I'm an R developer at my company, and we're currently hosting our Shiny apps on a separate Kubernetes platform. We use GitHub actions for building Docker images and deploying them for online access. However, I've been asked to explore alternatives, and I believe Azure's RStudio Server could be an ideal fit since we're already deeply invested in Azure services like DataFactory and DataBricks. I'm curious if anyone has real experience with Azure RStudio Server. Can anyone share their insights on its limitations or challenges? Are there any hidden costs to watch out for? Also, will I be able to focus solely on R code knowing that Microsoft manages backend tasks like patching and CVEs? Finally, does anyone know if Python and Reticulate work smoothly in this environment?
1 Answer
Hey there! I’d definitely say to proceed with some caution. It sounds like you're stepping into this migration with some uncertainty. While Azure RStudio Server could be a good fit, be aware that R isn't as prevalent in the data science community as it used to be, which could limit your support and resources. I’ve seen more demand leaning towards technologies like DataBricks for similar tasks, but if Shiny meets your specific needs for building front-end apps, you might find it useful!
I totally get that. We also use R within our DataBricks but there are specific needs that Shiny fulfills for us, like building user-friendly front-ends. If you're looking into hosting Shiny apps, I'd think about checking out how they perform in Azure RStudio Server. It might take some digging to find concrete answers on limitations.