A VxRail consists of a set of server nodes that are designed and engineered for VxRail. A VxRail physical node starts as a standard Dell PowerEdge server. The Dell PowerEdge server next goes through a manufacturing process following VxRail product engineering specifications to produce a VxRail node ready for shipment. The set of components manufactured into VxRail nodes is based on the customer purchase order. The set of VxRail nodes is delivered ready for data center installation and connectivity into the data center network infrastructure.

Once the data center installation and network connectivity are complete, and the equipment is powered on, the VxRail management interface is used to perform the initialization process, which forms the final product: a VxRail cluster. 

A VxRail cluster starts with a minimum of two nodes and can scale to a maximum of 64 nodes. The selection of the VxRail nodes to form a cluster is primarily driven by planned business use cases, and factors such as performance, capacity, and network connectivity.

The naming standards for VxRail models are structured to target specific objectives:

Table 1.      Previous VxRail model naming conventions

Starting in 2023, VxRail evolved to a new model naming standard to align more closely with PowerEdge naming standards. This new naming standard uses a two-character prefix instead of a single character, and consolidates three target objectives from the previous standard under a single model series. 

Table 2.      New VxRail model naming conventions

Each VxRail model series offers choices for network connectivity. Every VxRail node will include support for base connectivity through on-board Ethernet ports. Network expansion is supported on VxRail nodes using available PCIe slots for Ethernet adapter cards. The number and type of PCIe slots that can be used for networking purposes depends on the model selected and the hardware configuration ordered.

Figure 1.   Back view of VxRail E- and VE-Series nodes

With the 1U rackmount models, such as the VE-Series and E-Series, PCIe slots can be used for network expansion. Certain models will support both low-profile PCIe slots and full-height PCIe slots. The number and type of slots on each node will vary depending on the hardware options configured, such as whether to include support for GPUs.

Figure 2.   Back view of VxRail VP-, V-, P-, and S-Series nodes

The 2U models offer a higher number of PCIe slots in contrast to the 1U models. However, the 2U rackmount models can be configured at the time of ordering to address a wider range of use cases, which can reduce the number of available PCIe slots. For instance, network expansion options would be reduced in cases where those slots instead are populated with storage devices or GPUs.

Figure 3.   Back view of VxRail VD-Series rackable and stackable chassis options

The VD-Series support sleds are 1U and 2U in size. Both sled options are supported in either a rackable or stackable chassis. Only the 2U sleds support network expansion outside of the on-board Ethernet ports through PCIe slots.