Built by mapping the 8-module Nutanix Hybrid Cloud Fundamentals (NHCF 7.5) course against a 50-question NCA practice bank. Organised by what actually gets tested, not by course order.
This page decides what to study. ToastNix holds the notes for every blueprint objective, and every module below links straight to them. Blueprint cross-reference → · Hands-on labs → · Flashcards → · Dictionary →
Every one of the 50 practice questions was tagged to the NHCF module that teaches it, and separately to the NCA 7.5 blueprint objective it tests. The counts below are the result.
Three structural facts about this bank, and how they differ from NSE 4:
The real exam is 50 questions in 90 minutes, scored 1000–6000 with 3000 to pass. The timed practice exam below uses the same length and clock.
| Module | Course labs | Qs | Share | Verdict |
|---|---|---|---|---|
| M2 Hardware and Storage Concepts | Exploring the hardware and storage-container dashboards; creating a storage container. | 15 | 30% | Heaviest module by far |
| M1 Introduction | Accessing and exploring Prism Element and Prism Central; deciding when to use Prism Central versus Prism Element. | 9 | 18% | Which Prism, which product |
| M7 Monitoring Cluster Performance and Health | Adding built-in charts to the Analysis dashboard; creating a custom chart; creating, downloading and scheduling reports; monitoring cluster health; collecting logs. | 7 | 14% | Which view shows what |
| M8 Understanding Licensing and Performing Upgrades | Exploring the LCM dashboard; performing inventory at a connected site and at a dark site. | 6 | 12% | All LCM, no licensing |
| M5 VM Management | Exploring the VMs dashboard; creating VMs; creating a recovery point; cloning a VM; exporting a VM as an OVA. | 5 | 10% | Placement and policy |
| M3 Networking | Viewing network configuration in Prism Central; using the Network Visualizer. | 4 | 8% | Bond modes and IPAM |
| M4 Image Management | Exploring the images dashboard; uploading and importing an image to Prism Central; uploading an image using Prism Element. | 3 | 6% | Categories, twice |
| M6 Backup and Disaster Recovery Fundamentals | Creating protection domains and a local VM restore; configuring remote sites; protection domains using remote sites; VM migration. | 1 | 2% | A whole course day for one question |
Module M2 alone is 30% of the bank. M6 takes most of a course day and is worth one question here, though the blueprint tests snapshots under VM tasks.
M2 Hardware and Storage Concepts, M1 Introduction, M7 Monitoring Cluster Performance and Health.
These three are nearly two-thirds of the bank. Storage policy and failure behaviour (M2), which Prism does what (M1), and which monitoring view answers which question (M7). Get these to the point where you answer without thinking.
M8 Understanding Licensing and Performing Upgrades, M5 VM Management, M3 Networking.
LCM workflow, VM placement and the three bond modes. Fewer questions, but each one is a single fact you either know or don’t.
M4 Image Management, M6 Backup and Disaster Recovery Fundamentals.
Images and data protection. Learn the handful of facts and move on.
Also inToastNix 1.1 2.3 3.1 3.2·Cards 1.1 2.3 3.1 3.2
The course covers: Nodes, blocks and clusters; physical storage types; AOS distributed storage; storage pools, containers, redundancy and replication factor, capacity reservation, cluster resiliency, snapshots. Course labs: Exploring the hardware and storage-container dashboards; creating a storage container.
This is the heaviest module by a distance: 15 of 50 questions. Almost none ask you to do anything. They ask where a setting lives, how many copies or nodes a level needs, and what the cluster does on its own when something fails. Q30 is the only exhibit, and it is a disk diagram you read for failure domains.
Trap: “storage pool” is the distractor for every policy question. A cluster normally has one pool holding all the physical disks; it carries no RF or efficiency setting. The container is the logical slice that does.
Create two containers on your CE cluster, one with compression and one without, and open each one’s settings to see exactly which policies live there. Then power off a node and watch the Data Resiliency widget. Lab 2 and Lab 3 walk through both.
Also inToastNix 1.1 1.2 3.1 4.2·Cards 1.1 1.2 3.1 4.2
The course covers: Three-tier vs hyperconverged infrastructure; Nutanix Cloud Solution Packages; getting started with Prism Central and Prism Element; Pulse; when to use which Prism; cluster setup fundamentals; cluster-level security. Course labs: Accessing and exploring Prism Element and Prism Central; deciding when to use Prism Central versus Prism Element.
The bank uses this module for “which Prism” and “which product” questions. If a feature spans clusters (RBAC, categories, reports), the answer is Prism Central. If it is about one cluster’s own console, it is Prism Element on a CVM IP.
Trap: the Foundation VM IP and the AHV host IP both appear as distractors for “how do I reach Prism Element”. Foundation is only used to image nodes and build the cluster; the host runs the hypervisor, not Prism.
Log into Prism Element on each of your three CVM IPs and on the virtual IP, and confirm they are the same console. Then open Settings → Pulse and read what is sent and what obfuscation hides. Lab 1.
Also inToastNix 4.1 4.2 4.3 4.4·Cards 4.1 4.2 4.3 4.4
The course covers: Sessions, charts and reports in Prism Central; alerts and events; alert emails; log collection; creating a support case on the Nutanix Support Portal. Course labs: Adding built-in charts to the Analysis dashboard; creating a custom chart; creating, downloading and scheduling reports; monitoring cluster health; collecting logs.
