Kubernetes Runs Anywhere, but Does your Data?
summary
Kubernetes gets storage portability by running storage systems inside the cluster: Rook uses custom resource definitions and the operator pattern to deploy and manage Ceph, CockroachDB and Minio on any Kubernetes (demo runs on a laptop). The demo shows automated dynamic provisioning via storage classes, clientsets and informers, a Ceph-backed MySQL serving WordPress, a CockroachDB load generator doing real writes, and an S3-compatible Minio — all deployed and functioning.
AI-written summary · about
chapters (13)
- 0:00The portability promise
- 1:21Benefits of portability
- 2:34Kubernetes storage concepts
- 4:27Storage abstractions
- 7:26Limitations of current setup
- 10:13The need for portable storage
- 12:20Custom resources
- 14:38The operator pattern
- 17:21Building operator logic
- 22:13Summary of portable data
- 23:55Demonstration
- 26:42Running stateful apps
- 30:13Q&A and conclusion
related talks
37mBuilding a Storage Cluster with Kubernetes
37mIntro: Rook
35mDistributed storage with Rook
05mSponsored Keynote: Keep Kubernetes Caffeinated
37mDeep Dive: Rook
27mKubernetes The Fun Way
07mRook: Enterprise Storage for Kubernetes
32mData Without Borders - Using Rook Storage Orchestration at a Global Scale