FaaS: A Contemporary, Comparative Study
summary
Shows that breaking 400 microservices down can reclaim capacity from 4,000 app instances left idle overnight and avoids an autoscaler “event storm” when thousands restart. Compares open-source FaaS (OpenWhisk, OpenFaaS, OpenFaaS, riff, Knative) and tooling (Kubernetes, Docker, Prometheus, NATS) and finds Knative mature. Demonstrates Knative on Minikube demo building with Cloud Foundry buildpacks, an encoder/decoder event pipeline on channels, cold-start mitigation, and an AWS Lambda cost example with orders-of-magnitude savings.
AI-written summary · about
chapters (11)
- 0:00Introduction and context
- 1:53Evolution of serverless
- 3:24Defining serverless and FaaS
- 5:24Operational challenges
- 8:56Application behavior hurdles
- 10:36Breaking down applications
- 12:04Benefits of serverless
- 14:56Serverless challenges
- 17:59Use cases and adoption
- 20:01Platform selection criteria
- 28:10Knative and riff demo
related talks
31mFaaS Meets Node.js Frameworks: Developing Cloud Native Node.js Applications at Speed
05mLightning Talk: The State Of FaaS on Kubernetes - Michael Hausenblas, Red Hat
33mFaaS 101
33mGoing FaaSter: Cost-Performance Optimizations of Serverless on Kubernetes
35mServerless is Interesting But FaaS is Not Enough
34mFaaS and Furious - 0 to Serverless in 60 Seconds, Anywhere
19mDiscovering serverless with OpenFaaS
23mOpenFaaS: Serverless platform with no strings attached?