Serverless and containers are coming together: an AWS perspective
summary
Claims that serverless patterns are arriving inside container platforms and shows concrete tooling and limits: AWS Fargate provides an infrastructureless data plane for EKS/ECS by sourcing a dedicated one VM per pod runtime, sizing pods by request not limit and billing for pod resources, demoing a 1→10 pod scale with perceptible cold‑start lag. Also covers higher abstractions like Co-pilot and KNative for app‑first workflows, and the move to microVMs via Firecracker to cut startup time and improve fleet efficiency.
AI-written summary · about
chapters (17)
- 0:00Start
- 3:29Containers Ws and Serverless?
- 5:35Let's de-compose the container orchestrator
- 8:392 trends in the industry towards serverless containers
- 9:04Trend #1: how do we make the infra disappear?
- 10:59Cloud Abstractions
- 13:05EKS data plane options
- 13:52The EC2 flow at 33,000 feet
- 14:35The Fargate flow at 33,000 feet
- 19:54Simplified deployment flow
- 22:54Trend #2: a schizophrenic rush towards a better world
- 25:51What's AWS Copilot
- 26:17AWS Copilot (CLI based abstraction)
- 27:20What's knative
- 28:26Knative/Fargate: high level architectural diagram
- 29:41Conclusions
- 31:34Call to action: explore more - Useful links • Amazon ECS and AWS Fargate
related talks
51mServerless Compute at the Heart of Your EDA
43mBringing EDA to the Containers World
51mWho needs containers in a serverless world?
47mServerless Compute at the Heart of Your EDA
41mContainers or Serverless Functions - When to choose what?
24mContainers vs Serverless from a DevOps standpoint
17mCloud Computing Made Better: An AWS View on Serverless Computing
01h 01mServerless containers with Knative and Cloud Run