Where is Redundancy Factor set in Nutanix?

Replication/Redundancy Factor and the optimization features are set at the storage container level, not the storage pool. Containers present as datastores or shares to the hypervisor.

What are the three storage-optimization features in Nutanix?

Compression, deduplication, and erasure coding (EC-X). All three are configured per storage container to reduce the physical capacity that data consumes.

What is NC2 in Nutanix?

Nutanix Cloud Clusters (NC2) runs the Nutanix stack on public-cloud bare-metal instances, such as AWS and Azure, extending the same AOS and AHV experience into the cloud.

← NCA 6.10 Study Hub

NCA 6.10 Study Guide · Section 4 of 4

Describe Cluster Configuration Options

The biggest section: storage components, Replication and Redundancy Factor, the three storage-optimization features, supported hypervisors and NC2, networking, and use cases. Maps to Module 1, Module 2, Module 5 (DSF), and Module 6 (Networking).

🎯 Section 4 of 4 Maps to Modules 1, 2, 5, 6 Versions AOS 6.10 · pc2024.2

4.1Storage components

Component	What it is
Storage Pool	All physical disks (SSD/NVMe/HDD) across all nodes, pooled into one tier of capacity. Usually one pool per cluster.
Storage Container	A logical slice of the storage pool where you set policies (RF, compression, dedup, erasure coding). Containers present as datastores/shares to the hypervisor. Optimization and RF are set at the container level.
vDisk	A virtual disk a VM consumes, living inside a container. Made of extents; data is distributed across the cluster.

The storage layer is the Distributed Storage Fabric (DSF), served by the Controller VM (CVM) on every node. Data locality keeps a VM's hot data on the node it runs on.

4.2Replication Factor and Redundancy Factor

Setting	Copies / meaning	Tolerates	Minimum nodes
RF2 / FT1	2 copies of data (metadata RF3)	1 node or drive failure	3 nodes
RF3 / FT2	3 copies of data (metadata RF5)	2 simultaneous node or drive failures	5 nodes

Replication Factor (RF) is the data setting on a container (how many copies of data). Redundancy Factor / Fault Tolerance (FT) is the cluster-level resiliency. They line up: FT1 ↔ data RF2, FT2 ↔ data RF3. Default for a cluster is RF2 / FT1.

Exam shape

Memorize the minimums: RF2 needs 3 nodes, RF3 needs 5 nodes (RF3 keeps 5 metadata copies, hence 5). "Tolerate two simultaneous node failures" → RF3 / FT2.

4.3Storage optimization: compression, deduplication, erasure coding

Feature	How it works	Best for
Compression	Inline (large/sequential I/O, >64 KB) or post-process; uses LZ4 / LZ4HC. The recommended default optimization.	Almost everything; broadly safe to enable
Deduplication	SHA-1 fingerprint on ingest; removes duplicate blocks (inline and post-process)	Full clones, VDI, persistent desktops, repeated OS images
Erasure Coding (EC-X)	RAID-like parity striped across nodes; post-process on write-cold data (not overwritten for >7 days)	Cold / write-once data; reclaims capacity without RF's full copy cost

Exam shape

Match the workload: dedup → VDI / full clones; erasure coding → cold, infrequently changed data; compression → general default. Know EC-X is post-process on data cold for ~7 days. All three are set on the storage container.

4.4Supported hypervisors and NC2

Nutanix runs your choice of hypervisor: AHV (built in, free, KVM-based), VMware ESXi, and Microsoft Hyper-V. AHV is the default.

NC2 (Nutanix Cloud Clusters) runs the full Nutanix stack (CVM + AHV + AOS + Prism) on bare-metal instances in AWS and Azure, operated as one cloud with on-prem. It is the "extend/migrate to public cloud without changing your operating model" answer.

4.5Cluster sizes

Size	Notes
3-node (standard)	The normal minimum for a resilient cluster (RF2 / FT1)
2-node (ROBO)	Edge/remote; requires an external Witness VM in a separate failure domain
1-node (ROBO)	Lowest cost, reduced resiliency; only on selected hardware models, limited guest-VM count

4.6Networking basics (AHV)

AHV networking uses Open vSwitch (OVS). The default bridge is br0 with a default uplink bond br0-up; OVS supports VLAN tagging, load balancing, and LACP. Best practice puts the CVM and AHV host on the native/untagged VLAN, and a bond should have at least two physical NICs for high availability. Guest VM VLANs are set on the VM NIC (virtual network / VLAN).

4.7Building it: Foundation

Foundation is the tool that images bare-metal nodes (installs the chosen AOS and hypervisor) and forms them into a cluster. It is built into the CVMs and needs roughly three IPs per node (hypervisor, CVM, and IPMI/remote management) plus a cluster IP and a Data Services IP.

§Section 4 in one breath

Disks form a storage pool; policy (RF, compression, dedup, EC) is set on a container; VMs use vDisks. RF2/FT1 (3 nodes) survives one failure, RF3/FT2 (5 nodes) survives two. Optimize with compression (default), dedup (VDI/clones), or EC-X (cold data). Hypervisor is AHV/ESXi/Hyper-V, or run on cloud bare metal with NC2. AHV networking is OVS / br0. Foundation images nodes into a cluster.

§Self check

1. Where do you set replication factor and compression, the pool or the container?

2. Minimum nodes for RF2? For RF3? How many failures does each tolerate?

3. Which optimization fits VDI full clones? Which fits cold archival data?

4. What does a two-node ROBO cluster require that a three-node cluster does not?

5. What is the default OVS bridge name on AHV, and where should the CVM/host VLAN sit?

6. What does NC2 run on, and where?