Seven questions, all “which tool shows me this”. Learn where each view lives: charts in Analysis, history versus present, reports versus alerts, Prism Central versus Prism Element versus the support portal.
Trap: “create an alert policy” sounds like monitoring, but an alert only fires going forward. Any question that says historically or over time wants a chart or a report.
Build one metric chart and one entity chart in Analysis, run NCC, open the Health dashboard, and collect a log bundle. Lab 8.
release, or Direct Upload.Also inToastNix 2.4 3.3·Cards 2.4 3.3
The course covers: License Manager; license details in Prism and on the Support & Insights Portal; licensing a cluster; Life Cycle Manager and its workflow; the LCM dashboard; performing an inventory. Course labs: Exploring the LCM dashboard; performing inventory at a connected site and at a dark site.
Six questions, all LCM. The bank wants the LCM workflow in order (inventory → pre-checks → update), the two dark-site methods, and Redfish’s one selling point. Licensing itself is not tested in this bank, though the blueprint lists it (objective 2.4).
release directory. Q22Q10Trap: “Run LCM pre-checks” and “Settings → Upgrade Software” look like ways to update LCM. Pre-checks validate an update you already chose; 1-click upgrade is the older path. Inventory is what refreshes the framework.
Run an LCM inventory on the CE cluster and watch the task list: the framework update runs before the inventory. Then open LCM settings and find where a dark-site URL would go. Lab 9.
Also inToastNix 2.1 4.4·Cards 2.1 4.4
The course covers: Creating and monitoring VMs in Prism Central; updating, deleting, exporting and cloning VMs; snapshots and recovery points; categories and affinity policies. Course labs: Exploring the VMs dashboard; creating VMs; creating a recovery point; cloning a VM; exporting a VM as an OVA.
Five questions about placement and policy: where a VM runs, how to keep VMs apart or pinned, and how a policy reaches many VMs at once.
Trap: “VM labels”, “annotations” and protection policies appear as distractors for pinning a VM. None of them control placement.
Create two small VMs, put them in an anti-affinity group with aCLI, and watch where they land. Then hot-add a vCPU. Lab 6.
vs0, bridge br0, uplink bond.The course covers: AHV networking terminology; managed and unmanaged networks; monitoring cluster networks in Prism Central; the Network Visualizer in Prism Element. Course labs: Viewing network configuration in Prism Central; using the Network Visualizer.
Four questions: one exhibit read (the visualizer showing one active uplink per host) and three definitions. Know the three bond modes by their balancing method.
Trap: “Active-Passive” is not an AHV bond-mode name. The real names are Active-Backup, Active-Active with MAC pinning, and Active-Active (LACP).
Open the Network Visualizer, then run manage_ovs show_uplinks on a CVM and match the bond mode to what the visualizer shows. Lab 4.
The course covers: The image service; monitoring, uploading, importing and managing images in Prism Central. Course labs: Exploring the images dashboard; uploading and importing an image to Prism Central; uploading an image using Prism Element.
Three short questions, two of them the same fact: categories drive image placement.
Trap: FTP and SMB appear as upload options. The image service takes a file upload or an HTTP(S) URL.
Upload one ISO from your workstation and one from a URL in Prism Element’s image configuration. Lab 5.
Also inToastNix 2.1 1.1·Cards 2.1 1.1
The course covers: Data protection strategies; snapshots and recovery points; the Data Protection dashboard; protection domains and remote sites; Nutanix Disaster Recovery terms. Course labs: Creating protection domains and a local VM restore; configuring remote sites; protection domains using remote sites; VM migration.
One question. A whole course day for one exam point, but the blueprint tests snapshots under VM tasks (2.1), so don’t skip it entirely.
Trap: “Protection Policy” is the scheduled automation; the thing you restore from is the recovery point.
Snapshot a VM, change a file inside it, and restore in place. Lab 7.
The exam is written against the 14 blueprint objectives, not the course. Each row links to the objective’s notes in ToastNix, the modules that teach it, and the practice questions that tested it.
| Objective | Blueprint wording | Modules | Practice Qs |
|---|---|---|---|
| 1.1 | Recognize Nutanix solutions, use cases, and features of Nutanix Cloud Infrastructure (NCI) technologies | M1, M2, M6 | Q2Q16Q27Q34Q42 |
| 1.2 | Recognize Nutanix solutions, use cases, and features of non-Nutanix Cloud Infrastructure (NCI) technologies | M1 | Q21Q50 |
| 1.3 | Recognize the use cases and features of Nutanix tools | M1 | Q35 |
| 2.1 | Describe VM tasks | M4, M5, M6 | Q7Q17Q33Q37Q38Q39Q41Q49 |
| 2.2 | Identify virtual networking tasks | M3 | Q8Q9Q11Q31 |
| 2.3 | Discuss storage operations | M2 | Q1Q12Q19Q20Q23Q28Q40 |
| 2.4 | Describe licensing tasks | M8 | none |
| 3.1 | Describe valid cluster options | M1, M2, M7 | Q5Q6Q14Q24Q29Q30Q46 |
| 3.2 | Identify hardware maintenance tasks | M2 | Q44 |
| 3.3 | Identify software maintenance tasks | M8 | Q10Q22Q25Q26Q36Q43 |
| 4.1 | Perform and interpret a health check | M7 | Q13Q15Q47 |
| 4.2 | Review and interpret alerts and events | M1, M7 | Q18Q45Q48 |
| 4.3 | Leverage support functions | M7 | Q4 |
| 4.4 | Review and interpret workload performance | M5, M7 | Q3Q32 |
Passing the practice bank is not the same as covering the blueprint. These objectives have one practice question or none, and the NHCF course gives them little or no time, but the real exam can still ask about them. Read their ToastNix notes and rate yourself on the blueprint checklist.
