Gergely Orosz · 5 min read · Product Execution
CI/CD with Robert Erez
Read the original on newsletter.pragmaticengineer.com →
Robert Erez discusses how teams deploy software safely and efficiently at scale, covering Kubernetes, GitOps, platform engineering, progressive delivery, feature flags, cloud development environments, and the growing role of AI in CI/CD. The conversation highlights practical tradeoffs between deployment approaches and the ongoing relevance of self-hosted options.
Key Takeaways
- Roll forward, never backwards: for stateful systems with databases, treat a failure in v2 as a push to v3 with the fix, not a rollback to v1.
- GitOps isn’t really about Git: the four pillars (declarative, versioned/immutable, pulled, continuously reconciled) don’t require Git, and secrets should not live in repos.
- Continuous deployment can be overkill; continuous delivery is often more practical, with automatic or periodic pushes as appropriate.
- Feature toggles are a better safety net than rollbacks, but toggle debt can accumulate—regularly clean up unused flags.
- A Git repo can become a scalability bottleneck at large scale, prompting pull-based GitOps approaches and alternative state-sharing patterns.
Excerpt
Stream the latest episode Listen and watch now on YouTube , Spotify , and Apple . See the episode transcript at the top of this page, and timestamps for the episode at the bottom. Brought to You by • Antithesis – if you're using agents to code, the problem isn't writing the code but making sure it didn't break anything. Antithesis goes beyond code review and runs your whole system in faster than real-time and identifies hard to find bugs before your users hit them in production. Antithesis enables teams like Jane Street, Fly.io, and the etcd community to use…