Liqid Are Dynamic In The DC

Disclaimer: I recently attended Dell Technologies World 2019.  My flights, accommodation and conference pass were paid for by Dell Technologies via the Media, Analysts and Influencers program. There is no requirement for me to blog about any of the content presented and I am not compensated in any way for my time at the event.  Some materials presented were discussed under NDA and don’t form part of my blog posts, but could influence future discussions.

As part of my attendance at Dell Technologies World 2019 I had the opportunity to attend Tech Field Day Extra sessions. You can view the videos from the session here, and download my rough notes from here.

 

Liqid

One of the presenters at Tech Field Day extra was Liqid, a company that specialises in composable infrastructure. So what does that mean then? Liqid “enables Composable Infrastructure with a PCIe fabric and software that orchestrates and manages bare-metal servers – storage, GPU, FPGA / TPU, Compute, Networking”. They say they’re not disaggregating DRAM as the industry’s not ready for that yet. Interestingly, Liqid have made sure they can do all of this with bare metal, as “[c]omposability without bare metal, with disaggregation, that’s just hyper-convergence”.

 

[image courtesy of Liqid]

The whole show is driven through Liqid Command Center, and there’s a switching PCIe fabric as well. You then combine this with various hardware elements, such as:

  • JBoF – Flash;
  • JBoN – Network;
  • JBoG – GPU; and
  • Compute nodes.

There are various expansion chassis options (network, storage, and graphics) and you can add in standard x86 servers. You can read about Liqid’s announcement around Dell EMC PowerEdge servers here.

Other Interesting Use Cases

Some of the more interesting use cases discussed by Liqid included “brownfield” deployments where customers don’t want to disaggregate everything. If they just want to disaggregate GPUs, for example, they can add a GPU pool to a Fabric. This can be done with storage as well. Why would you want to do this kind of thing with networking? There are apparently a few service providers that like the composable networking use case. You can also have multiple fabric types with Liquid managing cross composability.

[image courtesy of Liqid]

Customers?

Liqid have customers across a variety of workload types, including:

  • AI & Deep Learning
    • GPU Scale out
    • Enable GPU Peer-2-Peer at scale
    • GPU Dynamic Reallocation/Sharing
  • Dynamic Cloud
    • CSP, ISP, Private Cloud
    • Flexibility, Resource Utilisation, TCO
    • Bare Metal Cloud Product Offering
  • HPC & Clustering
    • High Performance Computing
    • Lowest Latency Interconnect
    • Enables Massive Scale Out
  • 5G Edge
    • Utilisation & Reduced Foot Print
    • High Performance Edge Compute
    • Flexibility and Ease of Scale Out

Thoughts and Further Reading

I’ve written enthusiastically about composable infrastructure in the past, and it’s an approach to infrastructure that continues to fascinate me. I love the idea of being able to move pools of resources around the DC based on workload requirements. This isn’t just moving VMs to machines that are bigger as required (although I’ve always thought that was cool). This is moving resources to where they need to be. We have the kind of interconnectivity technology available now that means we don’t need to be beholden to “traditional” x86 server architectures. Of course, the success of this approach is in no small part dependent on the maturity of the organisation. There are some workloads that aren’t going to be a good fit with composable infrastructure. And there are going to be some people that aren’t going to be a good fit either. And that’s fine. I don’t think we’re going to see traditional rack mount servers and centralised storage disappear off into the horizon any time soon. But the possibilities that composable infrastructure present to organisations that have possibly struggled in the past with getting the right resources to the right workload at the right time are really interesting.

There are still a small number of companies that are offering composable infrastructure solutions. I think this is in part because it’s viewed as a niche requirement that only certain workloads can benefit from. But as companies like Liqid are demonstrating, the technology is maturing at a rapid pace and, much like our approach to on-premises infrastructure versus the public cloud, I think it’s time that we take a serious look at how this kind of technology can help businesses worry more about their business and less about the resources needed to drive their infrastructure. My friend Max wrote about Liqid last year, and I think it’s worth reading his take if you’re in any way interested in what Liqid are doing.

Western Digital Are Keeping Composed

Disclaimer: I recently attended Storage Field Day 18.  My flights, accommodation and other expenses were paid for by Tech Field Day. There is no requirement for me to blog about any of the content presented and I am not compensated in any way for my time at the event.  Some materials presented were discussed under NDA and don’t form part of my blog posts, but could influence future discussions.

Western Digital recently presented at Storage Field Day 18. You can see videos of their presentation here, and download my rough notes from here.

 

Getting Composed

Scott Hamilton (Senior Director, Product Management) spoke to the delegates about Western Digital’s vision for composable infrastructure. I’m the first to admit that I haven’t really paid enough attention to composability in the recent past, although I do know that it messes with my computer’s spell check mechanism – so it must be new and disruptive.

