What does Serverless mean for tomorrow’s abstracted infrastructure?
summary
Serverless moves the perimeter into provider-managed runtimes, which forces teams to build strong abstractions for triggers and persistence and to test both happy and fuzzed failure paths locally. Practical tools discussed include Firecracker microVMs for isolation, OpenTracing + Jaeger/Zipkin for distributed traces, and the Knative + Istio on Kubernetes stack that enables blue/green traffic steering by revision. Warning: early Knative/Envoy/ISTIO integrations have escape and visibility gaps, so enforce least-privilege IAM, scan supply-chain dependencies, and add behavioral-anomaly detection.
AI-written summary · about
chapters (9)
- 0:00Introduction to serverless
- 5:54Serverless as programming model
- 7:23Challenges in testing serverless
- 12:25Monitoring distributed systems
- 17:49Operating serverless architecture
- 20:16Blue-green deployments with Knative
- 26:17Security in serverless environments
- 31:19Developer accountability in cloud native
- 34:29Application security risks
related talks
53mServerless is the Abstraction We Deserve
58mServerless - the brief past, the bewildering present, and the beautiful (?) future
21mWhat is the Future of Serverless Computing?
48mThe future will be Serverless
52mConfusion In The Land Of The Serverless
18mA First Look at the Security of Serverless Applications
03mServerless is other people
28mThe Future of Serverless