Objective 1.2 deserves the most attention of these: the real exam leans on product-to-use-case matching (Files, Objects, Volumes, NDB, NCM, NKP, Enterprise AI, EUC), and this bank only asks about two of them.
These run on the three-node lab cluster at 10.0.0.10. Each lab is built from specific practice questions, adapted from the NHCF course labs. At every step, predict first, then do it and open What you should see. If your cluster disagrees with this page, trust the cluster: note the AOS version and check the linked doc.
The nine labs cover 26 of the 50 questions. The other 24 are listed at the end, each with the reason it can’t be labbed.
https://10.0.0.10:9440, CVMs on 10.0.0.12, .14 and .16, AHV hosts on .11, .13 and .15. No Data Services IP yet; you set one in Lab 2.eth0) per node. The exam is written for AOS 7.5, so a few screens may be named or placed differently; the concepts are the same.nutanix for the aCLI, manage_ovs, ncc and logbay commands.If your nodes are Community Edition or nested VMs, the hardware checks are fake, so expect NCC to complain about things a real node would not (IPMI, disk models). That’s fine: the point is learning where each answer lives.
Questions about reaching Prism and about Pulse are free points once you have clicked through them yourself.
https://10.0.0.12:9440, then .14, .16 and the virtual IP .10.
Predict: Do all four show the same thing?
Yes. The same Prism Element, the same cluster name and the same VMs. Every CVM serves Prism; the virtual IP simply follows whichever CVM is the Prism leader. That is Q2.
https://10.0.0.11:9440.
Predict: Does it answer?
No. Prism runs on the CVMs, not on the hypervisor host. That is the distractor in Q2.
Predict: Is Pulse on, and what does the masking option protect?
CE asks about Pulse at first login. The masking option anonymises entity names and IP addresses: that is the data obfuscation in Q48. Q45’s answer, “disable only with Nutanix Support”, is why it defaults to on.
Six questions ask where storage settings live or what they mean. Opening the dialogs answers all six.
lab-ctr, and open Advanced Settings.
Predict: Which policies can you set here?
Replication factor, compression (with its delay), deduplication and erasure coding: every storage policy is on the container. That is Q1 and Q40. On three nodes the RF choice is RF2; RF3 needs five (Q24).
Predict: Which one is on by default, and does Prism let you enable erasure coding on three nodes?
Inline compression is the recommended default for nearly every workload (Q23). On this cluster, NutanixManagementShare has inline compression (delay 0) while CE created default-container-… with compression off, so check the box on your new container. Erasure coding has a minimum node count; note exactly what Prism says on your three-node cluster before you trust any number in the docs (Q12).
Predict: Does the pool have its own RF or efficiency setting?
No. The pool is just the physical disks and their capacity, which is why “storage pool” is the wrong answer in Q1.
Predict: Is there a Data Services IP? Set one to a free address on 10.0.0.0/24 (for example .17).
Yours starts empty. The Data Services IP is the single iSCSI portal external clients use to reach Volumes (Q19); without it they would have to target individual CVMs.
HA, fault tolerance and data resiliency are easier to remember once you have watched them happen.
Predict: What does the widget say the cluster can tolerate?
One node or disk failure: FT1, because the containers are RF2.
Predict: What happens to the test VM, and how long does it take?
Prism marks the host down, and HA restarts the VM on a surviving host after a few minutes. That is Q14 and Q29: there was spare capacity, so it came back.
Predict: Could the cluster survive a second failure right now?
No. With one node down and RF2, some data has only one copy left until the rebuild finishes. That is the point of Q30: FT1 tolerates one failure domain.
Predict: Does the resiliency status recover on its own?
Yes. Once the CVM is back and data is re-protected, the widget returns to normal. Nothing needs to be done by hand.
Three of the networking questions are recognisable on sight once you have read your own bond and built one IPAM network.
manage_ovs show_uplinks.
Predict: Which bond mode is the uplink in?
Active-backup: the AHV default (Q8). On a nested node there is only one NIC in the bond, so it is the only possible active link.
Predict: How many uplinks per host are drawn as carrying traffic?
One per host. With active-backup and two or more NICs you would see one active link and the rest idle: exactly Q8’s exhibit. Active-active with LACP (Q31) would show both carrying traffic.
10.10.10.0/24, gateway 10.10.10.1, pool .100–.150. Attach a test VM to it and boot it.
Predict: Where does the VM’s address come from?
