Remote Content Execution with RStudio Connect and Kubernetes

on YouTube ↗19m

" summary

Announces Off-Host Content Execution for RStudio Connect using Kubernetes: content jobs run in external pods instead of the Connect host, bundled with the Job Launcher and currently offered as a beta. Requires a Kubernetes cluster, kubectl, Helm charts, a Postgres database and ReadWriteMany shared storage (NFS/EFS). Publishers keep the same UX, can target execution images or let Connect map versions automatically; expect slightly longer cold starts from image pulls and job spawning. Installation demonstrated via Helm values.yaml and helm upgrade --install.

AI-written summary · about

" chapters (6)

1. 0:00Introduction to Connect
2. 2:10Off-host execution launch
3. 4:51User experience demo
4. 8:28Administrator requirements
5. 10:22Speed-run installation
6. 18:11Conclusion and acknowledgments

" related talks

• 19mRemote Content Execution with RStudio Connect and KubernetesKelly O'Briant · rstudio::conf 2019
• 19mConnect on Kubernetes: Content-level ContainerizationE. David Aja · posit::conf(2023)
• 22mRStudio Connect Past, present, and futureJeff Allen · rstudio::conf 2019
• 18mJob Launcher Changing where we run R stuffDarby Hadley · rstudio::conf 2019
• 17mSimplifying K8s Connectivity With a Powerful CLI for SRE'sSai Sandeep Ogety · Conf42 Cloud Native 2025
• 12mUsing Jupyter with RStudio Server ProKarl Feinauer · rstudio::conf 2019
• 15mR Markdown + RStudio Connect + R ShinyKolbi Parrish, Andy Pham · rstudio::conf 2019
• 36mBuilding a Scalable Data Science & Machine Learning cloud using KubernetesMurali Paluru · Big Data Conference Europe 2019