Nutanix 4.6 Features Overview – Part 1 (Beyond Marketing)

Update: Part two of the series can be found here.


Only four months after releasing Nutanix Based Software 4.5 (read Prism Central and NOS 4.5, NCC 2.1, Foundation 3.0 and MORE Tech Preview Features – What’s New!), a major release with multiple new features and improvements, Nutanix is once again pushing out a major release with tons of new features and huge performance improvements (here Official Press Release).

The velocity with which new features and improvements are released by Nutanix can only be compared to what we see in the consumer space where apps are updated at an amazing pace for user’s delight. I personally have never seen such sophisticated SDLC in the enterprise space, and the Nutanix Engineering, Product Management and Release teams continuously amazes me with their speed and quality.

The benefit of Nutanix being a software-defined-platform running out-of-kernel is that innovation speed is drastically faster when compared to in-kernel solutions that rely on long and entrenched train releases with extensive compatibility requirements and functional tests. It is not difficult to break an entire solution because a single component was created with bad code.

Furthermore, it is proven that hypervisor direct-path provides efficient hardware access while providing similar or even better performance than using kernel datapaths.

Without further ado let’s dig into what has been added into this new major release that is available for download and upgrade via 1-click upgrade.


Distributed Storage Fabric (DSF)

Native File Services (Tech Preview)

Nutanix has been primarily used to host virtual machine storage in the format of vdisks, but now it is also possible to host file storage (file services). This release is introducing a new Tech Preview technology that make File Services an integral and native component of the Nutanix Distributed Storage Fabric, removing the need for Windows File servers VMs or external NAS arrays, such as Netapp or EMC Centera or Isilon.

File Service is a scale-out native solution that provides single namespace entry-point per group or clustered file servers. Providing single namespace Nutanix avoids the problem of fractured namespaces deployments in traditional infrastructure where every single storage volume is a unique namespace. This namespace is capable of scaling to billions of files, while at the same time enabling simple management via a single volume.



The scale-out structure of the File Services enable administrators to easily scale capacity on-demand. If you need more capacity, then the file server cluster will just use more capacity from the cluster (assuming the cluster has additional physical resources). However, if additional performance is necessary, such as IOPs or throughput because of the number of users or applications accessing the file system, then the file server automatically spins up more instances to handle that performance. Note: In NOS 4.6 Tech Preview the auto-scale capability is disabled.

Couple with erasure coding and storage only nodes the solution will deliver cheap and deep storage at cloud storage prices – well below a $1/GB. On top of that you may compute data reduction technologies such as compression and deduplication, making the pricing point even more attractive.



File Servers are a reality for all types of business and storage related use cases, but this solution is specially interesting for VDI uses cases given that user profiles and home directories can now stay co-located with the users VMs on the Nutanix cluster, reducing complexity and improving TCO of the entire solution. In the Nutanix case deployments will not require storage heavy nodes since administrator can use spare-cluster capacity to host user profiles and data. The most common features for users like SMB 2.1, quotas and self-service restores will be supported at GA.

Watch the below video where I explain how to instantiate a new file services cluster.

[Watch in full-screen 1080p] [Recording done with pre-GA code; things may have changed a little bit]


Performance Improvements

Performance improvements have been across the board for both random and sequential workloads and I am certain few of my fellow Nutanix colleagues will be publishing more detailed information (here). I want to leave you with this graph that demonstrated the performance improvements attained just with Nutanix software optimization while maintained the exact same hardware.

  • Better performance across a wide range of enterprise application workloads, e.g., powering up to 30,000 Microsoft Exchange mailboxes in just 8U of rack space, as validated through the Microsoft Exchange Solution Reviewed Program (ESRP)
  • Delivering the industry’s fastest storage I/O performance across hyperconverged all flash solutions with over 1 million storage IOPS in just 4U of datacenter space.
  • Up to 4x price/performance gains on existing investments to deliver as low as $0.35/IOPS, better than top-selling all-flash arrays which also suffer from complex management and additional networking requirements.

To get this massive performance improvement in your existing Nutanix cluster just hit 1-click upgrade.


Volume Groups in PRISM

Volume groups have been available since NOS 4.5 via CLI, but now the feature is also available and supported via the PRISM user interface.

