Availability of HPE GreenLake for Block Storage Built on Alletra Storage MP Release 3

By Simon Watkins, senior manager, WW product marketing, storage and cloud data services, HPE

Hewlett Packard Enterprise Development LP (HPE) has announced the general availability of release 3 of its GreenLake for Block Storage built on its Alletra Storage MP, delivering an industry’s 1st disaggregated, scale-out block storage with a 100% data availability guarantee. ⁽¹⁾

Powered by a shared-everything storage architecture, and managed via the GreenLake Cloud Platform, this block storage offering brings an on-premises cloud experience, scale, and extreme resiliency and performance to modern mission-critical workloads.

This announcement represents the next step in a multi-step journey to consolidate, standardize and simplify HPE block storage portfolio on a common software and hardware stack – all with a unified cloud management experience provided by the GreenLake platform.

With release 3, we’re taking the next step on this ‘journey to one’ by delivering on the promise of disaggregated storage with new, multi-node switched models that power 2x higher performance ⁽²⁾ and enable 2.5x more capacity ⁽³⁾Additional enhancements include a simpler cloud management experience with expanded AI-based performance reporting and analytics, plus greater cost efficiency via a new 4x data compaction guarantee program. Release 3 also unlocks the full power of NVMe for Ethernet networks by adding NVMe-oF/TCP support to existing FC, NVMe-oF/FC and iSCSi connectivity options.

Let’s take a closer look at how this next-gen HPE block storage platform redefines enterprise storage – removing many of the roadblocks and compromises associated with legacy block storage architectures today.

Reimagining block storage architectures
Storage providers have always strived for new ways to improve the performance, resilience, scalability, and cost efficiency of their architectures. Although certain improvements have been made in recent years, mainly driven by the introduction of solid-state media and de-dupe and compression (DECO) technologies, they have usually represented incremental enhancements rather than substantial breakthroughs in architectural design.

Legacy storage architectures typically consist of a pair of HA controller nodes connected to dedicated drives via a midplane within a single hardware chassis. The midplane is often a limiting factor and determines the maximum performance that can be achieved from the system. Traditionally, to scale performance, you would have to purchase a new system with another pair of HA controller nodes, even if you didn’t need the extra capacity. Beyond creating an ‘orphan capacity’ issue, this approach creates silos that are complex to manage and maintain. It also requires that applications have an awareness of where data resides. The result is a storage infrastructure that is unable to flexibly and cost-effectively scale-out to keep pace with the growing and unpredictable demands of traditional, modern, and consolidated workloads.

Now imagine a storage architecture in which the chassis midplane is flexible and delivers more than enough IO/s to eliminate bottlenecks. This storage architecture enables to add incremental controllers for more performance, independently and as needed, providing additional drives and JBOFs to deliver more capacity for your applications. It’s a storage architecture in which storage SSDs are shared by all controllers across a high-speed network, so you have no silos and the ability to tolerate multiple node failures.

That is what constitutes release 3 of GreenLake for Block Storage built on the company’s Alletra Storage MP.

It’s disaggregated, scale-out, shared-everything storage that reimagines block storage. The disaggregation is made possible by standardized composable building blocks, comprising separate compute (controller nodes) and capacity (NVMe expansion shelfs/JBOFs), interconnected by a high-speed, redundant NVMe-oF back-end switch fabric.

Click to enlarge

The evolution from a traditional node pair design to a modular, scale-out, disaggregated architecture built on the firm’s Alletra Storage MP enables significant scale, performance, and availability benefits.

Easily and cost-effectively scale to meet evolving workload demands with disaggregated storage. GreenLake for Block Storage built on Alletra Storage MP leverages a disaggregated, scale-out storage architecture that enables you to scale capacity and performance independently, for greater efficiency and lower cost. This gives user the flexibility to fine-tune, adapt, and elastically scale storage environments in line with workload requirements and evolving business needs. User can start small with cost-efficient, 2-node entry-level configurations. Then scale-up and scale-out to multi-node switched configurations with granular performance and capacity upgrades. And user can maximize ROI via non-disruptive, data-in-place upgrades without repurchasing existing storage.

