Keynote: Towards Future Ops
summary
Argues operations evolve toward heterogeneity and multicloud strategies: teams run VMs and containers across AWS, Google and Azure (example: one company keeps web on AWS and data on Google; Azure has 17 China regions). Recommends declarative infra with terraform and schedulers like kubernetes/nomad that treat a data center as compute, classifying state as ephemeral sticky persistent, using saved “plans” to build trust. Predicts schedulers gain ML-driven self‑management in 2–3 years; app‑level security via Vault and team‑embedded security cuts months‑long firewall delays.
AI-written summary · about
chapters (12)
- 0:00Start
- 10:59Heterogeneity
- 13:04Multi-Cloud Motivation
- 17:21Multi-Tech Motivation
- 18:15Multi-Compute Motivation
- 24:14Choices
- 27:11Challenges
- 28:23Solutions
- 34:52Benefits
- 39:56Historically
- 43:28App-Level Security
- 48:51DevOps
related talks
15mKeynote: A Glimpse Into the Future of Cloud Native: Emergent Infrastructure
15mKeynote: Scaling DevOps – Moving from Mechanization to Automation
11mKeynote: Like a Magic Eye Illusion: Seeing the Bigger Picture for Cloud-Native S...
24mHow to survive Cloud – An Ops perspective
42mCloud Native Applications
14mKeynote: Open Source & Operational Excellence
28mKeynote – Arun Gupta
06mKeynote: The Future of GitOps