Volume Groups provide block storage and are particularly important for enterprise applications that do not provide support for NFS datastores, or applications that require block storage “shared” across deployment instances. The use cases include: Microsoft Exchange on ESXi, Windows 2008 Guest Clustering, Microsoft SQL 2008 Clustering and Oracle RAC.

Volume Groups provide support for Snapshots, Asynchronous Replication and scripted DR orchestration.




Self Service Restore (SSR)

File restores are a reality for system and datacenter administrators. Current restore options always involve administrators to execute tasks.

One of the existing options is to enable administrators to restore the full VM, but while this is easy, it’s not the most optimal option since users typically just require a few files from snapshots or backups.

The second existing option is to restore individual files from clone VMs. This option allows for fine-grained restores but is cumbersome – as it involves spinning up a VM clone and notifying the user, who then copies files over from the clone to the production VM. The administrator has to the remember to clean up the clone afterwards.

The better way involves enabling the administrator to give control of a VM’s snapshots to the end-user. This way an authorized end-user can see only his/her snapshots within the guest, then mount the appropriate snapshot to the guest (which shows up like a drive letter) and then copy over lost files.

This feature requires NGT (Nutanix Guest Tools) installed in the guest and SSR enabled on a per-VM basis. The entire workflow is done in-guest without administrative intervention, hiding much of the complexity such as disk signature clashes and system reserved partitions. File ACLs are preserved and restored disks show up as a new drive letter within the VM.

Please note that SSR is not intended to replace full-fledged backup solutions since it leverages local snapshots, but it can be a good enough solution for ROBO and SMB deployment. SSR empowers the App/VM owner, rather than overburden the Nutanix administrator.

This initial release doesn’t yet provide a GUI, only CLI, but a browser-based GUI within the Guest VM capability is slated for an upcoming release.


1-Click Upgrades of Foundation, BIOS, BMC

Nutanix 1-click upgrade automatically indicates when a new component version is available and will auto-download binaries if the auto-download option is enabled. The upgrade process uses a highly parallel process restarting one component at the time using a rolling upgrade mechanism. The entire cluster upgrade can be fully monitored by the administrator.

In this new release 1-click upgrade include updates for Foundation, BIOS and BMC components, in addition to previously supported 1-click upgrade for NOS, disk firmware, hypervisors and NCC.

I recommend reading this article (New PRISM Central One-Click Upgrade and One-Click Smorgasbord!) and watching the videos to better understand how 1-Click upgrade works.



Metro Availability Non-Disruptive vMotion and Site Migrations

Prior to NOS 4.6 a VM power cycle was required once VMs were migrated from SiteA to SiteB. In this release VM migrations from one site to the other are entirely non-disruptive. The use case here is site maintenance.

Important: While VMs will non-disruptively be moved between sites, the storage datapath will remain with the original site until the VM is restarted or the Metro Availability is broken or promoted. This means that storage traffic will flow over the network between sites; however, nodes on the original site can still be placed in hypervisor maintenance mode.


Application Mobility Fabric (AMF)

Cross Hypervisor Disaster Recovery

Cross Hypervisor DR addresses use cases where additional hypervisor deployments are required for DR sites, ROBO replication into core datacenters, Service Providers offering DRaaS (Disaster Recovery as a Service) using AHV in their datacenters, and also to avoid vTax for secondary datacenters.

The new feature makes it simple and seamless to replicate and migrate VMs between vSphere and AHV without the requirement of external tools.

As with other Nutanix replication use cases this one is bi-directional with native two-way VM configuration conversion when migrating or failing over; native 1-click migration between sites. Scripted Workflow customization is also supported, and the best thing is that the feature is available to all license editions.


Project Dial: In-Place Hypervisor Conversion (Tech Preview)

This is a new feature in Tech Preview. 1-click conversion of cluster hypervisor, CVM and user VMs from ESXi to AHV without data migration required. The software executes all pre-conversion cluster level checks to ensure a successful conversion. When this feature goes GA it will be an amazing tool to convert ROBO and SMB deployments from ESXi to AHV. Future versions may enable any direction conversion as part of the App Mobility Fabric.


OpenStack Integration

Nutanix 4.6 now has full OpenStack integration and support for AHV hypervisor, offering drivers for Nova, Cinder, Glance and Neutron. The solution comes packaged as a disk image and is deployed as a standalone System VM (SVM) that directly talks to the Acropolis APIs and also enable any OpenStack distribution to work with AHV without any changes to OpenStack – a Nutanix cluster appears as a “hypervisor” to OpenStack.

