vSAN Hyperconverged Infrastructure

Dell EMC PowerEdge MX, the First Kinetic Infrastructure, is now a vSAN Certified ReadyNode

VMware vSAN, the largest HCI (Hyper-Converged) software vendor in the market, is always first in enabling new platforms. This enablement helps our mutual customers adopt this modern server architecture for their workloads (traditional and cloud-native) and operational management benefits. Dell EMC PowerEdge MX, the kinetic infrastructure on modern server architecture, is now certified for vSAN. This is a huge milestone towards helping our mutual customers run traditional and next generation workload.

You can find the listing of Dell EMC PowerEdge MX on vSAN VCG here and here.

What is the Dell EMC PowerEdge MX?

Dell EMC PowerEdge MX, the industry’s newest high performance, modular infrastructure, designed to support a wide variety of traditional and emerging data center workloads. PowerEdge MX offers the first modular infrastructure architecture designed to easily adapt to future technologies and server disaggregation.

With its unique kinetic infrastructure, customers can break free from the bounds of technology silos and time-consuming, routine operational management while also dynamically assigning IT to optimally match different applications and needs.

Dell EMC PowerEdge MX – Front

To understand the layout from simplicity perspective, the slots can be populated either with compute or storage sleds. Each compute sled can contain up to 6 drives and each storage sled can contain up to 16 drives that can be consumed for vSAN storage.

Why Dell EMC PowerEdge MX on vSAN?

In the past, I have discussed in a blog about composable infrastructure and how it meets the HCI need. There are many benefits and scenarios where composability is of great value to our mutual customer. Indeed, many of our mutual customers are running their vSAN for various workloads on such infrastructure. Dell EMC PowerEdge MX fits well into this. I would like to reiterate some of those key points and few more additional benefits PowerEdge MX brings in.

1. Scale out compute and storage independently:

Companies are becoming more dynamic and it is hard to plan and predict the nature of workload the business will require. They demand a platform which can grow dynamically (both compute and storage).The biggest advantage of this platform is storage and compute can scale independently. The PowerEdge MX is 7U chassis “in-a-box” and depending on your workload needs, you can either scale compute or storage. Each compute sled can hold up to 6 drives and the storage sled can hold up to 16 drives. This is a huge benefit in today’s business environment. It not only saves upfront expenditure by avoiding over-provisioning but provides the flexibility of independent scaling; e.g. if a new workload demands only compute, you can scale compute sleds in the frame and vice versa.

Indeed, you can build up to an 8-node vSAN cluster within a frame (7U chassis) and extend it by adding additional PowerEdge MX frames as business demand grows.

2. New network switches capability from SDDC perspective:

Network Switches are the next level of convergence in the HCI space.  The Dell EMC PowerEdge MX, provides the agility, resiliency, and efficiency to optimize a wide variety of traditional and new, emerging data center workloads and applications. Designed from the ground up for high performance and low latency, the MX networking architecture provides line rate 25GbE speeds, perfect for Software Defined Storage (SDS) technologies such as VMware vSAN to scale and provide best in class IO and redundancy for all data center needs. Supporting Open Networking along with OS10 Enterprise Edition with SmartFabric Services, PowerEdge MX provides fabric automation capabilities such as topology validation and automated QoS.

The MX Scalable Fabric Architecture builds on the scale-out architecture of the SDDC model supporting up to ten MX7000 Chassis @25GbE per server port with no over-subscription in a single network domain. The Scalable Fabric Architecture also supports heterogeneous rack servers as well, up to a total of 104 devices. This multi-chassis architecture enables any and all SDDC use cases for today, tomorrow and far into the future.

 Key capabilities:

  • MX9116n Fabric Switching Engine – 16 x 25GbE server facing ports, 2 x 100GbE/8 x 32G FC unified ports, 2 x 100GbE ports, and 12 Fabric Expansion ports
  • MX7116n Fabric Expander Module – 16 x 25GbE server facing ports and 2 Fabric Expansion ports
  • MX5108n Ethernet Switch – 8 x 25GbE server facing ports, 2 x 100GbE ports, 1 x 40GbE port, and 4 x 10GBase-T ports
  • MXG610s Fibre Channel Switch – 16 x 32G FC internal ports, 8 x 32G FC SFP+ ports, and 2 QSFP 4 x 32G FC uplink ports

3. Supported on VMware Cloud Foundation (VCF):

VMware Cloud Foundation is the unified SDDC platform that brings together VMware’s vSphere, vSAN and NSX into a natively integrated stack to deliver enterprise-ready cloud infrastructure for the private and public cloud. It simplifies deployment of SDDC and Hybrid Cloud with built-in automated life-cycle management capabilities.

Since the PowerEdge MX has been certified for vSAN ReadyNode, it is also qualified to run VCF which allows our customers to adopt this new generation infrastructure for their hybrid cloud journey.