Release 3 adds support for up to 8xJBOF expansion shelves, enabling user to scale from 15.36TB to around 2.8 PB⁽⁴⁾ without disruption. Support for 8 to 24 SSDs per enclosure, with the ability to upgrade drives in 2-drive increments and JBOFs in increments of one, empowers user to maximize efficiency and cost savings via granular capacity upgrades.

Accelerate mission-critical apps with a shared-everything architecture. GreenLake for Block Storage built on Alletra Storage MP leverages a parallel, all-NVMe, multi-node, all-active architecture with I/O processing active on all media, controllers, and host ports. This ensures consistent, predictable performance and ultra-low latency, at scale.

With release 3, we’re introducing new 2-controller node and 4-controller node switched models with a choice of 16-core or 32-core configurations – adding to the existing option of 2-controller node switchless models and driving a leap in performance of up to 2x.^v
Run any application without compromise with a 100% data availability guarantee. GreenLake for Block Storage built on Alletra Storage MP is an always-on service that’s built on an AI-driven, disaggregated, no single-point-of-failure platform to guarantee unrivalled 100% data availability for your mission-critical apps. Release 3 enables predictive performance even with the loss of multiple nodes in the cluster.

Greenlake for Block Storage Data Ops Manager

Simplify management with AI-driven cloud management experience on-premises
GreenLake for Block Storage built on Alletra Storage MP delivers a cloud operational experience that’s powered by AIOs for infrastructure. This reduces firefighting by predicting and preventing disruptions before they occur. Administrators can rely on AI-driven recommendations to take the guesswork out of managing data infrastructure while eliminating time-consuming, frustrating escalations via predictive support automation and direct access to experts.

With release 3, the company continue to eliminate the guesswork for customers when it comes to storage optimization via enhanced and extended AI-based performance reporting and analytics that improve troubleshooting and insights. Key reporting enhancements include headroom utilization trend analysis by volume set, top volume set hotspots by latency, improved workload drift detection, and resource contention detection. The firm have also improved capacity reporting, capacity savings, and efficiency metrics, and we’re introducing new sustainability metrics, including power consumption trends.

Watch a demo of the advanced performance reporting features in Data Services Cloud Console on the GreenLake platform.

Store more for less with the HPE StoreMore Guarantee
GreenLake for Block Storage built on Alletra Storage MP offers several data reduction and data compaction technologies to reduce storage capacity utilization, including advanced de-dupe with express indexing, an efficient compression algorithm, data packing, hardware accelerated and reservation-less Thin Provisioning, and Virtual Copy. Together these always-on technologies change the economics of flash without any performance trade-offs and have been designed to help user get the most out of user system’s flash capacity while improving flash media endurance.

With release 3, we are backing this up with a 4:1 data compaction ratio guarantee, calculated as a ratio of usable to effective capacity, in aggregate, across all reducible data. If not, the company will provide expertise related to data efficiency, additional storage capacity needed to reach the targeted data efficiency, or other compensation, at our discretion. The guarantee is valid for the term of the support contract.

This Store More Guarantee offers assured efficiency for user workloads, reducing physical storage costs, saving energy, and decreasing datacenter footprint. The firm also offers customized guarantees, valid for a year, for customers requesting a guarantee for a set amount of effective capacity.

Redefining mission critical storage with unique block storage architecture
At the end of the day, user’ organization must focus on eliminating the legacy storage complexity that continues to slow you down, and redefine user mission-critical storage to meet the performance, scale, manageability, availability, and cost efficiency demands of your traditional and cloud native workloads.

That’s precisely what makes GreenLake for Block Storage built on Alletra Storage MP so compelling. Unique among block storage offerings, it simplifies storage management and reduces risk with a disaggregated, shared-everything, scale-out storage architecture; an AI-driven cloud experience; and guaranteed 100% data availability – all of which frees your organization to dramatically accelerate user’s digital transformation journey.

⁽¹⁾ HPE Storage Substantiation
⁽²⁾Compared to existing 2 node (16-core/32-core) switchless configurations
⁽³⁾Compared to release 2 of Greenlake for Block Storage built on Alletra Storage MP iv Fully loaded configuration (8xJBOFs) with 192×15.36TB NVMe SSDs
⁽⁴⁾Compared to existing 2 node ( 16-core/32-core) switch less configurations