This video explains simplyblock’s core architectural approach: MAUS, or Modular Adaptive Unified Shared-Everything.
MAUS is the design principle behind how simplyblock separates compute and storage into independently scalable pools while maintaining shared access to NVMe volumes from any node in the cluster. The architecture enables the scale-out storage model that allows Kubernetes workloads to access high-performance block storage over NVMe/TCP without co-locating compute and storage hardware.
Key topics covered:
- What “shared-everything” means in a disaggregated NVMe storage context
- How simplyblock distributes data across storage nodes using erasure coding and replication
- How the MAUS design enables elastic scaling: add storage nodes without touching compute, and vice versa
- NVMe/TCP as the fabric connecting compute to the shared storage pool
- How multi-path and automatic failover work within the MAUS architecture
This architecture overview is useful context before evaluating simplyblock for Kubernetes persistent storage, database workloads, or OpenShift deployments.