4. All NVMe Support:

If we look at evolution of server-side storage in last decade, flash devices such as SAS/SATA SSD replaced traditional magnetic media (HDD) to great extent. Almost all if not all, enterprise workloads today run on all flash. This became possible with server-side economics brought in by HCI.

A similar scenario is evolving within flash storage with NVMe devices. Many customer have begun standardizing on all NVMe in their data center as it brings in better TCO considering diversified nature of their workloads. Very soon, we will start realizing that the next wave of device refresh going to be by NVMe devices.

The Dell EMC PowerEdge MX can support all NVMe vSAN today for customers to start on this journey. The next generation of PowerEdge MX release is planning for supporting even more NVMe drives than its current supported configuration. The table below shows how the customer today can consume all NVMe vSAN for their NVMe deployment for each node.

5. Multi-path Support:

One of the key reasons for customers adopting modular servers in HCI deployment is multipath support. It provides continuous access to storage sled within the chassis even when a path is down. Without multipath support, compute will lose the access to storage sled. The multi-path support is limited to SAS devices in the storage sleds only in the current portfolio of PowerEdge MX.

Multi-path SAS Drives in Storage Sled

As you can see the storage sled has two paths reaching to single drive. In case one of the SAS module goes down, the SAS module can serve the IO requests. One thing to keep in mind that this is active/passive meaning both paths are not used at the same time for IO. The second path is used only when a fail over situation occurs.

6. Boot Device in Compute Sled:

Many use cases demand boot drives to be in front slot. This is very useful from PowerEdge MX-series perspective. User can put the boot device in compute sled in a separate controller and can configure the entire vSAN from storage sled which is mapped from the backplane and presented to the compute sled as local drives. This makes boot device and vSAN datastore path completely separate from each other.

Example vSAN Deployment Choices for PowerEdge MX

Below, I would like to highlight few example scenarios on how to deploy vSAN on MX-series. This is not exhaustive list and customer can deploy in many other ways. But it would help understand how PowerEdge MX can be used flexibly to consume vSAN for their different workload.

One Frame with 1x Storage Sled (12 drives) and 3x Compute Sled (2 drives):

Each of these compute sleds are configured with 2 drives with two disk group (1 cache + 2 capacity) per server. The optimal choice would be to use SAS drive as cache tier and SATA drives as capacity tier.

One Frame with 1x Storage Sled (16 drives) and 7x Compute Sled (4 drives):

Each of these compute sleds are configured with four drives within two disk groups (1 cache + 2 capacity) per group. The optimal choice would be to use NVMe drive as cache tier and SATA drives capacity tier.

One Frame with 5x Storage Sled (16 drives) and 3x Compute Servers (No drives):

The entire storage is served from storage sled and there is no compute sled storage slot been used. The optimal choice would be to use SAS drive as cache tier and SATA drives as capacity tier.

One Frame with No Storage Sled and 8x Compute Servers (6 drives):

This could potentially an All-NVMe vSAN where cache and capacity both could be NVMe device.

I would put below table as an example for three possible combinations for all flash vSAN deployment on PowerEdge MX-series

Examples of vSAN Deployment Choices

Just to reiterate again, vSAN requires only three compute servers to begin with. The above configurations are just examples and not exhaustive list. As this new generation hardware gets deployed more and more, we will learn more about how users like to configure it.

What Use Cases is Not Meant for Today?

While the Dell EMC PowerEdge MX covers many HCI use cases within a single infrastructure, there are a few cases where it may not be a good candidate.

  1. NVMe device on Storage Sled: In the current release PowerEdge MX supports NVMe devices only on compute sled and not in storage sled. Though you can deploy all NVMe vSAN today on MX-series but the NVMe devices need to be on compute sled only.
  2. Multi-path not available on NVMe devices: The current generation of PowerEdge MX does not support multipath in storage as well as compute sled.


As organizations adopt this new generation HCI platform to run their traditional and next generation applications together, platforms like PowerEdge MX is a great fit. We will continue enabling as new features and functionality of PowerEdge MX are made available.

For any questions on vSAN hardware, please reach out to vSAN Hardware PM vsan-hcl@vmware.com.

To learn more about vSAN, visit VMware vSAN

To learn more about Dell EMC PowerEdge MX, visit Dell EMC PowerEdge MX

To learn more about Dell EMC PowerEdge MX vSAN deployment guide, visit vSAN ReadyNode Deployment Guide


2 comments have been added so far

  1. I have doubt, what is the difference between compute sled and storage sled. By looking at the names I understand that compute sled is for boot devices and storage sled for vsan disks.

    Then what is this means for creating DG for compute sleds?

  2. What Use Cases is Not Meant for Today?
    As organizations adopt this new generation HCI platform to run their traditional and next generation applications together, platforms like PowerEdge MX is a great fit

Leave a Reply

Your email address will not be published. Required fields are marked *