VCP-VCF common mistakes
VCP-VCF · Understand the traps and how to avoid them.
Studying VCF as a renamed vSphere stack and assigning vCenter the top-level orchestration role.
vSphere experience makes vCenter feel like the management plane, so candidates miss that SDDC Manager owns inventory, domain lifecycle, credential and certificate rotation, and lifecycle management, while Cloud Builder is only the initial bring-up appliance.
What to do instead
Build a responsibility matrix for Cloud Builder, SDDC Manager, vCenter, NSX Manager, and vSAN, then trace one management-domain bring-up and one workload-domain creation through those owners.
Conflating management domain and workload domain requirements, especially which services are shared and which are per-domain.
Both domains contain vCenter, NSX, and vSAN, so candidates assume they are interchangeable and miss the management domain's special role hosting SDDC Manager and core infrastructure.
What to do instead
Draw the two-domain model with per-domain versus shared components, list what must exist in the management domain before workload domains can be created, and note where NSX Edge clusters are required by design.
Launching Cloud Builder without validating DNS, NTP, VLANs, IP pools, MTU, and credentials against a planning workbook.
The deployment UI encourages field-by-field entry, so candidates treat it as a form-filling exercise and discover dependency failures only after deployment starts.
What to do instead
Create a pre-flight checklist from the VCF planning workbook; test forward and reverse DNS, NTP reachability, VLAN and MTU configuration, IP pool overlap, and credentials before powering on Cloud Builder.
Treating a VCF network pool as one VLAN and one IP range instead of mapping pool type to traffic type.
Candidates carry over a flat vSphere port-group mindset and miss that VCF separates management, vMotion, vSAN, NSX TEP, Edge, and other traffic with distinct VLAN, subnet, and MTU requirements.
What to do instead
Learn each network pool type and its traffic; verify VLAN IDs, gateways, MTU, and non-overlapping subnets end to end; practice assigning pools during workload-domain creation.
Commissioning hosts as if adding them to vCenter, without checking VCF host prerequisites and validation results.
Adding a host in vSphere is familiar and appears equivalent, but SDDC Manager validates hardware compatibility, credentials, network configuration, and version baseline before a host can join a domain.
What to do instead
Study the host commissioning workflow and prerequisites, then practice resolving failed validations before host assignment; know that commissioning is separate from assigning a host to a domain.
Rotating passwords or replacing certificates directly on vCenter, NSX, or ESXi instead of through SDDC Manager.
Component-level tools are familiar, and the direct change appears to work, but SDDC Manager stores credentials and certificate state and out-of-band changes break orchestration and lifecycle management.
What to do instead
Use SDDC Manager credential rotation and certificate management workflows; learn which certificates are managed where and verify component connectivity after each rotation or replacement.
Assuming NSX Manager cluster VIP, node IPs, host TEPs, Edge TEPs, and uplink IPs are interchangeable in the VCF address plan.
NSX uses many similarly named addresses and candidates conflate management, control, overlay, and underlay networks.
What to do instead
Build an address map for a workload domain showing NSX Manager VIP and node IPs, host TEP pool, Edge TEP pool, uplinks, overlay, and MTU; trace how the network pool supplies each range.
Treating lifecycle management as a single upgrade-all button and skipping bundle review, prechecks, and backups.
Candidates memorize that SDDC Manager performs lifecycle management but not that bundles have types, upgrades have supported sequences, and prechecks and backups are mandatory operational steps.
What to do instead
Walk through the lifecycle workflow: check the depot, download bundles, review compatibility, run prechecks, take backups, remediate, then upgrade in supported order; distinguish install, update, and upgrade bundles.
Troubleshooting VCF orchestration failures by starting in ESXi or vCenter logs because those are the most familiar.
The failure often originates in SDDC Manager task validation, DNS or NTP, certificates, or NSX, so component-level logs show only downstream symptoms.
What to do instead
Start with the failed SDDC Manager task and SDDC Manager or lifecycle logs, generate a support bundle, then correlate with vCenter, NSX, ESXi, DNS, NTP, and certificate state.
Assuming vCenter backup is a VCF backup and skipping SDDC Manager and NSX backups before changes.
Candidates equate virtualization backup with full VCF protection and do not realize SDDC Manager, vCenter, and NSX have separate backup responsibilities.
What to do instead
Document backup coverage for SDDC Manager, vCenter, and NSX; take required backups before upgrades or certificate and credential changes; practice a restore validation.