Scaling Shiny apps with asynchronous programming

June 7, 2018 Joe Cheng

This updated talk has new info since Joe spoke at rstudio::conf 2018.

Download Materials from the webinar

Abstract
One persistent challenge with developing Shiny apps for live deployment, has been the R language runtime’s single-threaded nature. Because of this, a given Shiny app process can only do one thing at a time: if it is fitting a model for one client, it cannot simultaneously serve up a CSV download for another client. Asynchronous programming offers a way to offload certain classes of long-running operations from the main R thread, such that Shiny apps can remain responsive.
The next major release of Shiny will include deep support for asynchronous programming via 'promises', inspired by the JavaScript abstraction of the same name, but with significant enhancements to integrate seamlessly with Shiny's reactive constructs and to allow app authors to write code that looks mostly like idiomatic R. Programming with these promises has its own learning curve, but the payoff can be huge if your app has a significant bottleneck.

About the speaker

Joe Cheng 
CTO, RStudio

Joe Cheng RStudio CTO. Joe was the original creator of Shiny, and leads the team responsible for Shiny and Shiny Server.

About the Author

Joe Cheng

Joe Cheng is the Chief Technology Officer at RStudio. Joe was the original creator of Shiny, and leads the team responsible for Shiny and Shiny Server. GitHub: https://github.com/jcheng5

Follow on Twitter Visit Website More Content by Joe Cheng
Previous Video
Best practices for working with databases
Best practices for working with databases

Get the most out of joining R forces with database forces. We will review key concepts, share the latest in...

Next Video
Part 1 - Introducing an R interface for Apache Spark
Part 1 - Introducing an R interface for Apache Spark

sparklyr facilitates a connection between R and Spark using a full-fledged dplyr backend with support for t...