From the pool, handed out by AHV itself. No DHCP server exists on that network. That is Q9.
The image questions are all about where images come from and where they go.
Predict: Which sources does it accept?
A file uploaded from your workstation, or a URL the cluster downloads from. That is Q39: a remote disk file comes in over HTTP.
Predict: Do both end up in the same list?
Yes. In Prism Central, Image Placement Policies would then decide which clusters hold each image, by category (Q7, Q37).
Placement questions have specific answers: anti-affinity by aCLI, Agent VMs for per-host appliances.
sql1 and sql2, then on a CVM:acli vm_group.create sqlpairacli vm_group.add_vms sqlpair vm_list=sql1,sql2acli vm_group.antiaffinity_set sqlpairPredict: Where do they land?
On different hosts, and HA still protects them. That is Q38.
sql1 while it is running: add a vCPU and 1 GB of memory, then try to remove them again.
Predict: Which changes apply live?
Adding vCPU and memory works on a running AHV VM. Removing them needs the VM powered off.
Predict: What does it promise?
An agent VM stays on its host and is powered off rather than migrated during maintenance: the right tool for one firewall VM per node (Q33). If your build hides the option, read the Managing a VM page linked from M5.
One question, but restoring a VM in place is a skill worth having before you need it.
Predict: After restoring the snapshot, will the file still be there?
No. Restoring reverts the VM in place to the snapshot’s state. Prism Central calls these snapshots recovery points, and that is Q49’s answer.
Four monitoring questions come down to knowing which view shows what.
ncc health_checks run_all. When it finishes, open the Health dashboard.
Predict: Where do the failed checks show up?
In Health, by entity. If Home’s heart has turned red, this is where you find out why (Q47). Nested CE will fail some hardware checks; that is expected here. A fresh CE also raises critical alerts for the default CVM and AHV host passwords; changing them is a worthwhile first fix.
Predict: One chart, several containers. Which chart type was that?
A metric chart: one metric across several entities (Q15). Now make an entity chart for one VM with CPU and memory; that is the other type. Charting memory for several VMs over time is how you answer Q13.
Predict: What stops working if this is empty?
Every email: alerts and NCC reports. Q46’s fix.
logbay collect --duration=-1h.
Predict: Where does the bundle go?
Logbay prints the path of the zip it wrote. That bundle is what you attach to a support case.
The LCM questions are workflow questions. Running it once fixes the order in your head.
Predict: What runs before the inventory itself?
An LCM framework update, if a newer one exists. That is how you update LCM: run an inventory (Q25).
Predict: What are the alternatives to fetching from the internet?
A local web server URL, or Direct Upload of bundles you downloaded elsewhere. That is Q22. On the web server, the framework bundle is extracted into release (Q10).
A mock exam from the full question bank: all 50 questions, 90 minutes, the real exam’s length and clock. Question order and answer order are shuffled every time, so you can’t pass by remembering that it was “B”. You get no feedback until you submit, and a blank counts as wrong. The real exam is scaled (3000 of 6000 to pass), so treat the 70% mark here as a target, not the official cut. Every attempt is saved to your account.
The exam needs the full questions. If they aren’t loaded yet, Start will ask you to sign in first.
This builds itself from the questions you answer in flashcard mode, or mark on the cards below. It ranks the modules you get wrong most often, and gives you the module, its labs and the exact questions to go back to. Once you load the full questions, your answers are saved to your account. Before that, they stay in this browser.
Each entry gives the question in one line, the keyed answer, and why that answer is right, in my own words, with the Nutanix page that settles it. Press Load full questions to swap in the real question text, options and exhibits. That bank is private and needs the site password.
Q1 M2 Lab 2
Where replication factor and storage-efficiency policies are applied in Prism Element.
Answer: A
On the storage container. Every vDisk in a container inherits its RF, compression, deduplication and erasure-coding settings, which is why moving a vDisk between containers is how you change them for one VM. Storage Container Management
Q2 M1 Lab 1
Which IP address reaches Prism Element directly.
Answer: B
Every CVM serves Prism Element, so the IP of any CVM works, as does the cluster virtual IP. The AHV host, the Foundation VM and Prism Central are different things. Logging into the Prism Element Web Console
Q3 M7
Where the cluster runway forecast lives.
Answer: C
Runway is an NCM Intelligent Operations feature on the Prism Central dashboard. Prism Element manages one cluster and has no runway view. Nutanix Cloud Manager (NCM) Intelligent Operations Overview
Q4 M7
What Discoveries on the Nutanix support portal are for.
Answer: A
Insights matches Pulse data against known issues and flags the clusters that are exposed, before you hit the bug. Nutanix Insights - Pulse Health Monitoring
Q5 M1
How to open a registered cluster from Prism Central.
Answer: B
Prism Central’s cluster quick-access widget launches that cluster’s Prism Element in context. Registration is what makes the cluster appear there. Register a Cluster with Prism Central
Q6 M1
Where role-based access control is configured.
Answer: D
RBAC is a Prism Central feature: roles are assigned to users or groups and scoped to clusters or categories. Security and User Management
Q7 M4
The Prism Central feature that maps images to target clusters using categories.
