Recently, the FlexPod team revamped our FlexPod® Express architecture and updated our documentation with a new design guide and two new deployment guides. Before we start talking about our implementation of VMware vSphere 6.5 and Microsoft Windows Server Hyper-V 2016, let’s take a look at the hardware components.

Storage

We’ve added new hardware components to the FlexPod Express architecture. First, we have the NetApp® FAS2600 storage array running ONTAP® 9.1. In this validation, we used the FAS2650 set up as a 2-node switchless cluster. These 2Us feature plenty of onboard connectivity, and up to 24 SAS drives. In FlexPod Express, we use NetApp ONTAP storage efficiency features like compression, deduplication, and compaction to get the most out of the storage system.

Network

We used the Cisco Nexus 31108 switch in this validation. This was my first time working with this switch, and I can’t find enough good things to say about it. Each switch has 48 ports that support 10G and 1G connectivity, and 6 ports that can support 40G or 100G connectivity. This switch is ready for anything you throw at it today — and tomorrow.

Compute

We use the Cisco UCS C-220 M4 server as the hypervisor node for this design validation.  The UCS C-220 M4 can be configured with the desired CPU and RAM specifications. The FlexPod architecture also supports other Cisco C-Series standalone servers. We use the VIC 1227 to provide connectivity to the hypervisor hosts, which allows us to create logical adapters to provide flexibility for the configuration. For example, we use the VIC 1227 to create dedicated iSCSI boot adapters.

 

So which hardware design reigns supreme — hardware architecture for VMware vSphere 6.5 or for Microsoft Windows Hyper-V 2016? Neither. That’s right. FlexPod Express uses the same hardware architecture for both hypervisors. This way, our customers can choose the hypervisor that meets their business requirements today, and tomorrow. If their requirements change down the road, that’s fine too. FlexPod Express is designed to scale; additional hypervisor nodes can be added as the infrastructure grows.

Deployments

The big difference between the two architectures, besides the hypervisor of course, is the protocol used for serving data. Both hypervisors are booted from iSCSI LUNs. In the case of VMware vSphere we use NFS for data; in the case of Microsoft Windows Server Hyper-V we use iSCSI. This means that if customers want to switch the hypervisor of their FlexPod express, they need to perform only a slight bit of additional storage configuration in either case.

 

Our FlexPod Express deployment guides outline the deployment of both hypervisors:

Because these architectures are so similar, they share a common design guide.  The FlexPod Express design guide gives an overview of the hardware included in the architecture and the software components used for each hypervisor.

 

FlexPod Express is a great choice for remote and branch offices, and for customers who need to provide infrastructure for a dedicated workload.

Melissa Palmer

Melissa Palmer is a solutions architect in the NetApp Infrastructure and Cloud Engineering team. She is also VMware Certified Design Expert (VCDX) #236. Prior to joining the Infrastructure and Cloud Engineering team, Melissa was a systems engineer for NetApp and a VMware engineer for a number of enterprise environments. Melissa has bachelor of engineering and master of engineering degrees from Stevens Institute of Technology.