In a single OpenStack deployment can support multiple AHV clusters and the SVM is protected with HA for availability. As part of the solution package there is support for multi-tenancy and authentication via KeyStone service. Additionally, placement decisions across the Nutanix clusters are taken by the Nova scheduler and placement decisions within the Nutanix Cluster are taken by the Acropolis scheduler.


Citrix Power Management Plugin

The current version of the Citrix Power Management plug-in integrates with XenDesktop 7.6 and enable power management operations for the desktops being managed by XenDesktop on AHV hypervisor, but it does not use MCS for desktop life-cycle management. The new release that is in Tech Preview offers complete desktop life-cycle management from within XenDesktop Studio and offer all the benefits of Citrix MCS technology, without any of the drawbacks.

I recommend reading this article by Kees Baggerman for more information and tech preview demo.

[Watch in full-screen 1080p]  [Recording done with pre-GA code; things may have changed a little bit]


AHV Guest Customization

More a necessity than an improvement, but good to know administrators can automate deployments and guest customization. AHV VM guest customization allow administrators to set unique identifiers on first boot such as AD object ID, network identity, hostname, custom ssh key injection, security customizations, application deployment and virtIO driver injection in Windows at unattended fresh installs. Any user provided script can be injected into the VM and custom specifications such as Sysprep for Windows and Cloudinit YAML for Linux are also supported.



AHV VLAN Trunking

Another must have item that can now be scratched from the list. VLAN trunking provides higher performance, by avoiding multiple NIC overhead for a VM. AHV also provide support for virtual networking appliances that commonly have multiple network interfaces.


Prism – the product

There are so many new features in PRISM that I will probably forget to include some of them in my list. Here my list with the most important updates in my opinion.


Actionable Capacity Forecasting

PRISM now offers cross-cluster forecasting for CPU, memory, and storage capacity runway (time remaining) based on machine-learned consumption behavior from the running workloads with Nutanix X-FitTM, a patent-pending machine learning technology. It also demonstrates detailed trending information of hosts (in CPU and memory) and containers (in storage) level with recommendation of appropriate scale-out options (node types and numbers).


Customizable Dashboards

PRISM now offers the ability to create and customize your own dashboard by dragging and dropping widgets. Widgets can be pre-defined or customizable for charting, alerts, cluster status, capacity runway, and top resource utilization top.

[Watch the video below]

Intelligent keywords search

PRISM now offers keywords based search with auto-complete and instant results. These searches provide contextual results, showing configuration, charts, alerts, appropriate actions, and help contents.

[Watch the video below]

Entity Explorer

PRISM now offers a single pane of glass for all cluster entities and VMs where administrators can easily execution of actions within the explorer. The entity explorer allows for powerful filters, including configuration, status, and usage, to narrow down the selection; all with different view formats (e.g. grid and tile) with grouping and focus perspective.

[Watch the video below]


[Watch in full-screen 1080p]  [Recording done with pre-GA code; things may have changed a little bit]


In the second part of the series I will cover:

  • SaltStack self-healing baseline/ STIG AHV
  • Single CG Restriction Removal for Linked Clone Deployment
  • Foundation 3.1 Updates
  • Network Integration – Phase 2
  • Nutanix Guest Tools (NGT)
  • Replication Factor Migration/Conversion
  • Application Consistent Snapshots for AHV
  • … and few others small features


Update: Part two of the series can be found here.


This article was first published by Andre Leibovici (@andreleibovici) at

1 comment

6 pings

  1. Hi Andre, nice article (as usual). One comment re the new Prism features: probably worth pointing out that these are Prism Pro features that (I believe?) will be a license-able option moving forward.

  1. […] Source: […]

  2. […] « Nutanix 4.6 Features Overview – Part 1 (Beyond Marketing) […]

  3. […] Read the full Nutanix AOS 4.6 launch overview here. […]

  4. […] Nutanix 4.6 Features Overview – Part 1 (Beyond Marketing) […]

  5. […] Nutanix 4.6 Features Overview (Beyond Marketing) […]

  6. […] Nutanix 4.6 Features Overview (Beyond Marketing) […]

Comments have been disabled.