1

I am currently in the process of bringing my R-Shiny app online. Unfortunately, I am not getting anywhere with my current approach (EKS Cluster + Shinyproxy) (see question: "ERR_EMPTY_RESPONSE" - ShinyApp hosted over AWS (EC2 / EKS / ShinyProxy) does not work).

Therefore, I have considered taking the (supposedly) easier route via Fargate.

But I wondered how Fargate works exactly when several users want to access the service at the same time?

Background: My goal is to integrate my ShinyApp into another website and to use Shiny as a tool for data processing, so to speak. Important: It may be that more users want to process their own data at the same time. So it would be important that several users can use the tool at the same time without being able to mess with each other's data. (This is why I have wanted to use ShinyProxy so far).

Can one of you explain to me how Fargate deals with (I'll just call it that) multi-user access in connection with RShiny? Is there any risk at all that users will get in each other's way? In my dream world, each user gets their own R session - their own Docker image, so to speak.

I hope I was able to formulate my question in a meaningful way so that you understand what exactly I mean. :)

Thank you!

LePyka
  • 181
  • 8
  • "In my dream world, each user gets their own R session - their own Docker image, so to speak." That's not how Fargate works. Fargate itself isn't even aware that your software has user accounts. Fargate is just a server running the Docker container(s) you tell it to run. Your question about how Fargate "deals with" multi-user access is that it isn't even aware your app has multi-user access, so it does not "deal with" it in any way. You're looking for some sort of higher-level service management that doesn't exist. – Mark B Oct 09 '21 at 16:23
  • I was a little unclear in the wording - there are no user accounts in my software. I'm more concerned that users will simply access the application via their browser. You don't have to log in or anything like that. You simply upload your dataset and receive an evaluation. – LePyka Oct 09 '21 at 16:37
  • None of that is anything Fargate is going to handle for you. Fargate runs a docker container for you, that's it. It's no different from running a docker container on your local computer. There's nothing special there to magically spin up multiple containers per user or anything like that. – Mark B Oct 09 '21 at 17:45
  • Then my Question goes more towards rShiny! Thanks for the information and clarification! – LePyka Oct 09 '21 at 17:56

0 Answers0