Skip to content

VCP-DCV Section 3 - Installation, Configuration, and Setup

Deploying ESXi and vCenter, then configuring networking and storage.

ESXi installation

  • Hardware requirements - a supported 64-bit x86 CPU with hardware virtualization enabled in firmware, the minimum RAM for the release, and a supported boot device. Always verify against the VMware Compatibility Guide; unsupported hardware is the first thing to check in a failed-install scenario.
  • Interactive installation - from ISO via ILO/iDRAC virtual media or USB.
  • Scripted installation - a kickstart file (ks.cfg) supplied by boot option, for repeatable deployment at scale.
  • Auto Deploy - PXE-boots stateless hosts from an image profile held by vCenter. Suits large, uniform estates.
  • Boot device - USB and SD cards are deprecated for the OSData partition in current releases; persistent local storage is expected.
  • DCUI (Direct Console User Interface) - the console interface for initial configuration: management IP, hostname, DNS, root password, and restart of management agents.
  • Lockdown mode - restricts direct host access so administration goes through vCenter. Normal lockdown permits DCUI access for defined users; strict lockdown disables the DCUI entirely, which can lock you out if vCenter is unavailable.

vCenter Server deployment

  • Two-stage deployment - stage 1 deploys the appliance OVA, stage 2 configures SSO and services.
  • Deployment size - tiny through x-large, chosen by host and VM count. Undersizing causes performance problems that look like other faults.
  • Target - deployed onto an ESXi host or an existing vCenter.
  • SSO domain - create a new one or join an existing for Enhanced Linked Mode.
  • DNS - forward and reverse records must exist and resolve correctly before deployment. DNS misconfiguration is the leading cause of failed vCenter deployments and is a reliable exam answer.
  • NTP - time synchronization is required; certificate validation and SSO token exchange fail with skew.

Networking configuration

Standard switch

  • Created per host, with port groups for VM traffic and VMkernel adapters for host traffic.
  • VMkernel adapter services - management, vMotion, Fault Tolerance logging, vSAN, provisioning, and NFS/iSCSI.
  • VLAN tagging modes - EST (external switch tagging, VLAN 0), VST (virtual switch tagging, the common choice, a VLAN ID on the port group), and VGT (virtual guest tagging, VLAN 4095, the guest handles tags).

Distributed switch

  • Created in vCenter, hosts added as members, with uplink port groups mapping physical NICs.
  • Distributed port group - policy applied consistently across all member hosts.
  • Features - private VLANs, NetFlow, port mirroring, LACP, Network I/O Control, and health check.
  • Rollback and recovery - vSphere reverts a network change that disconnects a host from vCenter, which is why VDS misconfiguration is usually recoverable.

Policies (set on switch or port group; port group overrides switch)

  • Security - promiscuous mode, MAC address changes, and forged transmits. All should be Reject unless a specific need exists, such as a nested lab or an IDS appliance.
  • Teaming and failover - load balancing method (originating virtual port, source MAC hash, IP hash, or physical NIC load on VDS), network failure detection, notify switches, and failback.
  • IP hash requires a correctly configured static EtherChannel or LACP on the physical switch. Choosing IP hash without switch-side configuration breaks connectivity, and that mismatch is examined.
  • Traffic shaping - average bandwidth, peak bandwidth, and burst size. Egress only on VSS; ingress and egress on VDS.

Storage configuration

  • iSCSI - software adapter, dependent hardware adapter, or independent hardware adapter. Configure targets (static or dynamic discovery), CHAP authentication, and port binding for multipathing.
  • Fiber Channel and FCoE - zoning and LUN masking are done on the fabric and array.
  • NFS - mount an export; NFS 4.1 supports multipathing and Kerberos, NFS 3 does not.
  • Multipathing (PSA/NMP) - path selection policies are Fixed, Most Recently Used (MRU), and Round Robin. Round Robin is the common default for active/active arrays.
  • Datastore creation - VMFS6 on a LUN, or mounting NFS.
  • Datastore cluster - a group of datastores managed by Storage DRS.
  • Storage I/O Control (SIOC) - datastore-wide fairness under contention using shares and limits.

VM creation and configuration

  • Virtual hardware version - determines available features; upgrading requires a supported ESXi version and a power cycle.
  • Virtual disk controllers - LSI Logic, and PVSCSI (paravirtual) for high-I/O workloads. NVMe controllers for very high performance.
  • Network adapters - VMXNET3 is the paravirtual default and requires VMware Tools. E1000e is emulated and used when drivers are unavailable during installation.
  • CPU and memory hot add - allows increases while powered on, enabled per VM in advance, with guest OS support required.
  • VM options - boot options, firmware (BIOS or EFI), and advanced parameters.
  • OVF and OVA - open virtualization packaging for import and export.

Choosing paravirtual devices (VMXNET3, PVSCSI) gives better performance but depends on VMware Tools being installed, which is the trade-off the exam tests.

Exam pointers

  • Verify hardware against the VMware Compatibility Guide first in installation problems.
  • DNS forward and reverse records must exist before vCenter deployment.
  • VST (a VLAN ID on the port group) is the normal tagging mode; VGT uses VLAN 4095.
  • Security policies should be Reject unless a documented exception applies.
  • IP hash load balancing requires matching physical switch configuration.
  • VMXNET3 and PVSCSI need VMware Tools.
  • Strict lockdown mode disables DCUI access entirely.

Official documentation

📖 vSphere installation and setup - ESXi and vCenter deployment 📖 vSphere networking - switches, port groups, policies 📖 VMware Compatibility Guide - supported hardware