Nutanix NCA cheat sheet

Nutanix Cloud Platform combines AOS distributed storage, AHV virtualization, Prism management, and optional services such as Flow, Calm, Files, and Objects.

Hyperconverged infrastructure places compute, storage, and virtualization in the same node; scale-out adds nodes rather than expanding a central storage array.

Prism Element manages one Nutanix cluster; Prism Central manages multiple clusters and centralizes categories, policies, RBAC, and dashboards.

AOS runs as the Controller VM on every node and pools all local disks into a single distributed storage fabric.

AHV is a KVM-based Type-1 hypervisor that uses the cluster distributed storage fabric for VM disks instead of external SAN or NAS.

Nutanix clusters can run AHV or VMware ESXi; the chosen hypervisor determines some management and upgrade workflows.

Nutanix Calm provides application automation and self-service; Flow provides networking and security; Files provides SMB/NFS; Objects provides S3-compatible storage.

Nutanix licensing is subscription-based and feature/capacity dependent; Prism Central is a separate deployment with its own requirements.

A Nutanix cluster is a group of nodes sharing a distributed storage fabric; each node runs a hypervisor and a Controller VM.

The Controller VM handles storage I/O, metadata, replication, and cluster services; VM I/O is redirected to the local CVM and then to the DSF.

AOS services include Stargate for I/O, Curator for scans, Cassandra for metadata, Zookeeper for cluster configuration, and Genesis for cluster management.

The cluster VIP provides Prism Element access and is separate from individual Controller VM and hypervisor management IP addresses.

RF2 keeps two data copies and tolerates one node or disk failure; RF3 keeps three copies and tolerates two failures.

Adding a node triggers automatic data rebalancing across the cluster; no manual LUN expansion or disk group creation is required.

All nodes need consistent network connectivity and reachable CVM and hypervisor management addresses for cluster formation and operation.

A cluster can be expanded online; Prism Element and NCC verify health before and after expansion.

AHV VMs are managed through Prism Element for one cluster or Prism Central for multiple clusters; Acropolis is the AHV management service.

AHV high availability restarts protected VMs on surviving hosts after a host failure, subject to available capacity and admission control.

Affinity policies control placement: VM-host affinity pins VMs to hosts; VM-VM anti-affinity keeps selected VMs on separate hosts.

Nutanix Guest Tools must be installed in the guest OS to enable VSS quiescing, self-service restore, and file-level restore.

VM snapshots are space-efficient metadata snapshots; application consistency requires NGT/VSS or guest quiescing.

A VM clone can be created from a snapshot or existing VM; clones share unchanged data blocks until modified.

AHV live migration moves a running VM between hosts in the same cluster using the distributed storage fabric; no external shared storage is required.

Prism Central categories can drive VM placement, protection, and security policies across managed clusters.

Deleting a VM removes its configuration; snapshots and clones may retain referenced data blocks until expired.

The distributed storage fabric pools SSD and HDD across nodes into one datastore; VM disks are distributed as extents.

A storage container defines policies such as replication factor, compression, deduplication, and encryption; VMs consume containers through datastores.

RF2 stores two copies of data; RF3 stores three; usable capacity is reduced by replication factor and other overheads.

Erasure coding EC-X reduces capacity overhead for cold data by striping data and parity across nodes; it does not replace RF for hot data.

Inline compression reduces data before writing; deduplication removes duplicate blocks; both are policy-controlled per container.

Snapshots are metadata-based and space-efficient; they do not copy the full VM disk at creation time.

Volume Groups present block storage over iSCSI to external hosts or VMs that require raw block devices.

Nutanix Files provides SMB and NFS file services; Nutanix Objects provides S3-compatible object storage.

Hybrid nodes tier hot data on SSD and cold data on HDD; all-flash nodes keep data on SSD.

AHV uses Open vSwitch on each host; virtual networks map to VLANs and connect VM NICs to physical uplinks.

The default AHV bridge is br0; it carries host management and CVM traffic and can be VLAN-tagged.

Nutanix Flow provides distributed firewall and microsegmentation; policies are applied at the virtual NIC level using categories.

VLANs segment traffic; AHV must be configured consistently with physical switch access or trunk ports.

Link aggregation LACP can provide uplink redundancy and throughput; the physical switch must match the configuration.

The cluster VIP, CVM IPs, and hypervisor management IPs must be reachable and correctly placed on the management network.

Flow Network Security can isolate VM groups without changing physical network segmentation.

AHV supports multiple VLAN-backed networks on a virtual switch; a VM NIC is assigned to one network at a time.

A protection domain groups VMs and defines snapshot/replication schedules and retention; it is the basic local backup unit.

Asynchronous replication sends scheduled snapshots to a remote cluster; RPO equals the schedule interval.

NearSync replication provides lower RPO, measured in seconds, than asynchronous replication.

Metro Availability provides synchronous replication across two sites in a stretched cluster with near-zero RPO.

Nutanix Leap automates DR with runbooks, testing, and one-click failover and failback.

Local snapshots enable fast restore by clone or in-place revert; remote snapshots enable site-level recovery.

Replication requires a remote cluster with sufficient capacity and network connectivity; container RF affects capacity.

Protection domain schedules can be hourly, daily, or weekly; retention sets how many snapshots are kept.

Leap runbooks define VM boot order and network changes during DR failover.

Prism Element Health shows cluster health, alerts, and tasks; Prism Central aggregates health across managed clusters.

NCC (Nutanix Cluster Check) runs automated health checks and should be run before upgrades and during troubleshooting.

Alerts can send email, SNMP traps, or syslog; alert policies set severity and recipients.

Pulse sends telemetry to Nutanix support; it can be enabled or disabled and does not affect cluster operation.

Performance metrics include IOPS, latency, throughput, CPU, memory, and storage capacity; Prism charts show historical trends.

Critical alerts require immediate action; warnings indicate degraded but operational conditions.

Health checks cover disk status, CVM service status, replication status, and network connectivity.

Logbay collects logs for support; nCLI provides command-line access to cluster services.

Prism supports local users and AD/LDAP authentication; RBAC roles determine what users can view or change.

Prism Central RBAC can grant roles across multiple clusters; Prism Element roles apply to a single cluster.

Categories are key-value pairs used to group VMs for policy, protection, and security.

Data-at-rest encryption can use self-encrypting drives and/or software encryption; key management must be configured.

Flow Network Security enforces microsegmentation policies at the VM NIC level; policies can be monitored or enforced.

Least-privilege roles and separation of admin duties reduce risk; the default admin account should be secured.

Audit logs record Prism user actions; syslog can forward logs to a SIEM.

Role assignments combine a user or group with a role and a scope; scope determines which clusters or objects are affected.

LCM (Lifecycle Manager) automates inventory, compatibility checks, and firmware/software upgrades for Nutanix and partner components.

AOS upgrades are rolling and non-disruptive; one node is upgraded at a time while the cluster remains available.

Before an upgrade, run NCC and verify LCM compatibility; unresolved health issues can cause upgrade failure.

Prism Central has its own upgrade path and must be compatible with managed Prism Element clusters.

AHV host upgrades can be performed through LCM; ESXi upgrades use VMware tools and must follow interoperability matrices.

Foundation images or re-images nodes; LCM handles ongoing firmware and software updates.

One-click upgrades in Prism use LCM; the upgrade plan shows prerequisites and affected components.

After an upgrade, verify cluster health, alerts, and service status; NCC can be rerun.