Replatforming a $4B Retailer Onto Kubernetes and Linkerd
summary
A $4B Nordic retailer migrates from Azure App Service to Kubernetes and Linkerd and reports a 75% hosting cost reduction while keeping faster deployments and better disaster recovery. The platform uses Flux GitOps for config, Prometheus + Alertmanager for alarms, and KEDA autoscaling to speed scale‑up. Linkerd provides mutual TLS via Linkerd and traffic insights that find a socket‑reuse bug during load tests. Onboarding emphasizes champions, docs as code, and templates.
AI-written summary · about
chapters (14)
- 0:00Introduction
- 1:20Goals
- 2:05Background
- 2:43Before Kubernetes
- 6:05Choosing technology
- 6:22Guiding principles
- 9:45Linux Admins
- 10:35Flux
- 11:36Kubernetes Tools
- 12:03Onboarding
- 14:02Why bother
- 14:59Alarms
- 16:04Service Mesh
- 17:37Load testing
related talks
01h 02mLearnings from the Enterprise-level Kubernetes transformation at Elkjøp Nordic
20mHow H-E-B Curbside Adopted Linkerd During a Pandemic
31mMigrating Legacy Infrastructure to Kubernetes: Olark’s Lessons From the Trenches
32mLinkerd and the Service Mesh
39mService Mesh: There and Back Again
36mKubernetes in the Enterprise Transforming Rabobanks On Premise Applications
43mReplatform to Amazon EKS
37mLinkerd End User Panel: Case Studies from Production