Nutanix NCP-CI-AWS flashcards

32 cards. Read the term and the definition here. The deck at the top of this page is optional.

NC2 bare-metal requirement

NC2 on AWS runs AHV only on bare-metal EC2 instance families such as i3.metal, because nested virtualization and direct NVMe access are required. Standard virtualized instances are not supported.

Minimum NC2 cluster size

A production NC2 cluster needs at least three nodes so that RF2 can tolerate a single node failure. Fewer nodes cannot satisfy resiliency requirements.

Prism Central placement on NC2

Prism Central runs as a separate EC2 instance in the same VPC as the cluster it manages, not as a VM inside the AHV cluster.

NC2 deployment orchestration

The NC2 console provisions clusters by creating CloudFormation stacks, IAM roles, and bare-metal EC2 instances inside the customer's own AWS account and VPC.

NC2 data plane ownership

All workload data stays in the customer's AWS account and VPC. Nutanix hosts only the management console and licensing metadata, not customer VM data.

Bare-metal quota prerequisite

Before deployment, verify the AWS account has sufficient bare-metal vCPU quota in the target region and availability zone. Exhausted quota is a frequent cause of failed builds.

Single AZ cluster scope

An NC2 cluster lives entirely within one AWS availability zone. Cross-AZ resilience requires a second cluster plus asynchronous replication between them.

Guest VM traffic path

AHV NATs guest VM traffic out through the node's elastic network interface, so AWS security groups and network ACLs filter all VM traffic.

Hybrid connectivity options

On-premises to NC2 connectivity uses AWS Site-to-Site VPN or Direct Connect attached to a virtual private gateway or transit gateway. NC2 does not create these attachments.

Subnet IP capacity planning

The cluster subnet must hold node IPs, Prism Central, and the IPAM pool assigned to guest VMs. Exhaustion blocks VM creation and cluster expansion.

Egress and NAT gateway

NC2 needs outbound internet or proxy access for licensing, Prism registration, and updates. A NAT gateway with a stable elastic IP supports allowlisting.

Reserved AWS subnet addresses

AWS reserves the first four and last IP address of every subnet. The AHV cluster virtual IP must not be assigned from these reserved addresses.

Prism Central authentication sources

Prism Central accepts local users, Active Directory or LDAP directories, and SAML 2.0 identity providers. Roles map to users or directory groups.

Prism projects for isolation

Projects bind VMs, users, and roles together so tenants see only their assigned resources. They provide multi-tenant isolation within one Prism Central.

NC2 IAM role dependency

NC2 relies on IAM roles and instance profiles created at deployment. Deleting or rotating them breaks snapshots, scaling, and CloudFormation updates.

Flow Network Security scope

Flow enforces distributed firewall rules inside AHV based on VM categories, names, or subnets. AWS security groups still filter at the ENI layer independently.

CloudTrail audit coverage

AWS CloudTrail records API calls made by NC2 service roles in the customer account, complementing Prism Central task and user audit logs.

LCM upgrade role

Lifecycle Manager inventories AOS, AHV, and firmware versions and performs rolling upgrades. Only NC2-validated version sets from the compatibility matrix should be applied.

NCC health checks

Nutanix Cluster Check validates cluster health and configuration. Run it regularly and resolve failures before upgrades or disaster recovery tests.

Scaling an NC2 cluster

Adding nodes is done from the NC2 console, which provisions additional bare-metal instances. Scale-down must keep the cluster above minimum node count and RF requirements.

Ephemeral instance store risk

NC2 storage uses local NVMe instance store tied to the EC2 instance. Stopping or terminating nodes destroys data unless snapshots or replication protect it.

Idle cluster cost behavior

Bare-metal instances bill hourly regardless of VM activity. Idle NC2 clusters remain expensive, so capacity and cost reviews should include AWS Cost Explorer.

Protection domain purpose

Protection domains group VMs so snapshots are crash-consistent across the group. Local snapshots remain on the same cluster and do not survive cluster loss.

Async replication RPO

Asynchronous replication sends snapshots to a remote NC2 cluster on a schedule. The RPO equals the interval between replication runs.

Nutanix Leap recovery plans

Leap orchestrates DR with recovery plans defining boot order, IP remapping, and one-click failover to a standby NC2 cluster. Test failover validates plans safely.

Nutanix Move migration

Move migrates VMs from VMware, Hyper-V, or AWS EC2 into AHV on NC2. It requires network reachability and sufficient target cluster capacity.

Leap test failover

Test failover clones protected VMs into an isolated network on the recovery cluster. Production replication continues and no production traffic is disrupted.

NearSync replication tradeoff

NearSync achieves RPO in seconds but demands low-latency links and specific licensing. Cross-region NC2 DR typically uses standard asynchronous replication instead.

Cluster build failure triage

Check the NC2 console task log and CloudFormation stack events first. Bare-metal quota, IAM permission errors, and subnet IP exhaustion are the leading causes.

Prism registration failure checks

When Prism Central cannot register a cluster, verify outbound NAT or proxy access, DNS resolution, and routability between the cluster VIP and Prism Central IP.

EC2 instance status check

A failed EC2 instance status check indicates host-level AWS failure. The node must be replaced rather than repaired from inside the guest.

Support log collection

Gather Prism log bundles and NCC output, then include the NC2 cluster ID and AWS region in the support case. Add CloudTrail logs for AWS-layer issues.