Answer: B
Image Placement Policies decide which clusters hold an image, by matching image categories to cluster categories. Managing Categories in the Categories Tab
Q8 M3 Lab 4
Reading host uplinks in a Network Visualizer exhibit.
Answer: B
One link per host carries traffic and the others are idle standbys: that is Active-Backup, the AHV default bond mode. Bond Types
Q9 M3 Lab 4
The network type that hands VMs IP addresses automatically.
Answer: C
On an AHV network with IPAM enabled, AHV answers the VM’s DHCP request itself from the pool you defined. IP Address Management
Q10 M8 Lab 9
Where to extract the LCM framework bundle on a dark-site web server.
Answer: A
Into the release directory that LCM’s dark-site URL points at. LCM looks for its framework and manifest there. LCM Framework Update using Dark Site Method
Q11 M3
Securing traffic between management and storage replication.
Answer: B
Network segmentation moves CVM backplane (replication) traffic onto its own VLAN or interfaces, apart from management. Network Segmentation
Q12 M2 Lab 2
The efficiency feature that replaces traditional RF copies for suitable data.
Answer: C
Erasure coding stores parity instead of full copies, for write-cold data, which saves capacity while keeping the same failure tolerance. Prism Element - Erasure Coding
Q13 M7 Lab 8
Checking whether memory load has historically been even across an app’s VMs.
Answer: C
Only Analysis charts show history per VM side by side. An alert policy or a filtered VM list shows you the present, not the trend. Adding a Chart
Q14 M2 Lab 3
What the cluster does when a node fails at about 50% utilisation.
Answer: D
AHV HA restarts the HA-protected VMs on the surviving nodes. At 50% load there is room for them, so nothing is left off. Cluster Fault Tolerance
Q15 M7 Lab 8
One chart showing logical usage for three storage containers.
Answer: A
A metric chart plots one metric across several entities. An entity chart is the reverse: several metrics for one entity. Adding a Chart
Q16 M2
The scale-out storage technology behind Nutanix HCI.
Answer: B
The Distributed Storage Fabric pools every node’s local disks into one storage layer served by the CVMs. Introduction to AOS
Q17 M5
What enabling Flash Mode does.
Answer: B
It sets the storage-tier preference to SSD for a VM or vDisk, so its data stays on flash in a hybrid cluster. Managing a VM (AHV)
Q18 M7
Sending weekly VM performance reviews to business managers.
Answer: D
Prism Central reports can be scheduled and emailed. Analysis charts and alerts are not a distribution mechanism. Creating a Custom Report Configuration
Q19 M2 Lab 2
The address external clients use to reach Volumes storage.
Answer: C
The Data Services IP is the cluster-wide iSCSI portal. Clients discover their volume groups through it, and the cluster redirects them to a CVM. Volume Group Configuration
Q20 M2 Lab 2
What Replication Factor 2 refers to.
Answer: D
The number of copies of data written in the cluster. RF2 keeps two, each on a different node. Replication Factor 2
Q21 M1
Storage service for petabytes of unstructured data with S3 APIs.
Answer: D
Objects is S3-compatible object storage. Files is SMB/NFS; Volumes is iSCSI block. Nutanix Objects Overview
Q22 M8 Lab 9
The LCM feature that simplifies upgrades on clusters without internet access.
Answer: C
Direct Upload lets you push downloaded bundles straight into LCM, without standing up a local web server. LCM Framework Update using Dark Site Method
Q23 M2 Lab 2
The storage-efficiency feature suitable for nearly every workload.
Answer: C
Compression. Inline compression is cheap on modern CPUs and helps almost everything; dedup and EC only pay off for specific data. Prism Element - Compression
Q24 M2
Minimum cluster size for redundancy factor 3.
Answer: C
Five nodes. Surviving two failures needs enough members left for the metadata quorum and three-copy placement. Replication Factor 3
Q25 M8 Lab 9
How to upgrade the LCM framework to the latest version.
Answer: C
Run an inventory. Inventory checks for a newer framework and updates LCM itself before it inventories anything else. Performing Inventory with the Life Cycle Manager from Prism Element
Q26 M8
The biggest benefit of Redfish with LCM.
Answer: C
BMC and BIOS updates finish about twice as fast, because the Redfish path needs fewer reboots than the older method. Life Cycle Manager (LCM) Overview
Q27 M2
What the cluster does about repeated Stargate and Cassandra failures on one node.
Answer: B
It redirects that node’s I/O path to another CVM (autopathing), so guest VMs keep their storage while the CVM is unhealthy. Introduction to AOS
Q28 M2
What to check before configuring Replication Factor 1.
Answer: D
The node count, along with RF1’s other prerequisites. RF1 keeps a single copy, so it is only for data you can afford to lose. Replication Factor 1
Q29 M2 Lab 3
A host in a four-node AHV cluster loses all networking.
Answer: C
HA treats it as a failed host and restarts its VMs on the other three, because there is enough spare capacity. Cluster Fault Tolerance
Q30 M2 Lab 3
How many failures an FT1 / RF2 cluster tolerates across domains (disk-diagram exhibit).
Answer: A
One. RF2 keeps two copies, so a second failure in a different node, block or rack can take the only remaining copy of some data. One Node or One Disk Cluster Fault Tolerance
Q31 M3 Lab 4
NIC teaming that balances on source/destination IP and TCP/UDP ports.
