NCP-CN-V6-10: Nutanix Certified Professional - Cloud NativeDemo

Prometheus collects and retains the time-series metrics used to analyze Kubernetes and platform-service resource utilization. As cluster workloads and the desired retention period grow, Prometheus needs higher container CPU and memory capacity for metric ingestion and querying, along with more persistent storage for retained metrics. NKP sizing guidance specifically provides Prometheus resource settings and storage values that increase with cluster size and workload.

NKP autoscaling is horizontal node scaling: Cluster API (CAPI) creates worker nodes when the cluster needs additional capacity and deletes nodes when they are no longer needed. It does not hot-add or hot-remove CPU or memory, nor merely change a node’s power state.

NKP Insights provides AI-driven Kubernetes analytics and anomaly detection with root-cause analysis, making it the component intended to identify anomalous cluster behavior proactively.

NKP can attach and centrally manage existing CNCF-conformant Kubernetes clusters. Grouping the 30 attached edge clusters in an NKP Workspace provides a single management scope for deploying and lifecycle-managing applications across the fleet, using the available kubeconfig credentials. NKP Ultimate includes the centralized fleet-management capabilities needed for this approach.

Konvoy Image Builder creates the vSphere VM template from a base OS image and installs the artifacts required for Kubernetes cluster nodes. It is the NKP-supported image-building tool for vSphere.

Rook Ceph provides the in-cluster persistent and S3-compatible object storage used by Velero for backup data. NKP/DKP documentation identifies rook-ceph as required when Velero uses the platform’s built-in storage; it is unnecessary only when Velero is explicitly configured with external cloud object storage.

For Nutanix infrastructure using CAPX, NKP supports Rocky Linux and Ubuntu operating-system images. CentOS and Flatcar are not supported platforms for this deployment method. The NKP supported-infrastructure matrix lists Rocky Linux and Ubuntu under Nutanix.

Velero uses velero backup describe <backup-name> to display details of a specific backup. Supplying testbackup targets the backup with that exact name.

For an AWS cluster managed by Nutanix NKP, the control plane must be updated to the target Kubernetes version using the AWS control-plane update command and a compatible AMI. Worker node pools are upgraded separately with the AWS node-pool upgrade command so their nodes roll to the target version.

Konvoy Image Builder requires an x86_64-based Linux or macOS operator machine. This host platform is a documented prerequisite for running the builder and producing compatible cluster images.

NKP uses kind during bootstrap and runs docker ps -a to list container-based kind clusters. A failure of that Docker command indicates that a Docker-compatible container runtime is unavailable or not running.

The pre-provisioned infrastructure provider (CAPPP) initializes Kubernetes control-plane and worker nodes on pre-existing hosts defined in a previously created inventory YAML file. The NKP pre-provisioned cluster command therefore requires a --pre-provisioned-inventory-file path.

nkp create appdeployment creates an AppDeployment in a workspace. Without an explicit cluster list, it targets all existing clusters in that workspace. The platform application must be identified by its application ID and version, and the target NKP workspace must be specified.

The NKP CLI creates a bootstrap cluster with nkp create bootstrap. Its --kubeconfig flag specifies the path to the kubeconfig for the resulting management cluster, so setting it to bootstrap-cluster.conf recreates the context file at that name.

Deleting an NKP cluster is performed from that cluster’s three-dot actions menu by selecting Delete. Detaching a cluster only removes its association with the management interface; it does not delete the cluster or free the worker capacity for reuse.

NKP uses the Kubernetes namespace supplied at cluster creation to place the workload cluster in its corresponding workspace. Omitting --namespace code-green causes the cluster resources to be created in the default workspace namespace instead, so the cluster is not shown in the code-green workspace.

NKP requires an Infrastructure Provider in the target workspace so it has the credentials and infrastructure configuration needed to provision a new Kubernetes cluster. After the provider is created for the workspace, the Create Cluster workflow becomes available.

An NKP management cluster must be able to communicate with the target cluster’s Kubernetes API server and services so it can establish and maintain management of the attached cluster. Exact Kubernetes-version matching, an N-1 version relationship, and creation of an NKP management-cluster administrator account on the target cluster are not required attachment prerequisites.

A bastion host provides controlled SSH access to protected systems, so SSH keys are required for secure key-based remote access. Enabling NTP keeps the host’s clock synchronized, supporting accurate audit logs and time-sensitive authentication. Docker and LDAP server software are not required for bastion-host operation.

Before provisioning an NKP cluster in FIPS mode, the operating system or OS images must be prepared for FIPS operation according to the OS vendor’s instructions. FIPS cannot be applied to an existing cluster; it must be enabled when creating a new cluster.

Kommander Project Continuous Deployment (CD) provides GitOps-based deployment from a source-code repository to the clusters associated with a project. It continuously reconciles repository changes, allowing a release merged to the production branch to be automatically deployed to both the Production and DR environments. Continuous Integration handles build and test workflows; secrets and ConfigMaps only supply credentials or configuration. D2iQ: Project Deployments

For AWS-based kOps configuration, the AWS region is obtained from the AWS_REGION environment variable when it is not obtained from EC2 instance metadata. Setting AWS_REGION supplies the specified us-east-1 deployment region.

AlertManager centralizes alerts for operational failures such as CrashLooping workloads and node problems, while NKP Insights proactively identifies cluster security and configuration anomalies. NKP Insights uses kube-bench to assess cluster security posture against CIS benchmarks, making the combined tools appropriate for continuous health monitoring and security/compliance detection.

NIB and KIB build machine images that include the components and configuration required for Cluster API to provision and bootstrap NKP cluster nodes. The resulting image is CAPI-compliant and intended for use by NKP node virtual machines.