Nutanix Visualizes ‘Truly Portable’ Cloud Applications

Data is distributed. Because we now work with data resources spread out across an inherently distributed topography, we need mechanisms that work with data in its various formats, locales, systems and workflows. Information resources today are spread across different databases, data warehouses, data lakehouses and a variety of different data repositories in remote edge computing units, across mobile devices (starting with, but not limited to smartphones) and through an increasing number of AI engines and the language models that serve them.

There’s a direct knock-on effect here, with the impact felt most directly in the cloud.

Because data exists in so many places, we need to have extremely flexible cloud computing applications that are capable of operating across bare metal (physical cloud server hardware with no virtualization layer), virtualized environments (that form the bulk of cloud services stemming from datacenters) and all shapes of containerized environments spanning the composable computing world of Kubernetes orchestration.

Combine these factors with the ever-competing nature of the three major cloud service provider hyperscalers – and the fact that many businesses will be customers of AWS, Google Cloud and Microsoft Azure – and it’s easy to see why cloud applications need to be inherently portable so that they can run anywhere. The above reality provides some of the backdrop for how hybrid multi-cloud infrastructure management vendors are now aligning their services. The quest for the truly portable cloud application is perhaps typified by developments now happening at Nutanix.

From Bare Metal To Virtualization

Aiming to elevate itself beyond competing offerings from Microsoft with its Azure Arc (Redmond’s take on making Azure management tools work outside of Azure), Google Anthos (multi-cloud tooling for various infrastructure locations) and the Red Hat OpenShift hybrid cloud application platform, Nutanix has now announced its Cloud Native AOS. This is technology which extends Nutanix enterprise storage and advanced data services to hyperscaler Kubernetes services and cloud native bare metal environments, without requiring a hypervisor.

As data becomes more distributed, IT teams are looking for a consistent way to protect, replicate and restore data across Kubernetes infrastructure in datacenters, bare metal edge locations and cloud native hyperscalers. This new service promises to simplify day two operations for Kubernetes applications and their data. Cloud Native AOS extends the company’s AOS software (which can be said to be the backbone of its platform for data, Platform-as-a-Service and AI) to stateful, native Kubernetes clusters in the cloud and bare metal environments.

Nutanix CEO Rajiv Ramaswami has said that Nutanix today is built around a mission to provide freedom of choice at every level of the computing stack with mission-critical security so that the platform becomes the long-term choice for enterprises who want to operate with cloud-native technologies. When it comes to firms that need to modernize their IT infrastructure, Ramaswami says he knows that IT departments need to have freedom of choice to use the hardware of its choice… and many of course have their own installed base of machines.

The technology analyst, advocate and evangelist community has a lot to say on Nutanix. Enterprise Strategy Group’s (now part of Omdia) Scott Sinclair identifies the need for organizations to be able to run their cloud workloads on the hardware of their choice in a hybrid cloud deployment format. This is a message which resonates very clearly with Nutanix’s cloud portability mantra. Sinclair suggests that this is part of Nutanix’s ploy to “chip away” at the VMware customer base i.e. providing customers with an eminently flexible approach to cloud portability. The ESG analyst thinks that this move is key for Nutanix in the face of the company having a comparably and understandably smaller partner and ecosystem.

Ploughing Into Brownfield

“Nutanix understands what it will take for the company to reach the next level, but the journey will not be easy. Nutanix will need to accelerate the already rapid pace it has been on expanding their partner ecosystem in order to help reduce potential showstoppers within brownfield opportunities. In addition, businesses have sunk tremendous levels of investments in people, processes and tools in managing and automating their existing hypervisor environments, making it costly to switch. The opportunity is there, but with each percentage point of share that Nutanix is able to take, the next one becomes increasingly more difficult,” said Sinclair, who is practice director for infrastructure, cloud, DevOps and networking at ESG.

Thomas Cornely, SVP of product management at Nutanix proposes the view that the company has built a “complete platform” for enterprise-grade infrastructure with advanced data services in virtualized datacenters. He notes that Cloud Native AOS provides resilience for Kubernetes infrastructure by protecting containerized applications and their data with integrated disaster recovery between availability zones, clouds and on-premises. He reminds us that customers can build and deploy cloud-native applications with seamless migration of applications and data optimally located across sites, including the ability to move applications back to on-premises containerized environments.

The services here are said to allow software developers to use Kubernetes APIs to automate and provide self-service control over all aspects of data management for their applications. So does all this make a compelling argument for leaving VMware in favor of Nutanix, a subject that has been so widely discussed over the last year?

More Than A VMware Offramp?

