5V0-31-22 VMware Cloud Foundation Specialist (VCF v2) Practice Questions

The Supervisor Cluster control plane consists of three virtual machines that run the Kubernetes API server, scheduler, and controller manager. Its primary function is to expose the Kubernetes API, allowing DevOps users and administrators to interact with the vSphere environment using standard Kubernetes tools (like kubectl) to manage the lifecycle of Tanzu Kubernetes Grid clusters, vSphere Pods, and namespaces.

NSX Federation architecture is built on a hierarchical model. A single, active Global Manager (GM) instance is deployed to manage global policies (like firewall rules and network segments) that span multiple locations. Each individual VCF site (location) has its own Local Manager (LM) instance that manages the local data plane components (hosts, edges). The Global Manager pushes configuration down to the Local Managers, which then enforce it locally.

After restoring SDDC Manager, its database may be out of sync with the current state of the environment (e.g., passwords rotated, hosts added). SDDC Manager provides a reconciliation workflow that compares its restored configuration with the live environment. This process identifies any drift and allows the administrator to sync the live configuration back into the SDDC Manager database, ensuring consistency.

In the context of vSAN, 'scaling up' refers to increasing the resources within existing hosts, typically by adding more disks to existing disk groups or adding new disk groups. This increases storage capacity and potentially performance. 'Scaling out' refers to adding more ESXi hosts to the cluster, which increases compute (CPU/RAM), storage capacity, and performance simultaneously.

NSX Federation is designed for centralized policy management across multiple sites. Global policies, such as a firewall rule that needs to be enforced consistently in Site A and Site B, are created on the Global Manager. The Global Manager then replicates this policy down to the Local Managers at each site for enforcement.

The Standard Architecture model deploys the management workload on a dedicated management domain, physically separating it from VI workload domains. This meets the compliance requirement. A vSAN-based management domain requires a minimum of 4 ESXi hosts to provide N+1 redundancy (FTT=1). A Consolidated Architecture runs management and compute workloads on the same cluster, which violates the stated requirement.

Cloud Builder validates NTP reachability from the ESXi hosts themselves. If FQDNs are used for the NTP servers, the hosts must be able to resolve these names. During the initial bring-up, ESXi hosts may not have DNS configured until later in the process. Using IP addresses for NTP servers in the Deployment Parameter workbook is the recommended approach to avoid this dependency. The other options are less likely as Cloud Builder itself, not just the ESXi hosts, needs connectivity, but the error is specific to host reachability.

When enabling vSphere with Tanzu, SDDC Manager requires several non-routable CIDR blocks to be defined for NSX-T to manage. The Namespace Network CIDR (for Pods), Ingress CIDR (for services exposed externally via NSX LB), and Egress CIDR (for SNAT'd outbound traffic) are the primary ranges. The Management Network is for the Supervisor Control Plane VMs and is typically a routable network, not a private overlay.

A vSAN Stretched Cluster provides synchronous replication between two sites, achieving an RPO of zero. Combined with vSphere HA, it offers automated failover. The 3ms latency is well within the 5ms RTT requirement for stretched clusters. NSX Federation provides network and security policy consistency across sites but does not handle storage replication or automated VM failover. Site Recovery Manager provides automated disaster recovery orchestration but typically does not achieve RPO=0 without synchronous storage replication.

VMware Cloud Foundation leverages vSphere Lifecycle Manager (vLCM) for managing host compliance. The correct, VCF-integrated approach is to modify the cluster's desired state image within vLCM to include the required driver. After updating the image, the cluster can be remediated to bring all hosts into compliance. Once vLCM reports compliance, the SDDC Manager pre-check will pass. Manual installation (ESXCLI) would cause drift from the desired state image and is not the recommended practice.