The New HPE Alletra Storage MP – The Era of Standardization

There’s the adage that variety is the spice of life, but, as usual, too much of anything can be a challenge.

Take, for example, storage hardware at any company that has multiple storage portfolio offerings.

Almost always, the hardware is different between the various offerings, sometimes extremely so. Not just in performance or capacity, but also in quality and reliability. And in many cases, the hardware is so dramatically different that there can be absolutely no sharing – the drives are different, the controllers are different, the chassis are different, the HBAs… maybe even the screws 😊 (lately I have been trying to keep things generic and not name names, but a prime example is Pure – FlashBlade is incredibly different vs FlashArray//X which is incredibly different vs FlashArray//C which is incredibly different vs FlashArray//XL).

At HPE we are taking major steps to standardize the storage solution hardware to provide a better customer experience.

In this blog I will discuss the new hardware and briefly touch upon the new software offerings that will all run on that same hardware – as usual, I will try to provide information and insights that may not always be found in official announcements or the product pages.

Alletra Storage MP: The New Underlying Hardware Platform for HPE Storage

Make no mistake – what I’m about to describe isn’t a small feat, it is something that’s unique in the storage industry.

When designing the standardized MP platform, we wanted to ensure it would be suitable for multiple kinds of applications, for now and the future. Using the exact same set of building blocks for File, Object and Block workloads, at any scale – plus perhaps some much more exotic things… I think of “MP” as Multi-Purpose or Multi-Persona.

We wanted to use common hardware as much as possible, and not have to rely on custom ASICs or media. All the supply chain issues companies faced the past few years have been more than enough trouble for everyone.

Upgradability between generations was also extremely important, as was coexistence of different generations.

The ability to start small and scale to crazy big was essential.

Being able to repurpose hardware? Why not? Of course, if you ask a Dell EMC customer that way overbought PowerScale if they could convert that surplus hardware to a PowerMax, or a Pure customer that overbought FlashBlade if they could convert a surplus chassis to a FlashArray//C, or a NetApp StorageGrid customer if they could convert that to ONTAP… they will look at you like you’re a crazy person, while secretly wishing they could do just that kind of conversion. But for them it’s not even in the realm of possibility.

The quality of the hardware had to be uniform – the same high-end hardware quality should be used for both the ultra-fast versions of Tier-0 platforms with 100% uptime guarantees as well as their entry versions and everything in between. If the only difference is performance and capacity, there is no need to sacrifice any reliability for the less expensive offerings.

Things like a plethora of PCIe lanes, huge memory throughput and high core density are important for performance: we standardized on AMD CPUs for the new hardware.

Low power consumption and a tighter footprint are also very important and help with sustainability efforts, so we used extremely efficient components all around, as well as going to OCP-format expansion slots.

And drawing from existing HPE server parts and know-how was a given – HPE sells hundreds of thousands of servers each quarter after all. Storage controllers (even if you add up all storage vendors together) are a tiny fraction of that number. Taking advantage of server economies of scale was a no-brainer.

Some other benefits of this approach include:

• Easier logistics
• Easier and faster development
• Single hardware field training
• Easier, more consistent installations
• Easier supportability
• Easier distribution
• Easier to stock parts
• Lower inventory costs
• Easier to develop automation for (anything from the management plane to much more complex automation)
• Can help lower costs due to economies of scale: if everything is based on the same stuff, then you buy more of the same stuff… we already see this with things like CPUs and RAM, but it takes more than that to build complete systems.
• Resistant to supply chain issues: Much easier to predict how much of each component is needed, and stock well in advance – and if something becomes truly hard to find, the lack of “special” components makes it far easier to source alternatives.

The Alletra Storage MP Building Blocks

The various MP building blocks allow multiple kinds of architectures to be built.

At a basic level, there are just two items: a chassis and (optionally) dedicated cluster switches.

The chassis can become just compute, or compute with storage in it.

After that, things can be combined as needed.

For example, the JBOF (Just a Bunch Of Flash) expansion shelf is itself built like a complete standalone system: in the MP chassis, it has storage media and two “controllers”, each with an OS, RAM and CPU. It’s just that the expansion slots, CPUs and memory used are minimal, just enough to satisfy JBOF duties.

Storage media is flexible – SSDs, SCM and NVDIMMs are all supported by the architecture currently. So I guess it’s not “just” a bunch of flash, but the JBOF name is here to stay 😊

Increase the JBOF compute power and expansion slots, and now you can use that same architecture as a compact storage array with two controllers and storage inside the same MP chassis.

For bigger capacity and performance needs, a disaggregated architecture can be used, connecting diskless compute to JBOFs in large clusters.

The HPE GreenLake for File offering is essentially the last combination mentioned: A disaggregated, shared-everything (DASE) cluster, using diskless compute nodes, 100Gb RDMA networking, and JBOFs to create a complete solution.

The HPE GreenLake for Block offering on the other hand, is starting as a smaller, stand-alone switchless system, with later iterations adopting the same disaggregated paradigm to build much larger solutions.

The following figure should make it all very clear (courtesy of our very own Pete Mattei, or StorePete as he likes to be called):

I will cover the Block and File offerings in more detail in separate posts. Go here for the Block MP post, the File post is coming.

Summary

With Alletra Storage MP, HPE has created a paradigm shift in storage hardware standardization. This will help customers achieve a better experience, with less waste, more flexibility, more predictability, more reliability – at any scale.

D