Lustre in Kubernetes
summary
Describes a working Lustre deployment inside Kubernetes that uses KubeVirt VMs with Lustre kernel modules baked into container images, YAML manifests to provision MGS/MDS/OSS and PV/PVC storage, and secrets with startup scripts to bootstrap servers; a single-MGS/MDS/OSS test surprisingly outperforms native Lustre (likely from client/storage colocation) but lacks node-failure tolerance and ties local PVs to nodes; automation exists but bootstrapping is inflexible; next steps are integrating an operator SDK and exploring Rook for autoscaling and failure recovery.
AI-written summary · about
chapters (7)
- 0:00Start
- 0:06Key Concepts
- 3:04Lustre Architecture
- 13:34Current Limitations
- 14:32Rook / Operator SDK
- 15:17OperatorSDK Integration
- 16:04What's Next?
related talks
07mLightning Talk - Say what? You're Running the Storage Platform IN Kubernetes?
37mBuild a Kubernetes Based Cloud Native Storage Solution From Scratch
31mElephant in the room
36mIt’s All for the Users. More Durable, Secure, and Pluggable. KubeVirt v0.53
26mRunning Kubernetes Clusters the Kubernetes-Native Way
33mPiraeus Datastore - Persistent Volumes for Kubernetes
42mVirtual Machine Orchestration in Kubernetes Utilizing KubeVirt
27mHigh Performance KubeVirt in Action
