Dell EMCPowerFlex Systems with VMware Cloud Foundation Reference Architecture
Back in September 2020, DellEMC released a white paper / reference architecture for customers that might want to combine VMware Cloud Foundation (VCF) with the DellEMC PowerFlex platform.
VCF is the foundational building block, the core SDDC infrastructure that can be used to create an environment that other products and services can easily integrate with. VCF makes the life of a data center engineer and architect much easier, because when you deploy VCF, you get a documented design and you know where everything is- meaning you know all the IP addresses, physical and logical locations and integrations that make up the system.
This is turn allows an organisation to "plug in" any external services or products that are compatible with the platform. VCF is a building block for the modern SDDC.
In modern data centers, we use a variety of products and services from various vendors. Each having their own use case. One of the more interesting products out their is the PowerFlex series of products available from DellEMC. But what is PowerFlex?
PowerFlex (previously VxFLEX) systems create a server and IP based SAN (an alternative to traditional SANs) by combining storage virtualization software, with Dell EMC PowerEdge servers to deliver flexible, scalable performance, and capacity on demand. PowerFlex combines SSD and NVMe media to create virtual pools of block storage with varying performance tiers. PowerFlex is essentially a software defined storage product, placing it in the hyperconverged infrastructure (HCI) universe.
As we discussed in previous posts, vSAN is a hard requirement for VCF. The key to this is that vSAN is required in the MGMT WLD, but you are free to utilize other storage types in the VI WLDs starting in VCF 4.X.
The key difference between PowerFlex and vSAN is that vSAN is exclusively a vSphere product and can only run with the ESXi hypervisor. PowerFlex allows you to run non-VMware hypervisors with software defined storage, such as RedHat Enterprise Linux (RHEL), CentOS or Microsoft`s Hyper-V hypervisors. Starting with VCF 4.0, you can add PowerFlex as supplemental storage to your VCF environment. This increases the amount and type of workloads that you can manage with VCF.
DellEMC has release a technical white paper here, which details the solution: https://infohub.delltechnologies.com/section-assets/powerflex-storage-vcf-supplementalstorage-3
New terms introduced with VCF 4.X are "principle" and "supplemental" storage. The difference between the two storage types is with the automation and deployment.
To quote the white paper:
"The terms “principal” and “supplemental” simply indicate when the storage is configured with the workload domain. The characteristics of these storage types have no relation to features or performance profiles, as both can be equally resilient and performant. Principal storage (vSAN, NFS, VMFS on FC) is used to initially deploy a workload domain, while supplemental storage is attached after the workload domain is deployed.Supplemental storage including PowerFlex, iSCSI, vVols, and NFSv4 are attached later using standard vCenter management tools".
The conceptual view of the solution:
And a detailed view showing the placement of the supplemental storage along with VCF workload domains:
This post was meant to inspire engineers to think beyond the "core" functionalities provided by VCF, and to look into other systems and products that might work well with this solution. There are many solutions available, and I will be documenting these as I continue down the VCF journey.