The presentation embedded below, titled “Running Kubernetes on Ephemeral Infrastructure”, talks about what happens when you run large kubernetes clusters for mission critical enterprise apps and you destroy every node every 48 hours. The presenters, Greg DeArment and Vlad Ungureanu of Palantir Technologies, goes into why they do that, some key benefits, and insights into challenges they ran into in taking that approach.
Kubernetes is an open-sourced container orchestration system for automating application deployment, scaling and management. It was conceptualized and developed by Google, and is now maintained by the Cloud Native Computing Foundation. Palantir is a user of Kubernetes and a major contributor back to the open source project.
By leveraging Kubernetes as the scheduling and execution engine for all distributed compute, Palantir was able to modernize their infrastructure in ways that provides enhanced resiliency, reduces risks, and optimizes the delivery of compute resources to developers and end clients.
For more on how see:
For more on related topics see:
- The CTOvision guide to Cloud Computing
- A Video Overview of Kubernetes
- Cyberwar: a 3 way gage fight
- Hacking The Cloud