GitOps - Operations by Pull Request
summary
Conference claim: a 15–20 person team moves from deploying once or twice a week to automated, multiple daily deployments and reports productivity doubled; full-cluster recovery happens in minutes by reconstructing declarative config from Git. Workflow uses Kubernetes plus Terraform/Ansible, stores all runtime config in a separate Git repo with one branch per environment, and uses an operator (eg Flux operator) to converge cluster state. Secrets use key-based encryption (eg sealed-secrets), diffs and observability drive canaries and alerting, and merges from staging to production trigger reviewed rollouts.
AI-written summary · about
chapters (19)
- 0:00Introduction
- 0:48Cordoba
- 2:47Git Ops
- 4:49declarative infrastructure
- 7:21pillars of GitOps
- 8:14Operator Pattern
- 9:21Extracting Config
- 10:44Sealed Secrets
- 15:32Pipeline
- 20:05Envoi
- 20:35Control
- 21:38Convergence
- 22:00Wipeout
- 22:21Life Cycle
- 23:22GitOps Theorem
- 23:47Summary
- 24:33Developers
- 25:15Questions
- 28:21Examples
related talks
47mGitOps - Using Git as your source of truth for build, deploy and observability
35mGitOps - the holy grail of devops
06mGitOps - At a Glance
35mGitOps in Action
34mGitOps: Git Based Application Deployment Patterns for Kubernetes
26mGitOps Lifecycle Management
26mGitOps: yea or nay?
55mGitOps, Fostering Next-Generation Operations Teams