Ship It Weekly Show Notes

In this episode of Ship It Weekly, Brian digs into 3 big themes for anyone running Kubernetes or building internal platforms.

First, Kubernetes is officially retiring Ingress NGINX and moving it into best-effort maintenance until March 2026. We talk about what that actually means if you’re still using it and how to think about choosing and rolling out a replacement ingress.

Second, we look at how CNCF is defining platform engineering and what “platform as a product” looks like in practice, plus some hard-earned lessons from running Kubernetes in production.

Third, we talk about AI as a first-class workload on Kubernetes. CNCF’s new Certified Kubernetes AI Conformance Program aims to standardize how AI runs on K8s, and recent writing on SRE in the age of AI looks at what reliability means when systems learn and drift.

In the lightning round, we hit good reads on database migrations, Postgres upgrades, and a distributed priority queue on Kafka. We wrap with the human side of incidents: fixation during incident response and using incidents as landmarks for the tradeoffs you’ve been making over time.

If you’re on a platform team, responsible for SLOs, or the person people ping when “Kubernetes is weird,” this one should give you concrete questions to take back to your roadmap and runbooks.

Links from this episode

https://kubernetes.io/blog/2025/11/11/ingress-nginx-retirement/

https://www.haproxy.com/blog/ingress-nginx-is-retiring

https://www.cncf.io/blog/2025/11/19/what-is-platform-engineering/

https://www.cncf.io/announcements/2025/11/11/cncf-launches-certified-kubernetes-ai-conformance-program-to-standardize-ai-workloads-on-kubernetes/

https://devops.com/sre-in-the-age-of-ai-what-reliability-looks-like-when-systems-learn/

Lightning round

https://www.cncf.io/blog/2025/11/18/top-5-hard-earned-lessons-from-the-experts-on-managing-kubernetes/

https://www.tines.com/blog/zero-downtime-database-migrations-lessons-from-moving-a-live-production

https://palark.com/blog/postgresql-upgrade-no-data-loss-downtime/

https://klaviyo.tech/building-a-distributed-priority-queue-in-kafka-1b2d8063649e

https://sreweekly.com/sre-weekly-issue-497/

https://ferd.ca/ongoing-tradeoffs-and-incidents-as-landmarks.html

In this special kickoff episode of Ship It Weekly, Brian walks through three major outages from the last few weeks and what they actually mean for DevOps, SRE, and platform teams.

Instead of just reading status pages, we look at how each incident exposes assumptions in our own architectures and runbooks:

Topics in this episode:

• Cloudflare’s global outage and what happens when your CDN/WAF becomes a single point of failure

• The AWS us-east-1 incident and why “multi-AZ in one region” isn’t a full disaster recovery strategy

• GitHub’s Git operations / Codespaces outage and how fragile our CI/CD and GitOps flows can be

• Practical questions to ask about your own setup: CDN bypass, cross-region readiness, backups for Git and CI

This episode is more of a themed “special” to kick things off.

Going forward, most episodes will follow a lighter news format: a couple of main stories from the week in DevOps/SRE/platform engineering, a quick tools and releases segment, and one culture/on-call or burnout topic. Specials like this will pop up when there’s a big incident or theme worth unpacking.

If you’re the person people DM when production is acting weird, or you’re responsible for the platform everyone ships on, this one’s for you.

Links from this episode

Cloudflare outage – November 18, 2025

https://blog.cloudflare.com/18-november-2025-outage/

https://www.thousandeyes.com/blog/cloudflare-outage-analysis-november-18-2025

AWS us-east-1 outage – October 2025

https://aws.amazon.com/message/101925/

https://www.thousandeyes.com/blog/aws-outage-analysis-october-20-2025

GitHub outage – November 18, 2025

https://us.githubstatus.com/incidents/f3f7sg2d1m20

https://currently.att.yahoo.com/att/github-down-now-not-just-211700617.html