In a VxRail deployment where the "Predefined" virtual distributed switch (VDS) configuration is selected alongside a "2x10GbE or Higher" dual-NIC layout, the automated installer applies deterministic network load- balancing and failover policies. Because a 2-NIC profile aggregates all foundational infrastructure traffic- including Management, vMotion, vSAN, and Virtual Machine networks-onto a single pair of physical uplinks (Uplink 1 and Uplink 2), logical segregation is critical to prevent contention and ensure data path reliability.
To achieve this, the predefined architecture utilizes an explicit failover ordering mechanism rather than a standard active/active load-balancing state across the uplinks. For the vSAN port group, which carries high- throughput storage synchronization traffic, the network design configures Uplink 1 as Standby and Uplink 2 as Active. This configuration maps directly to a Standby/Active setup. Conversely, the Management port group operates in an inverse fashion, utilizing Uplink 1 as Active and Uplink 2 as Standby. This layout optimizes bandwidth usage across both physical connections during standard operations, while guaranteeing that full path redundancy remains maintained if an upstream switch link failover event occurs.
References: Dell VxRail Deploy Study Guide; Predefined VDS Network Topologies; Teaming and Failover Policies.

When initializing a VxRail cluster, proper network name resolution is an absolute dependency for communication between the VxRail Manager, ESXi hosts, and the vCenter Server. During the deployment wizard execution, administrators are provided with two distinct architectural pathways for handling DNS services: an external enterprise DNS or an internal DNS service. If the internal DNS option is selected, the VxRail Manager virtual appliance (Choice B) is directly responsible for running and hosting this internal service.
Under the hood, VxRail Manager provisions an embedded, lightweight DNS containerized service to dynamically manage host name resolution across the management network segment. This internal service automatically populates forward and reverse lookup records for all newly initialized ESXi nodes and the vCenter Server, ensuring that deployment and subsequent node additions complete without external infrastructure dependencies. This model is particularly beneficial in isolated environments or dark sites where an external enterprise DNS server is completely unavailable. Neither individual guest virtual machines, the vCenter Server, nor a standalone Platform Services Controller host this automated internal system component.
References: Dell VxRail Deploy Study Guide; VxRail Architecture Overview; Internal DNS Configuration and Rules.