R in Containers: Rocking Kubernetes
Details
Microservices, containers, and orchestration: Are containers just lightweight virtual machines? What is the impact of containers and containerization on the work of data scientists? How does the average R User use these technologies?
This CRUG meetup will attempt to answer these questions.
- Containers for Data Scientists - Troy Hernandez - 6pm
Troy will give a brief introduction to containers, images, orchestration, and microservices. He will then lay out 3 popular use-cases for data scientists using containers.
- The Rocker Project - Dirk Eddelbuettel - 6:15pm
As usual, Dirk was ahead of the curve. He started the Rocker Project (Docker Containers for the R Environment) with Carl Boettiger in June of 2014. After 5 years the project has over 200 forks and 1000 stars on Github, and over two million downloads on Docker Hub.
https://www.rocker-project.org/
CRUG is thrilled to have Dirk present the Rocker project, a little of its history, and will then do a demo of a couple of his favorite applications of Rocker containers.
- Rocking Kubernetes - Ray Buhr - 7pm
Ray has been using Rocker containers in production since 2017. In a prod environment, container images are scaled up to 10s, 100s, or 1,000,000s of deployed containers. That kind of scale requires some management/orchestration: Enter Kubernetes (k8s).
Ray will take a Shiny app running locally and show you how to run it on k8s in the cloud. He will also discuss some considerations for working with data science container workloads in production environments.
###############################################
As usual, IBM is sponsoring the meetup and providing food and beverages. If you'd like to do some pre-learning, there is an IBM meetup focused exclusively on Kubernetes/container orchestration on Thursday, November 14th:
https://www.meetup.com/IBM-Cloud-Chicago/events/265932434/