This section provides a high-level architecture overview of PowerFlex rack and PowerFlex appliance that describe the VMware vSphere architecture to meet VMware NSX-T requirements.
The following table contains VMware vSphere requirements for PowerFlex rack and PowerFlex appliance .
Note: These requirements are recommendations for
PowerFlex appliance if the
PowerFlex controller nodes or NSX-T Edge nodes are provided by the customer.
| Management cluster
Created as a single VMware vSphere cluster and added within the PowerFlex data center.
- HA and Distributed Resource Scheduler (DRS) enabled at the VMware vSphere cluster level.
- VMware vSAN vSphere cluster enabled, if three or more nodes are present.
- Includes the VMware vSphere core VMs (virtual channels, PowerFlex management controller , and any Dell EMC appliances). The three NSX-T Manager VMs also run in this cluster.
| NSX-T Edge cluster
- Created as a single VMware vSphere cluster and added in the PowerFlex data center.
- HA enabled and DRS disabled at the VMware vSphere cluster level, if running the vSAN storage option.
- Deployed across multiple physical racks (split between two racks).
- By default, local storage with RAID 10 is enabled. However, vSAN is supported if VMware services or the customer chooses this option.
- Includes NSX-T Edge gateway VMs, deployed at the customer site, that provide logical routing.
- At minimum, four of the six NIC interfaces that are used for transport and external edge traffic must be configured as an individual trunk. The other two NIC interfaces that are used for VMware ESXi management or vSAN traffic are configured with Link Aggregation Control Protocol (LACP) enabled vPC.
Note: Do not deploy non-NSX-T Edge workloads in the NSX-T Edge VMware vSphere cluster.
| Compute cluster
- Created as a single VMware vSphere cluster and added within the PowerFlex data center.
- Can be deployed across multiple physical racks.
- HA and DRS are enabled at the VMware vSphere cluster level.
- Includes the customer production VMs.
- Can have more than one compute cluster.
- On a compute cluster, the management distributed switch must be configured without LACP. The switch carries overlay NSX-T ready transport traffic, and VMware does not recommend using LACP.