Nutanix NCP-CI-AWS common mistakes

NC2 provisions a Nutanix cluster on bare-metal EC2 instances that are spread across multiple Availability Zones and subnets; each node sits in its own subnet and the cluster needs cross-subnet and cross-AZ connectivity. Candidates who only know on-prem Nutanix assume one L2 segment is enough and mis-answer placement and sizing questions.

NC2 supports specific bare-metal instance families, and the cluster shape (number of nodes, instance type) determines what AWS capacity must be reserved. Candidates carry over N+1 or N+2 habits from on-prem and forget that NC2 capacity is purchased as a subscription tied to specific instance types and regions, so scaling questions and cost questions are answered with the wrong assumptions.

The wizard validates prerequisites but does not create every dependency, and failures at deployment often come from missing permissions or exhausted quotas. Candidates who have only clicked through a demo do not know which permissions the NC2 service needs and answer prerequisite questions vaguely.

Candidates reuse standard three-tier VPC patterns without accounting for the fact that Nutanix cluster replication and Prism Central traffic is east-west and cross-AZ. They also forget that cross-AZ data transfer has cost and latency implications, and that a single NAT gateway is an AZ-scoped failure domain.

NC2 has two overlapping security planes: AWS-native controls on the underlay and Nutanix-native controls (such as flow policies or microsegmentation) on the guest and management traffic. Candidates who learn only one plane answer with the wrong tool and miss that Prism access should be restricted at both the AWS security group and the Nutanix identity layer.

NC2 environments commonly integrate with an external directory, and Prism Central uses role-based access control with scoped roles. Candidates who only used the default admin account cannot answer questions about least privilege, SSO, or which identity provider governs which layer, and they conflate AWS IAM (which governs the NC2 service and infrastructure APIs) with Prism identity (which governs cluster management).

Nutanix upgrades are rolling but still require healthy cluster state, and NC2 operations interact with AWS capacity availability (a new node needs a bare-metal instance in the right AZ). Candidates who have not run an upgrade answer as if it is a single click with no prerequisites and miss questions about prerequisites, health checks, and rollback.

Candidates mix up which layer provides which protection: Nutanix replication and protection domains for guest data, AWS EBS snapshots for the underlying volumes, and application-consistent backups for workloads. They also forget that a DR plan must specify failover order, DNS changes, and how the NC2 cluster itself is recovered.

Many NC2 failures originate below the Nutanix layer: an instance that fails a status check, a missing IAM permission recorded in CloudTrail, a service quota limit, or a security group blocking cluster communication. Candidates who only know the Prism UI miss the AWS evidence and pick the wrong root cause.

NC2 support spans two vendors' boundaries, and the candidate must know what to gather (cluster logs, Prism Central logs, AWS case details, timestamps, affected resource IDs) and when the issue belongs to AWS versus Nutanix support. Candidates without operational experience give vague answers that do not demonstrate the boundary knowledge the exam tests.