The Konflux Chronicles: Productizing Observability the Hard Way
summary
Productized the Observability Operator on Konflux (Kubernetes-native) using Tekton pipelines, producing snapshots, a release bundle/FBC and images that land in advisories; v1.0 ships after delays and retries, now at 1.2. Migration replaces Delorean/Zuul/CPASS and integrations with Koji/Brew/Errata, but initial run hits cluster capacity limits, slow UI, many YAMLs to maintain, flaky nudges and CVE false negatives that required exclusions and manual fixes; automation and security scanning now work but ramp-up cost is high.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 0:46Productizing cloud native is hard
- 2:15From legacy to Konflux operator
- 4:11Switching to Konflux
- 8:57Migrating to Konflux challenges
- 13:42Konflux config and repos
- 18:07Red Hat’s strategic move
- 20:59Early Konflux pain points
- 23:11Konflux: what worked vs failed
- 25:17Q&A
related talks
23mWhen Kafka Goes Cloud Native: Observability That Actually Works!
27mRethinking Observability for Off-the-Shelf Tools
07mKeynote: Observability Simplified
26mFlux + Observability: Featuring Prometheus Operator and Pixie
01h 17mIntro to Kubernetes, GitOps, and Observability Hands-On Tutorial
38mA story about platform engineering — Niklas Voss
09mKeynote: Scaling Observability Pipelines in the Enterprise
12mKeynote: From Outage To Observability: Lessons From a Kubernetes...