Nutanix – Taking It to the .NEXT Level

I was happy to participate in the opening “Nutanix Champion” event to “ring in” the day one keynote.  When I got back stage during the rehearsal, it was evident a lot of people worked real hard to do something fun for the opening acts (Angelo Luciani, Julie O’Brien  and surely many more…so, kudos to you!).



Picture credit to

And now, my take on some of the announcements today….

Acropolis File Services:

This announcement fits into the recurring theme of “power through software” – leveraging commodity hardware to deliver additional value based on software upgrades and enhancements.  Acropolis File Services allows you to leverage your existing investment to expose the Nutanix file system for scale out file level storage.

Initially SMB 2.1 will be supported with other protocols (NFS, never versions of SMB, etc) on the roadmap.  AHV and ESXi hypervisors will be supported at GA.  Other features include user driven restore leveraging file level and server level snapshots (think file level recovery and disaster recovery, respectively) with asynchronous replication on the roadmap for Q4.

There are some interesting use cases I can see for this, such as user profile storage and replication for desktop and application virtualization environments and low cost scale out file services using the included Acropolis Hypervisor + Nutanix storage nodes.

Acropolis Block Services:

Similar to Acropolis File Services, Acropolis Block Services exposes the Nutanix file system as an iSCSI target for bare metal servers and applications.  Though the file system is exposed to a bare metal workload, all the great Nutanix features are preserved for them (snapshot, clone, data efficiency services, etc).  Again, this is a demonstration of “power through software” and the evolution of the platform – these features and support were first available for VM’s, then files, and now bare metal.

The Nutanix file system is presented via the iSCSI protocol a little differently than iSCSI is normally implemented.  Instead of resiliency built into the protocol being leveraged (ALUA, multipathing, etc).  Multipathing is handled by the back end, paths are managed dynamically and in the event of a node failure, failover is handled on the backend.  While “best practices” for iSCSI are usually well documented based on the vendor or platform, not having to rely on as much client side configuration and optimization removes the human element and thus risk for “PEBKAC” issues.  I’ve seen the “human element” manifest itself in more than one iSCSI implementation.

I see this feature being a big deal for shops that have both hyperconverged and traditional 3 tier deployed in the same datacenter.  Due to some extenuating circumstance (like a crappy software vendor that STILL doesn’t support virtualization in the year 2016) or an investment in physical servers the business wishes to extract value from, physical servers and/or traditional SAN storage must exist in parallel.  Being able to present traditional “block” storage to a bare metal server or app may remove that last roadblock on the journey.

“All Flash on All Platforms”

As the cost of flash continues to plummet, it continues to become more pervasive in the datacenter.  The capacity of SSD’s has surpassed that of traditional spindles (though at a cost right now) and I foresee a day where “flash first” is the commonplace policy.  As such, starting with the Broadwell / Nutanix G5 platform, all flash config will be available on all platforms. 

Microsoft Cloud Platform System Loaded from Factory

This was announced last week, but in a nutshell Nutanix and Microsoft collaborated to offer the Microsoft Cloud Platform Standard (CPS) installed from the factory.  This offers a more turnkey private cloud that accelerates the time to value and allows for “day one” operation. All “patching operations” are integrated into the Nutanix “One Click Upgrade” platform, further streamlining the day to day care and feeding that historically has burned up so much administrator time.

In addition, Nutanix will support the entire stack from the hardware up to the software, just like they have been doing for vSphere, Hyper-V, and AHV already.  There’s a lot to be said for “one throat to choke” when it comes to technical support.  I’m sure we’ve all been an unwilling participant in the “vendor circular firing squad” at some point.  My experience with Nutanix support has always been excellent, both from the technical capability of the support engineers as well as the customer service they deliver.

Prism Enhancements

Some big enhancements are coming to Prism.  Building upon “Capacity Planning” in Acropolis Base Software 4.6, “What if?” modeling will be added.   Instead of just projections based on existing workloads, you’ll be able to model scenarios such as onboarding of a new client or introduction of a new application or service at a granular level.

One of the benefits of the “building block / right sized” hyperconverged model is being able to accurately size for your existing workload, allowing for sufficient overhead, without overbuying based on best effort projections where your environment might be 3 years out.  Calculation based on existing utilization and growth was the first step, and then modeling “what if _____” is the next evolution of accurately projecting the next “building block” required to meet compute, IOPS, and capacity needs for “just in time forecasting”.  Maybe a “buy node now” button is in order through Prism 😛

Network Visualization

Enhancements to Prism will allow for quick configuration and visualization of the network config in AHV – both the config in the hypervisor and the underlying physical network infrastructure.  This makes finding the root cause of an issue much quicker and lowering the overall time til resolution by being more aware of the underlying network infrastructure.  Josh Odgers has a great blog post covering this with some nice screenshots of the Prism UI so I won’t bother reinventing the wheel

Community Edition:

Another notable milestone in the Nutanix ecosystem is 12,000+ downloads of Community Edition to date (and nearly 200 activations a week).  Hosted Community Edition trials will be available free in two hour blocks through the Nutanix portal as a “test drive”.  Another option for getting your hands on Nutanix CE are to install it on your own “lab gear” – Angelo Luciani has a great blog post on using an Intel NUC (are multiple NUC’s NUCii?)

Or….maybe on a drone?


Other Interesting Notes:

During the general session, some statistics were presented regarding the adoption of Acropolis Base Software 4.6.  500 clusters were updated to 4.6 within 7 days of release, and an overall 43% adoption within 100 days.  It was also noted that there was a significant performance increase available in 4.6 and as such 43% of customers received up to a 4x performance increase at no cost – I’ll say it again, power through software.


Another statistic I found interesting was that 15% of the customer base is now running AHV.  I suspect that percentage will increase significantly over the next 12 months with all the new features now native to AHV combined with the ease of “One Click” online hypervisor conversion.

A Picture to Sum It Up:

As someone who’s dealt with infrastructure that was everything but invisible, I think this says it all…


“Power Through Software”:


Other Blogs to Check Out:

I know I didn’t capture all the announcements here, but some other good blog posts I’ve seen today are worth a read…

Josh Odgers: (there’s a whole series here titled “What’s .NEXT 2016”

Marius Sandbu:

Eduardo Molina:

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s