🧊 ToastNix
Home Β· Blueprint

Section 2 Describe Nutanix Platform Administration

trapmemorisedefault valuetiplab Rate yourself: β—‹ not yet Β· ◐ shaky Β· ● know it (saved in this browser)

  1. 2.1 Describe VM tasks
  2. 2.2 Identify virtual networking tasks
  3. 2.3 Discuss storage operations
  4. 2.4 Describe licensing tasks

Objective 2.1 Describe VM tasks

2.1.1 Deploy, update, and configure a VM

Create VMs in Prism Central (VMs β†’ Create VM) or in Prism Element. You set vCPUs, cores per vCPU, memory, disks, NICs, boot type (legacy BIOS or UEFI, optionally Secure Boot), and optional guest customization (Sysprep for Windows, cloud-init for Linux). Windows guests on AHV need the Nutanix VirtIO drivers.

On AHV you can hot-add memory and vCPUs to a running VM, but you can't hot-remove them. VM templates capture a configured VM so you can deploy copies of it.

Images: the Image Service stores ISOs and disk images. You upload them from your workstation or from a remote URL (HTTP). Image Placement Policies use categories to decide which clusters an image lives on.

2.1.2 Configure storage for VMs

A VM's disks are vDisks on a storage container, so they inherit that container's RF and efficiency settings. Disk bus is usually SCSI. You can also attach Volume Groups for shared or clustered disks.

Flash Mode pins a VM's or vDisk's data to the SSD tier. Use it for latency-critical workloads on hybrid nodes.

2.1.3 Configure network attributes

Each NIC attaches to a subnet (a VLAN-backed subnet, or an overlay/VPC subnet). On an IPAM subnet AHV hands out the address itself. Without IPAM, the address comes from your own DHCP server or a static configuration.

2.1.4 Assign categories

Categories are key:value tags such as Environment:Production or AppType:SQL. Policies point at categories rather than at individual VMs. That covers security (Flow), protection and DR, Playbooks, Image Placement, and VM-host affinity. To apply one policy to dozens of VMs, tag them with a category and attach the category to the policy.

TIP

Agent VMs are the AHV setting for "one VM per host that stays on its host", for example a virtual firewall or an antivirus appliance. They aren't migrated and are powered off during host maintenance.

TRAP

VM-VM anti-affinity on AHV, for example keeping two SQL nodes on different hosts, is created with aCLI (acli vm_group.create then vm_group.antiaffinity_set). VM-host affinity is set in Prism.

2.1.5 Change vDisk Placement

Live vDisk migration moves a running VM's vDisks to another storage container without downtime, for example to change RF or efficiency settings. Do it from Prism Central (Move vDisk) or aCLI.

2.1.6 Take a snapshot

In Prism Element you can take a manual VM snapshot (VM β†’ Take Snapshot) and restore or clone from it. In Prism Central these are recovery points. A recovery point can be restored in place (revert the VM) or cloned into a new VM. Scheduled snapshots come from protection policies.

TRY IT IN THE LAB

Snapshot a VM in the CE lab, change a file in the guest, then restore from the recovery point in Prism Central.

Official docs for 2.1 (17)

Objective 2.2 Identify virtual networking tasks

2.2.1 Create subnet

Create subnets in Prism Central (Network & Security β†’ Subnets) or in Prism Element. A VLAN basic subnet maps a VLAN ID on a virtual switch. Turn on IP address management (IPAM) to have AHV assign addresses to VMs automatically, and set the network address, gateway, DNS and IP pools.

2.2.2 Identify types of networking

Types of networking: VLAN basic subnets, managed per cluster with optional IPAM, and overlay (VPC) subnets from Flow Virtual Networking, which are tenant-isolated and need Prism Central. Network segmentation splits traffic types onto separate VLANs or interfaces. Examples are the CVM backplane (storage replication) versus management, plus services such as Volumes or DR.

MEMORISE

To secure traffic between management and storage replication, use network segmentation.

