HPE GreenLake for File Storage is an all-NVMe, scale-out file storage solution designed for data-intensive enterprise workloads. The platform utilizes a disaggregated, "shared-everything" architecture to decouple compute and storage resources, allowing for independent scaling of performance and capacity.

Key Capabilities

Disaggregated Shared-Everything (DASE): The platform separates stateless compute nodes (C-Boxes) from storage nodes (D-Boxes) via a high-speed 100GbE fabric. This architecture eliminates traditional storage controller bottlenecks and allows resources to be added non-disruptively.
Stateless Compute Nodes: Controllers manage file services and data management software without local state. According to the vendor, this design ensures that controller failures do not require metadata re-indexing or cause service disruption.
Inline Data Reduction: The system employs similarity-based deduplication and compression. The vendor states that these data reduction techniques are intended to maximize effective capacity and reduce the overall storage footprint.
Global Namespace: Provides a unified view of data across the entire storage cluster, supporting native S3 and NFSv4.1 protocols for structured and unstructured data access.

Audience & Use Cases

Audience: Infrastructure architects, data scientists, and storage administrators managing large-scale unstructured data environments.
Use Case: AI/ML Training Pipelines, High-Performance Computing (HPC), and the consolidation of massive Data Lakes requiring high throughput and low-latency access.

Technical Specifications

Architecture: Disaggregated Shared-Everything (DASE)
Storage Media: All-NVMe Flash
Networking: 100GbE (Aruba or NVIDIA Spectrum)
Scalability: Supports expansion from terabytes to exabytes of effective capacity.
Availability: HPE states the solution provides a 99.9999% ("six-nines") availability guarantee.

HPE GreenLake for File Storage

What is HPE GreenLake for File Storage?

Categories & Use Cases