A Modern Architecture for Enterprise Resilience

Numerous high-profile incidents involving a lack of IT resilience have occurred in recent years, leading to significant financial losses and severe reputational damage.

Because infrastructure solutions are not resilient by default, IT organizations continue to spend thousands of hours and millions of dollars trying to achieve resiliency. Admin teams work to configure infrastructure and servers for High Availability (HA) and Disaster Recovery (DR), but it can be difficult to configure traditional solutions for full resiliency and hard to restore full operations after a problem occurs.

A Resilient, Clustered Architecture

The fundamental architectural unit in the Nutanix environment is the cluster. All functions within a Nutanix cluster can be distributed to any node in the cluster. This provides resilience, mobility, and performance—and simplifies maintenance and upgrades. 

• Cluster state: All cluster state information is stored in a distributed database. If a single node fails, important state information remains accessible on the other nodes.  
• Cluster management: The cluster control plane, Nutanix Prism, is also distributed across every node in the cluster, a design that allows  management functions to remain accessible if some nodes become unavailable. 
• Stateless hypervisor: Unlike many other hypervisors, the Nutanix AHV hypervisor is stateless. When AHV boots, it gets its state from the cluster database and takes action accordingly. The AHV state is designed to never be out of sync with the control plane.
• Storage: Storage within a Nutanix cluster is also distributed, and each node in the cluster has access to all storage. All functions related to storage and data protection can be managed from within Prism. All data is replicated within the cluster for resiliency based on a set replication factor. This proven architecture for clustering was born from HCI, with distributed storage. Today, even Nutanix Cloud Infrastructure with external storage benefits from this clustered approach, with a resilient management and control plane.

To begin to understand the implications, it’s worth comparing Nutanix operations with analogous operations using VMware solutions. A Nutanix cluster running AHV is similar to multiple VMware hosts running ESXi managed by a vCenter management instance.

• System state: ESXi hosts are not stateless, they store their own internal state, and vCenter also stores the state of all managed hosts. There’s no consensus protocol, so host failures and network partitions can lead to inconsistent state that requires manual intervention and recovery (as documented in Broadcom knowledge base articles 317923 and 344682.
• Cluster management: vCenter is external to ESXi, so it must be deployed and managed separately. If vCenter is down, there’s no centralized management, and operations are impacted. 
• Storage: Storage can be provided either with a traditional three-tier architecture using external storage or via a hyperconverged storage architecture using VMware vSAN.

Prism Central

For larger and multi-cluster Nutanix deployments, Nutanix provides a central management plane called Prism Central (PC). PC uses the same distributed database as individual Nutanix clusters. The clusters share state with PC using a consensus protocol to ensure that state always remains consistent. 

Prism Central can be clustered  for high availability. Clustered PC is active-active, making it easy to architect and deploy, and easy to recover. Management remains accessible if one of the hosts is down.

Nutanix also provides complete backup and restore of Prism Central instances in case of a catastrophe. PC instances can be backed up to another Nutanix cluster, or to S3-compatible object storage.

vCenter can also be clustered, but vCenter clusters are active-passive-witness. This could make them difficult to architect and deploy. When vCenter is down, centralized management is down. (See vCenter High Availability for reference.)

When WesBanco needed visibility and control over its IT operations, the company turned to Nutanix. Watch the video:

Benefits of the Nutanix Approach

The Nutanix approach offers many benefits. 

• Resiliency out of the box: When you deploy a Nutanix cluster running AHV, it automatically starts out in a resilient state, with management capabilities distributed across the cluster. AHV provides load balancing and high availability (HA). Should a VM fail, it is automatically restarted. Should a node fail, its VMs are automatically restarted on another node. 
• Less management effort: From a software standpoint, it’s simple to stand up a resilient Nutanix environment because resiliency is designed in.
• Control plane: Built into the cluster. In the VMware environment, vCenter must be deployed, managed, and updated separately as described earlier and as documented in the vCenter Installation and Deployment Guide.
• CAPEX AND OPEX: Simple management and built-in resilience drive efficiencies that can translate into CAPEX and OPEX savings. A recent IDC study found that participating customers had up to 41% lower infrastructure costs and 42% lower 3-year cost of operations; each IT team member was able to manage 90% more VMs.

Resiliency at Every Level

The clustered, distributed architecture allows Nutanix to provide resiliency at every level.

Standalone Cluster: A standalone Nutanix cluster running AHV offers great built-in resilience features:

• HA: Best effort HA is built-in and enabled by default. Resources can be reserved to guarantee HA for critical VMs. In the VMware environment, vSphere HA and/or Distributed Resource Scheduler (DRS) have to be explicitly enabled and configured. (See the vSphere 9.0 documentation for details.)
• Data protection: Each AHV VM can be protected according to a designated snapshot schedule. Immutable snapshots are supported for resiliency and ransomware protection.
• Encryption: Nutanix offers flexible data-at-rest encryption options to help prevent data loss or theft. 
• Nutanix Files Storage and Nutanix Objects Storage: Nutanix clusters support on-cluster file and object storage, eliminating the need for separate, external storage. Both the Nutanix Files Storage and Objects Storage solutions support write-once-read-many (WORM) for immutability, resiliency, and ransomware protection.

Standalone Cluster plus Prism Central: Adding Prism Central, even with a single cluster, offers additional resiliency capabilities in addition to those described above:

• Multicloud Snapshot Technology (MST): Replicate VM snapshots to a cloud object store or other cloud storage
• Disaster Recovery: Take advantage of online DR to the cloud or to a Nutanix Service Provider.

Multiple Clusters plus Prism Central: Having multiple clusters (including Nutanix Cloud Clusters running in the public cloud) adds the following resiliency capabilities:

• Replication: Take advantage of native async, NearSync, or sync replication to replicate data between clusters for DR or other needs.
• MST to Nutanix Objects Storage: Replicate snapshots to a cluster running Nutanix Objects Storage or any other S3-compatible object store.
• Centralized policy across clusters: Establish the same storage, security, and cost governance policies across all clusters for consistency and resilience.

Multiple Prism Central Instances: Having multiple PC domains adds additional levels of resilience:

• DR between Availability Zones: Create availability zones and enable DR between zones.
• Policy replication: Distribute protection and security policies across availability zones for consistency and protection.

Resilience that Makes a Difference

Nutanix offers a simple and flexible architecture, providing a resilient foundation and offering resiliency at every level to satisfy demanding enterprise availability and cyber resiliency requirements. Blog 2 in this Engineered Differently series, Built-in Security Across Platform, Data, Network and Applications, describes how Nutanix delivers comprehensive security at the platform, data, network, and application levels.

The final blog in this series will explain how Nutanix enables streamlined AI deployment.

To find out more about Nutanix management, visit www.nutanix.com/nutanixcentral or www.nutanix.com/prism. To learn more about Nutanix resiliency features, view our Infrastructure Resiliency Tech Note.

View the recent topics covered in this blog series: 

• Engineered Differently: Nutanix Delivers a Uniform Experience Across Datacenter, Edge and Cloud
• Engineered Differently: Built-in Security From Platform to Data to Network to Application
• Engineered Differently: Zero-Touch Provisioning, Configuration Management and Software Updates at Massive Scale

_{©2025 Nutanix, Inc. All rights reserved. Nutanix, the Nutanix logo and all Nutanix product and service names mentioned are registered trademarks or trademarks of Nutanix, Inc. in the United States and other countries. Nutanix, Inc. is not affiliated with VMware by Broadcom or Broadcom. VMware and the VMware product names recited herein are registered or unregistered trademarks of Broadcom in the United States and/or other countries.}