Building a CI/CD driven infrastructure for managing Kubernetes clusters
summary
A CI/CD driven Kubernetes platform built with GitLab CI pipelines, Terraform, Matchbox and Container Linux automates full cluster redeploys in ~15 minutes and weekly rolling node reinstalls. Rack‑centric clusters use three masters, per‑rack BGP speaker for external routing and metalLB for backend services while fronting uses an F5 container connector. Namespace templates, per‑team Prometheus and RBAC enforce multi‑tenancy for a fleet serving ~34 million monthly users and it runs in three datacenters.
AI-written summary · about
chapters (10)
- 0:00Introduction and company overview
- 7:31Motivation for Kubernetes adoption
- 10:06Cluster design and rack setup
- 13:01Front-end load balancing with F5
- 16:37Back-end load balancing with MetalLB
- 19:24Ingress controllers and traffic setup
- 24:00Cluster deployment and configuration
- 27:13CI/CD driven cluster operations
- 31:04Multi-tenancy and namespace management
- 35:43Closing remarks and Q&A
related talks
36mMulti-Cluster Kubernetes Quick Delivery with IaC and GitOps
30mMission Critical Kubernetes
35mKubernetes journey
24mSelf-Service Multi-Cloud K8s Platform at a Regulated Financial Enterprise
26mRunning GitLab and GitLab CI in Kubernetes for the Enterprise
28mBuilding and Managing Kubernetes with Kubernetes
32mRunning Kubernetes in a Manufacturing Line – What Could Possibly...
02h 47mA GitOps Architecture from Scratch: Gitlab, Ansible, Terraform, Kub and AWS