Load balancing, scaling, Multithreading R Shiny app on Azure

We're huge fans of R Shiny and have a production application running on Azure within a docker container. We have a few issues, however, that would love some assistance with. We're looking for a work for hire (not sure this is the right forum) person who can help us harden our R Shiny Instance.

Our models are mostly conjoint analysis and Discrete Choice experiments and we're having some issues.

  1. We don't believe that the R Shiny application is using all of the VM resources, as the VM is an 8-core 32GB machine which is at low-single digit %'s.
  2. We don't believe the auto-scaling is working. Despite having 3 instances of the application, If one R Shiny instance is frozen, all other users become locked out.
  3. We don't believe we have multithreading/parallel processing working correctly. We have some beefy CLM and OLS models running, however they're taking very long to calculate (minutes) for only a few hundred data points.

Any guidance or assistance would be deeply appreciated. We figured this would be the best forum to find an expert to help us productionize our system. Join our Discord here: subconscious.ai and we would love to chat further!!

Best Regards,

What do you mean by that? Are you using the open source Shiny Server? Do you run Rscript -e "shiny::runApp()" directly in the Docker container?

Hi Avi,
If you are still on the lookout to hire someone, I would be glad to help you out as I am an experienced R Shiny consultant.
You can reach out to me on my email here

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.

If you have a query related to it or one of the replies, start a new topic and refer back with a link.