2.2.3 Create vSwitch

AHV networking is built on Open vSwitch (OVS). A virtual switch (the default is vs0) wraps a bridge (br0) and its uplink bond. Create or edit virtual switches in Prism Element (Settings β†’ Network Configuration β†’ Virtual Switch), not on each host by hand. That way the configuration stays consistent across nodes.

Bond (uplink) modes:
- Active-Backup (the default): one link carries traffic and the others stand by. No switch configuration is needed.
- Active-Active with MAC pinning (balance-slb): VMs are spread across the links by MAC. No LACP is needed.
- Active-Active with LACP (balance-tcp): hashes on source/destination IP and TCP/UDP ports. Needs LACP on the switch.

MEMORISE

Load-balancing on IP address and L4 port means Active-Active with LACP.

2.2.4 Describe network visualizer

The Network Visualizer in Prism Element draws hosts, their uplinks, VMs and the connected physical switch ports. It uses LLDP/CDP, plus SNMP if the switch is added. Use it to spot which uplink is active (Active-Backup shows one active link per host) and where a VLAN is missing.

2.2.5 Monitor physical switches

Add the top-of-rack switch in Prism Element (Settings β†’ Network Switch) with its management address and SNMP credentials. Prism then shows that switch's port statistics and errors next to the hosts.

Official docs for 2.2 (22)

Objective 2.3 Discuss storage operations

2.3.1 Identify storage optimizations

The data-efficiency features are all set per container:
- Compression. Inline compression suits almost every workload and is the safe default.
- Deduplication. It pays off where many VMs share the same blocks, such as full clones and VDI.
- Erasure coding (EC-X). It replaces part of the RF overhead with parity for write-cold data and needs a minimum node count, so check the doc.

MEMORISE

The efficiency feature suitable for nearly every workload is compression. The one that replaces traditional RF copies for suitable data is erasure coding.

2.3.2 Identify redundancy requirements

The replication factor is the number of data copies: RF2 = 2, RF3 = 3. It is configured on the container. RF3 needs a cluster that can tolerate two failures (FT2, 5 nodes minimum). RF1 exists for disposable or scratch data and has prerequisites, so read the RF1 page.

The redundancy (fault tolerance) factor is the cluster-level setting. Rebuild capacity reservation holds back free space so the cluster can re-protect its data after a node failure.

TRAP

RF is a container setting. The answer is not "VM" and not "storage pool".

2.3.3 Identify the different storage types

The storage objects:
- Storage pool: all physical disks, usually one per cluster.
- Container: a logical slice of the pool that carries the policies.
- vDisk: the disk a VM sees.
- Extent / extent group: how the DSF lays data out. Extent groups are the files stored on the CVM-owned physical disks.
- Volume Group: iSCSI block storage for VMs or external hosts.

Official docs for 2.3 (15)

Objective 2.4 Describe licensing tasks

2.4.1 Identify current license state

Check license state in Prism Central's Licensing / License Manager page, or on the Prism Element licensing page. It shows each cluster's licensed tier, its add-ons, and whether it is unlicensed, in violation, or expiring.

2.4.2 Assign, reclaim, and unassign licenses

Seamless licensing (connected sites) lets you apply, upgrade, reclaim and unassign licenses from Prism Central with one click, synchronised with the Nutanix portal. Manual licensing (dark sites) uses a cluster summary file that you upload to the portal, then you download and apply the license file.

TRAP

Reclaim the licenses before you destroy a cluster. Otherwise they stay tied to a cluster that no longer exists.

2.4.3 Inspect available licenses

The Nutanix portal's licensing page and Nutanix Central list the licenses you own, which ones are assigned, and which are still available to apply.

2.4.4 Verify licensing expiration

Expiry dates show in License Manager, and Prism raises alerts as term licenses approach their end. License tags help group and track entitlements.

2.4.5 Identify correct license by use case

Licenses come as NCI tiers (infrastructure: AOS, AHV, Prism) and NCM tiers (management: Intelligent Operations, Self-Service, Cost and Security), plus add-ons for things like Files, Objects and NDB. Some products don't need a license at all, so check Products that do not require a license. Pick the tier that includes the feature the scenario needs.

Official docs for 2.4 (11)