Answer: B
Active-Active with LACP (balance-tcp) hashes on L3 and L4 headers. It needs LACP on the switch. MAC pinning balances by VM MAC instead. Bond Types
Q32 M5
The AHV feature that watches for compute and storage I/O hotspots over time.
Answer: B
ADS (Acropolis Dynamic Scheduling) detects CPU and storage-controller contention and live-migrates VMs to relieve it. AHV Overview
Q33 M5 Lab 6
Keeping a per-host firewall VM on its host, most efficiently.
Answer: D
Mark it as an Agent VM. Agent VMs stay on their host, are not migrated, and are powered off for maintenance. Managing a VM (AHV)
Q34 M2
What is stored as a file on CVM-owned storage devices.
Answer: C
The extent group. Extents (logically contiguous data) are grouped into extent groups, and each group is a file on a CVM-owned disk. Storage Components
Q35 M1
What a two-node ROBO cluster needs to stay available through a node failure.
Answer: A
A Witness VM at a third site arbitrates which node keeps serving, so the pair cannot split-brain. Witness VM for Two-Node Clusters
Q36 M8
The resource for a safe AOS and hypervisor upgrade on vSphere.
Answer: C
The Compatibility and Interoperability Matrix lists which AOS, hypervisor and firmware versions are supported together. Determining Compatibility Between Hardware and Supported Products
Q37 M4
What Image Placement Policies use to target clusters.
Answer: D
Categories on the clusters, matched against categories on the images. Managing Categories in the Categories Tab
Q38 M5 Lab 6
Keeping SQL Server cluster VMs on different hosts while keeping HA.
Answer: C
A VM-VM anti-affinity rule, created with aCLI (vm_group). HA still restarts the VMs after a failure. AHV Overview
Q39 M4 Lab 5
How to upload a remote disk file to the image service.
Answer: A
Give the image service a URL over HTTP. The alternative is uploading a file from your workstation. Creating a VM (AHV)
Q40 M2 Lab 2
The entity on which replication factor is configured.
Answer: D
The storage container, the same answer as Q1 asked a different way. Replication Factor Configuration
Q41 M5
Applying one security policy to dozens of VMs efficiently.
Answer: B
Tag the VMs with a category and attach that category to the policy. New VMs join by getting the tag. Managing Categories in the Categories Tab
Q42 M1
The technology that lets applications and data move freely between environments.
Answer: C
The App Mobility Fabric, Nutanix’s name for moving workloads between hypervisors and clouds. Introduction to AOS
Q43 M8
What happens to VMs during an AOS upgrade on vSphere without automatic live migration.
Answer: D
Disputed key. An AOS upgrade restarts the CVMs one at a time, not the hosts. Guest VMs stay on their host while the other CVMs serve their I/O, so B (“continue to run on the same host”) is the defensible reading. The bank keys D. Hosts only need emptying for a hypervisor upgrade. Life Cycle Manager (LCM) Overview
Q44 M2
What happens when an SSD fails in a hybrid AHV node.
Answer: D
The bank keys a reboot of that node’s CVM; its I/O is served by other CVMs meanwhile. Check the Data Drive Failure page for how your AOS release handles a metadata SSD. Data Drive Failure
Q45 M1 Lab 1
When Pulse should be disabled.
Answer: D
Only under the guidance of Nutanix Support. Pulse feeds Insights, proactive support and Discoveries. Pulse Health Monitoring
Q46 M7 Lab 8
NCC alert emails missing after a cluster expansion.
Answer: B
Configure SMTP. Without a working SMTP server there is no email path for alerts or NCC reports. Configuring an SMTP Server
Q47 M7 Lab 8
Which dashboard isolates the cause of a red heart in Prism Element.
Answer: A
The Health dashboard lists every failing check by entity. The heart on Home is only the summary. Verify the Cluster Health
Q48 M1 Lab 1
The Pulse feature that protects a customer’s private information.
Answer: D
Data obfuscation masks entity names and IP addresses before telemetry leaves the cluster. Pulse Health Monitoring
Q49 M6 Lab 7
A manual in-place restore of a mission-critical VM from Prism Central.
Answer: D
Restore from a recovery point. Recovery points are Prism Central’s snapshots, and they can revert the VM in place or clone a new one. Creating a VM Snapshot Manually
Q50 M1
The best fit for deploying a Microsoft SQL cluster on Nutanix.
Answer: A
Volumes: shared iSCSI block storage that clustered Windows and SQL nodes can all attach to. Volume Group Configuration
Every term below is marked in the text above with a dotted underline. Click one anywhere on the page and its definition opens in place, so you never lose your spot. Each entry gives the one-line meaning first and then how it actually works.
These definitions are written from the behaviour the question bank tests and the Nutanix documentation linked on each entry. Where a number or a default matters, check it against your AOS release.
The Nutanix operating system that runs in every CVM and turns the nodes into one cluster.
AOS provides the Distributed Storage Fabric, the cluster services and Prism Element. It is upgraded independently of the hypervisor, as a rolling restart of the CVMs.
