Parallel and Asynchronous Code compatibility on shinyapps.io plans

Posit Professional & Hosted Products shinyapps.io
, ,
1

Hi RStudio Community,
I'm wondering if someone could clarify as to what the capabilities of the shinyapp.io resource tiers are in regards to running in parallel with future/furrr or using asynchronous code with future/promises.
By the sound of the Performance Boost feature on tiers >= Basic:

Multiple worker processes per application, up to 8GB of RAM and the ability to add additional instances to keep your applications responsive as more people use them.

it seems like parallelization and asynchronous code will only work on these tiers and not on the tiers < Basic since those are single threaded.

Is this true? Is anyone using parallelized or asynchronous code that can confirm or elaborate on what the situation is with package functionality that relies on multi-threaded operations?

2

Hi,

As far as I know, the asynchronous code lets you run some slow parts of code in the background and free the R process. So, you could benefit from the future/promises options even if your app has only access to one worker.

3

Hi @bagrami,
Thanks for the response. I know that the async backend often requires the background process be offloaded to another worker. Have you had success with this in action?

4

Hi @yogat3ch,

I am sorry, but I am still a newbie in this topic. I have only tried the async backend running a shiny app in a local machine, and it worked fine. But I don't know if it offloaded the background process to another worker.

If I get more information about this, I will let you know.

1 Like