Reliability as Code (RaC), A Startup Case Study
summary
Argues reliability moves into source control by baking tests, SLIs/SLOs and infra into deployments and measuring MTTD/MTTR. Deploys a sock shop microservices sample to minikube, uses Terraform (notes Pulumi preference), raises replica count 8, runs Litmus chaos for controlled failures, captures protocol traces with Pixie -> New Relic, and wires load tests and error budgets. Demo is pre-recorded and mostly succeeds; setup and observability show measurable recovery times and enforceable service levels.
AI-written summary · about
chapters (8)
- 0:00Introduction and welcome
- 1:07What is Reliability as Code
- 2:53Key metrics and KPIs
- 7:49Core disciplines for RaC
- 12:48Demo overview and setup
- 16:02Live demo walkthrough
- 20:40Service levels and golden signals
- 21:42Recap and closing
related talks
27mContinuous Reliability. How?
30mAchieving DORA outcomes by building reliability
35mKeynote: Reliability in Cloud Native DevOps
42mArchitecting for Reliability
32mBuilding Reliability into Kubernetes using Chaos Experiments with Terraform
25mLive Q&A Session #2
27mTrusting Mission Critical Infrastructure: Building Reliable CD P...
30mImproving Reliability w/ Error Budgets & SRE