Where We’re Going, We Don’t Need Servers!
summary
Public cloud doesn’t mean universal adoption — about 70% of IT spend remains on premise, so choices matter. Serverless is broader than functions: S3 and SQS predate Lambda (Lambda in 2014) and offer true pay-per-use primitives, and managed databases like Aurora are easier lifts than Dynamo‑style shifts. OpenFaaS works on Kubernetes; Knative is immature and tied to service meshes. Best practice is pragmatic: map one microservice instance to one function, avoid thousands of tiny functions, watch timeouts (now up to ~30 minutes), and expect limited observability and runtime control.
AI-written summary · about
chapters (6)
- 0:00Intro
- 1:55Agenda
- 2:55Part 1: cloud and serverless
- 33:14Part 2: microservices and functions
- 46:32Part 3: what you should do about it
- 49:15Outro
related talks
31mThe Evolution of Serverless: Past, Present and Future
53mServerless is the Abstraction We Deserve
51mFaaS or not to FaaS? Visible and invisible benefits of the Serverless paradigma
40mWho Needs Serverless?
01h 00mServerless - reality or BS - notes from the trenches
51mConfusion In The Land Of The Serverless
42mNo Server Is Better Than No Server: Building Blocks for Serverless Architectures
17mCloud Computing Made Better: An AWS View on Serverless Computing