See also: Controller VM (CVM) · Distributed Storage Fabric · Prism Element
Nutanix’s own hypervisor, built on KVM and included with AOS.
AHV hosts run the guest VMs and the CVM. Several features (ADS, Agent VMs, Image Placement, Flow) exist only on AHV, which is a frequent distractor on ESXi questions.
See also: Controller VM (CVM) · ADS · Open vSwitch (vs0, br0)
The VM on every node that owns its local disks and serves storage I/O.
Each CVM runs the AOS services: Stargate (data I/O), Cassandra (metadata), Curator (background tasks), Zookeeper (configuration) and Prism. Prism Element is reachable on any CVM’s IP. If a CVM goes down, autopathing sends its node’s I/O to another CVM.
See also: Stargate · Cassandra · Autopathing · Prism Element
The scale-out storage layer that pools every node’s local disks.
The CVMs present one shared pool to every host. Writes are replicated to other nodes according to RF, and data locality keeps a VM’s reads on its own node.
See also: Replication factor (RF) · Data locality · Extent group
The CVM service that handles data I/O.
Stargate receives reads and writes from the hypervisor (NFS, iSCSI or SMB internally) and writes the replicas. Repeated Stargate failures on a node trigger autopathing.
See also: Controller VM (CVM) · Cassandra · Autopathing
The distributed metadata store inside the CVMs.
Cassandra tracks where every extent lives. It needs a quorum of CVMs, which is part of why higher fault-tolerance levels need more nodes.
See also: Controller VM (CVM) · Stargate · Fault tolerance / redundancy factor
Redirecting a node’s storage I/O to another CVM when its own CVM is unhealthy.
The host’s storage path is pointed at a healthy remote CVM, so guest VMs keep running without a restart. When the local CVM recovers, I/O moves back.
See also: Controller VM (CVM) · Stargate
The web console for one cluster, served by its CVMs.
Reach it on https://<any CVM or the virtual IP>:9440. It covers the cluster’s own VMs, storage, network, health, NCC and LCM.
See also: Prism Central · Controller VM (CVM)
A separate VM that manages many clusters from one console.
Clusters are registered to it. It owns the cross-cluster features: RBAC, categories and the policies built on them, reports, runway, Intelligent Operations and Image Placement Policies.
See also: Prism Element · Categories · Runway · Role-based access control
Granting users roles scoped to specific clusters or categories.
Configured in Prism Central. A role bundles permissions; an authorisation policy binds a role to users and to a scope.
See also: Prism Central · Categories
Telemetry the cluster sends to Nutanix for proactive support.
Pulse feeds Insights, Discoveries and support cases. It is on by default; disable it only under Nutanix Support’s guidance. Data obfuscation masks names and IPs first.
See also: Insights / Discoveries · Data obfuscation
Masking entity names and IP addresses in Pulse data.
Protects customer-identifying information while still letting Nutanix spot issues by pattern.
See also: Pulse
Nutanix’s analysis of Pulse data on the support portal.
Discoveries flag clusters exposed to known issues before they hit them, and support can open cases proactively.
See also: Pulse
How many copies of each piece of data the cluster keeps.
RF2 keeps two copies on different nodes, RF3 three. RF is a container setting. RF1 (one copy) exists for disposable data and has prerequisites.
See also: Fault tolerance / redundancy factor · Storage container · Erasure coding (EC-X)
How many simultaneous failures the cluster survives.
FT1 survives one node or disk failure and pairs with RF2. FT2 survives two, pairs with RF3, and needs at least five nodes. The Data Resiliency widget shows the current level.
See also: Replication factor (RF) · Cassandra
The group of physical disks the cluster stores data on.
Normally one per cluster. It carries capacity, not policy: RF and efficiency settings belong to containers.
See also: Storage container
A logical slice of the storage pool that carries the storage policies.
RF, compression, deduplication and erasure coding are all set here, and every vDisk inside inherits them. Live vDisk migration moves a disk between containers.
See also: Storage pool · Replication factor (RF) · Compression · Erasure coding (EC-X) · Deduplication
The unit stored as a file on a CVM-owned disk.
Data is written as extents (logically contiguous chunks of a vDisk), which are packed into extent groups; each extent group is a file on a physical disk the CVM owns.
See also: Distributed Storage Fabric · Storage container
Shrinking data as it is written (inline) or later (post-process).
Cheap on modern CPUs and helpful for nearly every workload, which makes it the safe default efficiency setting.
See also: Deduplication · Erasure coding (EC-X) · Storage container
Storing identical blocks once.
Pays off where many VMs share the same data, such as full clones and VDI. Elsewhere it costs metadata and memory for little gain.
See also: Compression · Erasure coding (EC-X)
Replacing full replica copies with parity for cold data.
Data that is written once and rarely rewritten is re-protected as data plus parity strips across nodes, saving capacity at the same tolerance. It needs a minimum node count.
See also: Replication factor (RF) · Compression
Pinning a VM’s or vDisk’s data to the SSD tier.
On hybrid nodes, it keeps latency-sensitive data on flash instead of letting it age down to HDD.
See also: Storage container
Keeping a VM’s data on the node it runs on.
Reads are served locally; the replica goes to another node. When a VM moves, its data follows over time.
