Around the Storage Block

HPE Nimble Storage dHCI - Technical Overview

We just wrapped up an absolutely fantastic week at HPE Discover! It was a week full of sessions, demos, and technical conversations – not to mention a product announcement or two. It would have been hard to miss that we announced our new disaggregated HCI solution which is appropriately named: HPE Nimble Storage dHCI.

My colleague David Wang wrote a really good overview of HPE Nimble Storage dHCI in his recent blog post over on Around the Storage Block. To add to that, today I would like to cover the technical aspects of the solution in a bit more detail.

The solution uses a variety of components to offer the simplified management experience of HCI along with the flexibility and independence of scale associated with a converged infrastructure. Let’s go over each of these components.



On the storage layer, you can pick and choose between HPE Nimble Storage Hybrid-Flash or All-Flash. This initial release of HPE Nimble Storage dHCI is offered as an iSCSI based solution running on 10GbE.

Since it’s an HPE Nimble Storage array, you get all the great features you’ve come to expect such as:

  1. Inline deduplication and compression
  2. Data Replication
  3. Integration with HPE Cloud Volumes
  4. Encryption at rest
  5. 99.9999% of measured data availability


On the server side, dHCI leverages the latest versions (Gen10) of both the HPE Proliant DL360 and DL380. The previous generation (Gen 9) of both these server models is also supported. The servers can be customized to your heart’s content – there aren’t any prepackaged bundles that you are forced to choose from. As long as it’s one of these two models, configure it with any of the available features and options and it can be used as part of dHCI.  Did I mention flexibility?


We support VMware ESXi 6.5 or 6.7


Stack Setup:

Stack Setup is the tool you’ll use to set up a new dHCI cluster. It’s a built-in part of NimbleOS and runs directly on the array, giving you the tools you need right out of the box. After the Nimble array initialization, you will be redirected to Stack Setup automatically to initialize the dHCI cluster. One of the primary objectives of Stack Setup is to ensure that best practices are used across the server, VMware and Nimble storage solution components. After the deployment, you don’t have to go through a number of whitepapers to check that your environment uses the correct best practices, we take care of that as part of the deployment.

Stack Manager

Stack Manager is our vCenter plugin. This plugin offers visibility into the HPE ProLiant servers and the HPE Nimble Storage array. You can manage traditional VMFS datastores along with VVol datastores, meaning that you can create, delete, grow, clone and snapshot from that plugin. Particularly when using VVols, there are some extremely granular clone/recovery options for individual VMs and even individual virtual disks within a VM. You can expand your vSphere cluster by adding new HPE ProLiant servers directly from within the plugin! It’s as simple as:

  1. Plug it
  2. Discover it
  3. Add it

VMware vCenter

VMware vCenter is the proverbial single pane of glass in dHCI. With the help of the above-mentioned plugin, you have control over all of the elements in the dHCI stack from directly within vCenter.

Network switches:

The HPE Nimble Storage dHCI stack uses two redundant top-of-rack switches that have 10Gbit/s connectivity. We have qualified a number of different switch models such as the HPE FlexFabric 5700, 5710, 5940, 5945, Aruba 8325 and all HPE StoreFabric M-Series models.

HPE InfoSight:

In addition to the fabulous integration which HPE InfoSight already has at the storage, server, and hypervisor level, InfoSight has added two new dashboards exclusively for HPE Nimble Storage dHCI. These dashboards provide:

A centralized view of all the HPE Nimble Storage dHCI clusters in your environment

A unique dashboard that offers a deeper view regarding the CPU, memory and storage usage.

Eliminate guesswork with regards to planning and scaling and easily determine the wellness of your solution at a glance!

As you can see, the HPE Nimble Storage dHCI solution offers a lot of great features and flexibility. To extend that flexibility, we have two deployments options.

Deployment using new HPE ProLiant servers

This deployment is characterized by the fact that you buy a new HPE Nimble Storage array, new HPE ProLiant server and optionally new network switches from HPE.

At the HPE Factory we will preload the correct vSphere ESXi image (6.7) using a build that’s bespoke for the HPE ProLiant series, do some customization, and ship the solution.

Once you’ve racked and cabled the components and done the initial network configuration, you can start with the HPE Nimble Storage array initialization. After array initialization, you will be redirected to Stack Setup which will guide you through the rest of the setup.


Deployment using existing HPE ProLiant servers

So what if you want the simplicity and flexibility of HPE Nimble Storage dHCI for your infrastructure, but you’ve already purchased the DL360 or DL380 servers and don’t currently have additional compute requirements which would compel you to purchase more? Not a problem! We can work with your existing servers, provided that they are Gen9 or Gen10 and DL360 or DL380.

In this deployment, you already have the servers and we ship you the HPE Nimble Storage array. Once you’ve racked and cabled your shiny new array, you may have to go through a process of network conformance meaning that you may have to create VLANs and assign them to the correct ports. After network conformance, you can go over the HPE Nimble Storage array initialization and once it completes, you will be redirected to Stack Setup.

The difference here is that instead of creating a new VMware vSphere cluster, like using new hardware, you will choose an existing cluster where you would like to integrate your HPE Nimble Storage array. Remember how dHCI puts best practices in place automatically? Here, you’ve already installed the servers, so we’ll just check your work a bit.  In cases where the best practices requirements are not yet met, we’ll guide you through putting them in place.  Once done, you can move forward with the rest of the deployment.

Of course, if you wanted to add new servers to your existing batch, we can do that too!


That’s enough for a brief technical introduction but as you can see, HPE Nimble Storage dHCI offers the simplified management experience of HCI with the disaggregated scaling of a converged solution.

Stay tuned - in my next blog post, I will do a technical deep dive of all deployment options.

0 Kudos
About the Author


Fred specializes in HPE Converged Solution and HPE Nimble Storage dHCI solution.