app on ShinyProxy can tell the user more explicitely there was a 410 error code?

Thank you

dylancis avatar Nov 10 '23 10:11 dylancis

Hi, in most cases the front-end of ShinyProxy should be able to detect that the either the app or ShinyProxy has crashed and show an error page. It might be that this does not work optimal in your specific situation, could you therefore provide the following information:

  • which ShinyProxy version are you using? The mechanisms to detect app or shinyproxy failures were greatly improved in the 3.0.x release.
  • are you using Redis for user or app sessions (https://shinyproxy.io/documentation/configuration/#session-and-app-persistence)?
  • which loadbalancer or proxy are you using in front of ShinyProxy?
  • which authentication backend are you using?

I hope you can provide this additional information, since it would be good if we could find and fix this issue.

LEDfan avatar Jan 02 '24 09:01 LEDfan

Hi @LEDfan , please find in italic my answers, thanks for your help.

Hi, in most cases the front-end of ShinyProxy should be able to detect that the either the app or ShinyProxy has crashed and show an error page. It might be that this does not work optimal in your specific situation, could you therefore provide the following information:

  • which ShinyProxy version are you using? The mechanisms to detect app or shinyproxy failures were greatly improved in the 3.0.x release. shinyproxy-2.6.1
  • are you using Redis for user or app sessions (https://shinyproxy.io/documentation/configuration/#session-and-app-persistence)? No
  • which loadbalancer or proxy are you using in front of ShinyProxy? AWS
  • which authentication backend are you using? Auth0

I hope you can provide this additional information, since it would be good if we could find and fix this issue.

dylancis avatar Jan 08 '24 08:01 dylancis

Hi @dylancis , could you try upgrading to ShinyProxy 3.0.2? This releases improves the detection of crashed/stopped apps.

LEDfan avatar Jan 08 '24 12:01 LEDfan