Deep Dive: Kubespray (a sig-cluster-lifecycle's project)
summary
Kubespray uses an Ansible-based CLI to deploy and manage production Kubernetes on bare metal and clouds, with ~6,400 GitHub stars, ~450 contributors and ~4,400 commits and a year-over-year +200 contributor growth. It supports six OSes, seven network plugins, ARM, local LBs and the local-path provisioner, runs CI across 10 test clusters on OpenStack/OVH/GCE created by Terraform, and handles lifecycle ops (create, upgrade, scale, remove) but upgrades can take 4–5 hours and binaries are problematic in air-gapped installs. Separate etcd for metadata and events is recommended past ~2,000 nodes. No API, command-line only.
AI-written summary · about
chapters (18)
- 0:00Start
- 1:32Mission of kubespray
- 1:42Kubespray at a glance
- 2:51Deployment workflow
- 4:07Lifecycle of cluster operations
- 4:54Certificate Management
- 5:59High Availability
- 6:54Local LB (default)
- 7:28User options
- 9:35Cross-platform
- 10:18Multi-arch
- 11:26Project principles Ensure that Kubespray is production-ready
- 12:56Continuous Integration
- 14:59Contribution guidelines
- 15:35Larger contributions
- 16:47Long-term support in Kubespray
- 17:35Kubespray Community In 2019
- 19:17Roadmap 2019
related talks
32mDeep Dive: Kubespray (a sig-cluster-lifecycle's project)
34mDeep Dive: Kubernetes/Kubespray (Cluster Lifecycle) SIG
36mKubespray Intro
44mSIG Cluster Lifecycle: What's New in Kubespray
12mKubespray: Driving Cost-Efficiency for AI on Kubernetes
04mLightning Talk: Kubespray CI with KubeVirt on Baremetal
40mBuilding Custom GPU Clusters at Scale: Using Kubespray To Create High-Performance AI Infrastructure
19mHardening automation with Kubespray - A journey into the open-source security world