I want you, dear reader, to stop and pay close attention.  I’ve said it before, I will say it again:

SDS and HCI architectural approaches are now ready for the majority of x86 workloads.  

There is a sustained space for external purpose built storage platforms (think SAN, NAS, Object appliances) for certain workloads (low in count, but high in value) where there is a need for:

• specific data services (can be anything – SRDF replication behaviors, inline dedupe that is a machine, protocol support) that are not available in the SDS/HCI worlds.
• capacity densities that cannot be met with general-purpose x86 server building blocks.
• latencies that are very VERY low (think consistently <1ms) or workloads that are very sensitive to latency jitter (think hundreds of microseconds) where distributed storage stacks that use ethernet fabrics don’t cut it.

But… Recognize: that second grouping is a subset (and important subset!) of workloads.

Say it with me everyone: SDS and HCI architectural approaches are now ready for the majority of x86 workloads.  

This means that every (!) customer should start to evaluate SDS/HCI models – and then find the workloads (very important ones) that are NOT a fit.

Look – the SDS/HCI market is dwarfed by the external SAN/NAS/Object appliance markets – but that shouldn’t be confused with technical fit.   It’s a function of inertia.

That’s why having the strongest SDS portfolio (ScaleIO and vSAN – and extending for Isilon, ECS, and others) for Dell Technologies is so important.   It’s also why we will continue to double down on making those SDS stacks available in the 3 consumption forms: software-only, software bundled with PowerEdge, and in the form of turnkey systems (which manifest as HCI Appliances – like VxRail, HCI Rack-Scale systems – like VxRack).

Citibank was at Dell EMC World sharing their ScaleIO story.  85PB deployed.   Full production.  Running massive numbers of workloads.   Hundreds of millions of dollars saved by their on-premises cloud, running on ScaleIO.   Not my words – the customers words.

It’s NOT about being low cost/GB (though it is very compelling).  It’s about starting small, and growing as you need.   It’s about being ORDERS OF MAGNITUDE easier to scale, pool, share, automate.  It’s about never needing to do a migration ever again.  It’s about being able to tap into the hardware ecosystem innovation – FAST.

That’s why ScaleIO is so important for Dell Technologies (VMware and Dell EMC, specifically).  

vSAN is great for customers who are all about VMware, and who want an HCI operational model – all the time.  Awesome.   On stage – AIG talked about how this is the way forward for them. 

ScaleIO is for people who want a Server SAN model – something that replaces their SAN, and matches and beats it’s operational model: supporting VMware and non-VMware, scaling/sharing/pooling storage independently of compute.   On stage – Citi talked about how this is the way forward for them.

Customers are different – we uniquely support both.

That’s not my opinion – it’s the company opinion, and our strategic position.

This is why ScaleIO 3.0, announced at Dell EMC World is so important – for 7 reasons.

1) More Effective Usable Capacity

ScaleIO 3.0 introduces multiple space efficiency features including inline compression, space- efficient thin provisioning and snapshots to maximize storage investment.

2) Performance and Acceleration Using Dell PowerEdge 14G and NVMe Drives

ScaleIO 3.0 leads the software-defined storage market in usage of the latest Dell 14G servers as Ready Nodes, including advanced performance and metadata acceleration using NV-DIMMs and NVMe drives.

3) Balance Cost and Performance with Seamless Volume Migration

Seamless volume migration in ScaleIO 3.0 simplifies storage operations by providing the flexibility to easily rearrange and optimize data placement across storage pools and protection domains. This for example enables easy movement of volumes between flash-only, hybrid, and HDD-only pools.

4) Simplify VMware Deployment with vVols Support

ScaleIO 3.0 introduces full VMware vVols support enabling software-defined storage to be managed at a per-VM level which provides a better granularity of data services and a simplified way to manage VMs.

6) Boost Data Copy Management with Improved Snapshot Functionality

ScaleIO 3.0 increases storage efficiency and extends the use of snapshots by enabling the creation of more snapshot copies, automating snap management and adding unrestricted refresh or restore capabilities.

7) Streamline Provisioning and Management of ScaleIO Ready Nodes, and VxRack FLEX Systems

New Automated Management Services (AMS) in ScaleIO 3.0 delivers simple and complete automated lifecycle management for hardware and software when deploying ScaleIO with RHEL 7 on a physical storage node.   AMS is used in both ScaleIO Ready Nodes (ScaleIO lifecycle, hardware reporting, and OS imaging), and also in VxRack FLEX systems (ScaleIO lifecycle and plugs into a broader M&O for the full stack) that incorporate full system design, ToR for Spine-Leaf fabric multi-cabinet scaling and more.

ScaleIO 3.0 GA is targeted towards late this year.

Customers – if your storage partner (even if it’s us – who have the industry leading arrays!) is continuing to push arrays on you without helping you evaluate which workloads are a fit for SDS and HCI models – point them to my blog post, and ask for more.   Dell EMC field… If you’re trying to win over a new customer using a competitor – if you’re not leading with the most disruptive thing we have, it’s becoming as skill-testing question, an IQ test.

Still not sure?   Read Storagereview here, and here.   Still not sure?   Wow, you’re a skeptic 🙂  Fine, we’ve dropped the gauntlet.   Accept our challenge!   You can download and try ScaleIO and find out for yourself, right here.

Are you a ScaleIO customer?  How is it going?