YAML Is My DBA Now: Our Postgres Journey From DIY To Autopilot Self-Service
Talkat Data on Kubernetes Day NA 2025 · on Monday, 2025-11-10· at 14:10· Building B | Level 2 | B211-212·
summary
Chronicles a migration from VMs to Docker to Patroni operators and finally CloudNativePG, describing concrete wins and failures: moved from node-pinned containers to operator-managed clusters, regained block-level snapshots and multi-volume snapshots, introduced PGbouncer for reconnectless restarts, and shifted from ticket-driven ops to GitOps self-service. Reports a real wrong PVC incident and an old CloudNativePG split-brain bug that gets fixed in the next release, notes persistent gaps like unset shared_buffers defaults and one DBA per ~200 developers.
AI-written summary · about
chapters (10)
- 0:00Postgres automation journey
- 1:44Postgres in containers
- 4:15VMs to Kubernetes with Patroni
- 7:50Postgres on Docker Swarm
- 9:55Postgres operators limits
- 13:06Declarative vs imperative
- 16:17Developer view of operators
- 19:03Memory gaps in operators
- 21:13Myth of full automation
- 23:41Q&A
related talks
50mDemystifying Kubernetes for Postgres DBAs; A Guide to Operators
26mBeyond the DBaaS Trap: Achieving Data Sovereignty With Kubernetes and CloudnativePG
50mFrom VMs to Cloud-Native PostgreSQL in Kubernetes...
30mBridging the DevOps to Data Divide with a Common Cloud Native Stack
31mLightning Talk: Kubernetes as Your DBA
36mPostgreSQL: Ups, DevOps...
27mFrom Zero to Hero: Scaling Postgres in Kubernetes Using the Power of...
30mRunning PostgreSQL on Kubernetes in Production: Day 2 Operations with CloudNativePG