See also: Distributed Storage Fabric
iSCSI block storage presented to VMs or physical hosts.
Shared disks for clustered applications such as Microsoft SQL and Windows failover clusters. External clients connect through the Data Services IP.
See also: Data Services IP · Objects · Files
The cluster-wide iSCSI portal address.
iSCSI initiators discover targets through it and are redirected to a CVM, so the address stays the same whichever CVM is serving.
See also: Volumes (volume groups)
S3-compatible object storage.
For petabyte-scale unstructured data: backups, archives, data lakes. Files is SMB/NFS; Volumes is block.
See also: Volumes (volume groups) · Files
SMB and NFS file shares served by file-server VMs.
Scale-out NAS on the cluster, for home directories and departmental shares.
See also: Objects · Volumes (volume groups)
The virtual switching layer inside every AHV host.
The default virtual switch vs0 wraps bridge br0 and its uplink bond. Configure it cluster-wide from Prism so every host matches.
See also: Bond modes · IPAM
How an AHV host spreads traffic over its uplinks.
Active-Backup (default): one link carries traffic, the others stand by. Active-Active with MAC pinning (balance-slb): VMs spread across links by MAC, no switch config. Active-Active with LACP (balance-tcp): hashes on IP and TCP/UDP ports and needs LACP on the switch.
See also: Open vSwitch (vs0, br0) · Network Visualizer
AHV assigning VM addresses itself on a managed network.
When enabled on a network, AHV intercepts the VM’s DHCP request and answers from the pool you defined. No external DHCP server is needed.
See also: Open vSwitch (vs0, br0)
Separating cluster traffic types onto their own VLANs or interfaces.
Moves CVM backplane (replication) traffic away from management, and can isolate services such as Volumes or DR.
See also: Open vSwitch (vs0, br0)
Prism Element’s map of hosts, uplinks, VMs and switch ports.
Uses LLDP/CDP and SNMP to draw the physical connections. In Active-Backup you can see one active uplink per host.
See also: Bond modes
Key:value tags that policies target instead of individual VMs.
Security, protection, Image Placement and affinity policies all point at categories. Tag a new VM and it joins every policy bound to that category.
See also: Prism Central · Image Placement Policy
Rules for which clusters hold which images.
A Prism Central policy that matches image categories to cluster categories.
See also: Categories · Prism Central
A VM tied to its host, such as a per-node firewall or antivirus appliance.
Agent VMs are not migrated: they stay on their host and are powered off during host maintenance.
See also: Affinity and anti-affinity
Rules that keep VMs on certain hosts, or apart from each other.
VM-host affinity is set in Prism. VM-VM anti-affinity, for example for clustered SQL nodes, is created with aCLI vm_group commands.
See also: Agent VM · VM high availability
Acropolis Dynamic Scheduling: automatic hotspot relief on AHV.
Watches CPU and storage-controller contention over time and live-migrates VMs to balance the hosts.
See also: VM high availability · Affinity and anti-affinity
Restarting a failed host’s VMs on the surviving hosts.
On AHV, HA-protected VMs restart elsewhere when a host fails or loses all connectivity, as long as there is spare capacity.
See also: Fault tolerance / redundancy factor · ADS
Prism Central’s name for a VM snapshot.
Restore one in place to revert the VM, or clone a new VM from it. Protection policies create them on a schedule.
See also: Protection domain
Prism Element’s grouping of VMs for scheduled snapshots and replication.
Protection domains replicate to remote sites. In Prism Central the equivalent model is protection policies and recovery plans.
See also: Recovery point
A tie-breaker at a third site for two-node clusters and Metro Availability.
Without it a two-node pair can’t tell a dead partner from a broken link, so it can’t keep serving safely after a node failure.
See also: VM high availability · Fault tolerance / redundancy factor
The framework that upgrades firmware and software across the cluster.
Workflow: inventory, pick updates, pre-checks, then a rolling update. Inventory updates the LCM framework itself first.
See also: Dark site · Redfish · Compatibility and Interoperability Matrix
A cluster with no internet access to Nutanix.
LCM then uses a local web server (framework bundle extracted into release) or Direct Upload of bundles you downloaded elsewhere.
See also: Life Cycle Manager (LCM)
A standard BMC management API that LCM uses for firmware updates.
On Redfish-capable hardware, BMC and BIOS updates finish about twice as fast because they need fewer reboots.
See also: Life Cycle Manager (LCM)
The support-portal list of which versions work together.
Check it before any AOS, hypervisor, Prism Central or firmware upgrade.
See also: Life Cycle Manager (LCM)
Nutanix Cluster Check, the health-check framework.
Runs hundreds of checks from Prism or with ncc health_checks run_all. Failures appear on the Health dashboard; results can be emailed if SMTP is set.
See also: Logbay
The log-collection tool on the CVMs.
logbay collect gathers logs for a time window into one bundle to attach to a support case.
See also: NCC
The two Analysis chart types.
A metric chart plots one metric across several entities. An entity chart plots several metrics for one entity.
See also: NCC
Prism Central’s forecast of how long capacity lasts.
An Intelligent Operations feature on the Prism Central dashboard. It projects CPU, memory and storage growth to the day they run out.
See also: Prism Central