There’s Work To Be Done

Hamilton spoke a little about the increasingly dynamic workloads in the DC, with a recent study showing that:

  • 45% of compute hours and storage capacity are utilised
  • 70% report inefficiencies in the time required to provision compute and storage resources

There are clearly greater demands on:

  • Scalability
  • Efficiency
  • Agility
  • Performance

Path to Composability

I remember a few years ago when I was presenting to customers about hyper-converged solutions. I’d talk about the path to HCI, with build it yourself being the first step, followed by converged, and then hyper-converged. The path to Composable is similar, with converged, and hyper-converged being the precursor architectures in the modern DC.

Converged

  • Preconfigured hardware / software for a specific application and workload (think EMC Vblock or NetApp FlexPod)

Hyper-Converged

  • Software-defined with deeper levels of abstraction and automation (think Nutanix or EMC’s VxRail)

Composable

  • Disaggregated compute and storage resources
  • Shared pool of resources that can be composed and made available on demand

[image courtesy of Western Digital]

The idea is that you have a bunch of disaggregated resources that can be really used as a pool for various applications or hosts. In this architecture, there are

  • No physical systems – only composed systems;
  • No established hierarchy – CPU doesn’t own the GPU or the memory; and
  • All elements are peers on the network and they communicate with each other.

 

Can You See It?

Western Digital outlined their vision for composable infrastructure thusly:

Composable Infrastructure Vision

  • Open – open in both form factor and API for management and orchestration of composable resources
  • Scalable – independent performance and capacity scaling from rack-level to multi-rack
  • Disaggregated – true disaggregation of storage and compute for independent scaling to maximise efficiency, agility snd to reduce TCO
  • Extensible – flash, disk and future compassable entities can be independently scaled, managed and shared over the same fabric

Western Digital’s Open Composability API is also designed for DC Composability, with:

  • Logical composability of resources abstracted from the underlying physical hardware, and
  • It discovers, assembles, and composes self-virtualised resources via peer-to-peer communication.

The idea is that it enables virtual system composition of existing HCI and Next-generation SCI environments. It also

  • Future proofs the transition from hyper-converged to disaggregated architectures
  • Complements existing Redfish / Swordfish usage

You can read more about OpenFlex here. There’s also an excellent technical brief from Western Digital that you can access here.

 

OpenFlex Composable Infrastructure

We’re talking about infrastructure to support an architecture though. In this instance, Western Digital offer the:

  • OpenFlex F3000 – Fabric device and enclosure; and
  • OpenFlex D3000 – High capacity for big data

 

F3000 and E3000

The F3000 and E3000 (F is for Flash Fabric and E is for Enclosure) has the following specification:

  • Dual-port, high-performance, low-latency, fabric-attached SSD
  • 3U enclosure with 10 dual-port slots offering up to 614TB
  • Self-virtualised device with up to 256 namespaces for dynamic provisioning
  • Multiple storage tiers over the same wire – Flash and Disk accessed via NVMf

D3000

The D3000 (D is for Disk / Dense) is as follows:

  • Dual-port fabric-attached high-capacity device to balance cost and capacity
  • 1U network addressable device offering up to 168TB
  • Self-virtualised device with up to 256 namespaces for dynamic provisioning
  • Multiple storage tiers over the same wire – Flash and Disk accessed via NVMe-oF

You can get a better look at them here.

 

Thoughts and Further Reading

Western Digital covered an awful lot of ground in their presentation at Storage Field Day 18. I like the story behind a lot of what they’re selling, particularly the storage part of it. I’m still playing wait and see when it comes to the composability story. I’m a massive fan of the concept. It’s my opinion that virtualisation gave us an inkling of what could be done in terms of DC resource consumption, but there’s still an awful lot of resources wasted in modern deployments. Technologies such as containers help a bit with that resource control issue, but I’m not sure the enterprise can effectively leverage them in their current iteration, primarily because the enterprise is very, well, enterprise-y.

Composability, on the other hand, might just be the kind of thing that can free the average enterprise IT shop from the shackles of resource management ineptitude that they’ve traditionally struggled with. Much like the public cloud has helped (and created consumption problems), so too could composable infrastructure. This is assuming that we don’t try and slap older style thinking on top of the infrastructure. I’ve seen environments where operations staff needed to submit change requests to perform vMotions of VMs from one host to another. So, like anything, some super cool technology isn’t going to magically fix your broken processes. But the idea is so cool, and if companies like Western Digital can continue to push the boundaries of what’s possible with the infrastructure, there’s at least a chance that things will improve.

If you’d like to read more about the storage-y part of Western Digital, check out Chin-Fah’s post here, Erik’s post here, and Jon’s post here. There was also some talk about dual actuator drives as well. Matt Leib wrote some thoughts on that. Look for more in this space, as I think it’s starting to really heat up.