Multi-Kubernetes, Multi-Region — Help!
summary
Describes a reproducible real-world multi-region Kubernetes deployment spanning Paris and California using Terraform and Ansible for infra-as-code and Argo CD for GitOps. Global routing is handled with Gateway API and GSLB, Istio with Envoy sidecars provides multicluster east–west mesh, and PostgreSQL with Patroni manages replication with a CI job that performs live disaster‑recovery promotion to California without data loss; CNCF multicluster primitives are noted as still evolving.
AI-written summary · about
chapters (13)
- 0:00Introduction
- 1:46Multirégion definition and stakes
- 4:12Why multirégion
- 5:56Cluster deployment
- 7:19GitOps and infrastructure as code
- 9:56North-South traffic routing
- 12:15East-West service communication
- 18:11Multirégion data management
- 26:19CAP theorem and replication
- 30:09Disk vs application replication
- 32:54Failover and disaster recovery
- 37:45Summary and conclusion
- 38:49Q&A
related talks
36mThe Architecture of a Multi-Cloud Environment with Kubernetes
31mFrom Region To Multi-AZ: Building Resilient Cloud Infrastructure With OpenStack, Kubernetes, and OVN
27mKubernetes Everywhere: Lessons Learned From Going Multi-Cloud
20mScale production grade K8s Multi-Cluster environments using GitOps
33mKubernetes-Driven Multi-Cloud Deployments: Maximizing Scalability and Flexibility
33mWhat We’ve Learned Building a Multi-Region DBaaS on Kubernetes
35mKubernetes journey
26mScaling Multi-cluster Kubernetes at Teleport Cloud