Building a high-performance analytics ecosystem on Amazon EKS

on YouTube ↗14m

" summary

Shows a production analytics platform on Amazon EKS using Kubernetes, GPU nodes, and global regions, with concrete optimizations: an idling system that shuts inactive pods and resumes them in ~7–10 seconds, yielding a 50:1 concurrency reduction in resource use; a daemonset (“cedar”) that prefetches and caches container images on nodes to eliminate cold starts; an egress controller plus SELinux and encrypted NFS volumes for outbound and data protection; and a CMDB-sync pod that maps dynamic namespaces to configuration records. Tools, numbers and modules are provided and tested in a running demo.

AI-written summary · about

" chapters (4)

1. 0:00Introduction
2. 2:41Deployment optimization
3. 6:42Security
4. 10:28Demo

" related talks

• 41mBuilding a modern data platform on Amazon EKSAvichay Marciano, Omer Haim · AWS Summit Tel Aviv 2023
• 42mDrupal, Kubernetes, ProductionDaniel Kossakowski · DrupalCamp Poland 2023
• 17mKeynote: Kubernetes: Lessons from the TrenchesArun Gupta · KubeCon + CloudNativeCon + Open Source Summit China 2019
• 17mKeynote: Kubernetes: Lessons from the TrenchesArun Gupta · KubeCon + CloudNativeCon China 2019
• 13mOne Cluster, End-to-EndAnoop Saha, Shreyas Adiyodi · Ray Summit 2026
• 31mAn Adventure in Multi Cloud with Kubernetes and Distributed DataLeon Kuperman · Distributed SQL Summit Asia 2021
• 29mThe Frugal k8s Engineer: Strategies for Amazon EKSPiyush Mattoo, Dariush Azimi · Conf42 Kube Native 2024
• 29mAmazon Elastic Container Service for Kubernetes (EKS): A case study - M. Re Ferre' and G. CarabelliM. Re Ferre', G. Carabelli · ContainerDay 2018