How to Migrate 700 Kubernetes Clusters to Cluster API with Zero D...
summary
Mercedes‑Benz migrates its legacy Terraform/Jenkins/Ansible fleet management into Cluster API, replacing a pipeline that managed 700 clusters and ~3,500 machines across four datacenters. The team uses Cluster API provider OpenStack, creates dummy kubeadm control plane objects to adopt existing control planes, migrates infra → worker machines → control plane in stages, and automates with Flux later. Migration completes with zero downtime, a Cluster API metrics exporter is built and donated, and the process proves operationally successful.
AI-written summary · about
chapters (25)
- 0:00Intro
- 2:16MercedesBenz Tech Innovation
- 3:05Agenda
- 5:56Workload Cluster
- 6:46Zero Downtime
- 7:27Legacy Cluster provisioning
- 8:22Target Cluster provisioning
- 9:00Preparation
- 9:40Three Steps
- 10:56Cluster Infra Migration
- 11:28Worker Node Migration
- 12:04Fake DM Control Plane
- 13:02Patch Status Fields
- 13:55Machine Set
- 14:28Machine Deployment
- 14:58Customer View
- 15:45Create a New Control Plane
- 16:24Adopt Legacy Provision Control Plane Machines
- 16:55Create OpenStack Machines
- 17:19Pause KCP
- 18:14Done
- 18:24Best Practices
- 21:47Future Plans
- 23:09Join us
- 23:27Questions
related talks
32mDon't Start With 500 Clusters: Patterns in Scaling Multi-Cluster Kubernetes…
12mCluster API To Streamline Kubernetes Clusters Lifecycle Management...
35mClusters as Cattle: How to Seamlessly Migrate Apps Across Kubernetes Clusters
31mSaving Money Exposed Our Architecture - Escaping ClusterAPI with Karpenter
46mThe Journey of Migrating 2500 Apps to Multi-Cluster Kubernetes and Zero Trust Istio
21mCluster API Deep Dive
45mLarge-scale kubernetes deployment, which tools to use and how - Rachid Zarouali
41mA Future Journey: How to Migrate 100 Clusters Between Clouds Without Downtime?