“More important than just being some ‘offramp’ for VMware, Nutanix now has to demonstrate an intrinsic value to customers looking to make a move. I am a fan of the company’s Dell PowerFlex and Pure Storage partnership announcements around disaggregated storage. Partly because this allows Nutanix to play in the larger, more business-critical application space – where it has been less than competitive. This can be attractive to a large enterprise IT executive because with VMware – they have to buy into the whole VMware Cloud Foundation experience – and that’s very closed and very expensive,” detailed Matt Kimbal, vice president & principal analyst for datacenter compute & storage at Moor Insights and Strategy.

But can Nutanix innovate like it did back in 2011 when it first started claiming to be a “VMware killer”, a headline which it positively encouraged and jumped on.

“There are always going to be products and solutions that are more ‘pointy’ in nature and deliver greater capabilities in a specific area. For instance, I think it’s fair to say that Red Hat OpenShift is deeper from a cloud-native perspective relative to Nutanix Kubernetes Platform. But does this mean a company should choose OpenShift over NCP with NKP? Well, that is highly dependent on an IT staff that is capable of deploying, optimizing and supporting over time. There is the balance that IT tries to strike between cost, complexity and performance (capability) – that fulcrum point is different from company to company,” added Kimbal. “I think Nutanix has succeeded on the stability front. NCP has a lot of capabilities and is supporting organizations of all sizes and across all verticals – and external storage support deepens its position.”

RedMonk co-founder James Governor comments in line with Kimbal and agrees that there is movement in the market right now, but that we need to consider all the moving parts in this equation.

“Nutanix is executing reasonably effectively, certainly in selling into customers concerned about Broadcom’s acquisition of VMware. It moved faster than some competitors such as Red Hat, for example. But in general, cloud repatriation [and so-called cloud portability] seemed more like vendor hope and hype than reality until recent geopolitical events pushed data sovereignty dramatically forward as a concern for global customers. Now, a strong hybrid cloud story is absolutely a selling point,” said Governor. “Nutanix doesn’t have the same breadth of ecosystem offerings as competitors and pricing transparency or lack of it can be an issue for clients, but if an organization is migrating from Broadcom that’s less of an issue.”

Why Is Cloud Portability Really Needed?

In terms of why cloud portability is really something that enterprise IT departments actually require, can Nutanix justify this practice and validate the requirement for it in real world operational environments? Happy to be voluble on this topic is Induprakas Keri in his role as SVP and GM of hybrid cloud at Nutanix. Having previously worked at TurboTax maker Intuit, Keri reminds us that in practical terms, US citizens really only use this annual tax app and its service layer across around four key days per year.

“Those spikes in usage meant that Intuit’s estate of IT hardware was 95% idle for 95% of the time. That made it a perfect scenario for public cloud, where maximum service flexibility could be exploited for a more cost-effective deployment,” noted Keri. “But let’s be realistic, most applications don’t operate like that. The vast majority of enterprise apps will go through an initial development phase where an organization is working to assess the value it can potentially extract from the app once it’s live. – and this initial stage needs to feature really fast [development] cycles.”

Keri leads us into the second phase of enterprise application deployment, a period where an enterprise that sees value in an app then needs to work out how many resources the app itself needs (is it an app suited for 10, 100 or 10,000 users?). Which leads us to the third stage, which can be referred to as the “steady state” when an application’s value, shape and size is known.

“The first two stages are once again well-suited to public cloud deployment for flexibility during prototyping and experimentation,” clarified Keri. “But in that third stage, the organization should know enough about their application use case to determine whether it should stay in the public cloud, or can be moved to an on-premises deployment. In other words, the third stage enables the company to see whether it is paying too much. The degree of application elasticity and data intensity will in turn dictate the degree to which portability is needed and affordable from a cost-effective analysis perspective.”

The affordability quotient here will be dictated by how much it costs to repatriate data out of a cloud service. Data uploaded to a new cloud service is typically very cheap or free; the cost comes when an organization wants to move its data and be portable. Keri calls it the Hotel California conundrum i.e. you can check out, but you can never leave, because cloud services providers always operate on this totally flexible but pay-to-leave basis. All of which operational mechanics means, cloud does sometimes need to be portable… and when it does, it needs to be very portable.

The View From Partners

“Today’s data is increasingly distributed – it’s not just in one data lake, warehouse or application, which means it’s crucial that companies have flexibility with their cloud infrastructure and have a ‘true multi-cloud’ infrastructure environment that supports ‘cross-cloud’ deployments. This is more than just a convenience; it is a transformative capability that eliminates the boundaries between cloud providers and creates one unified cloud environment that removes the operational complexity of managing data replication and full migrations between providers. With a true multi-cloud environment, the hardest part of any application to move – the data – now becomes the easiest,” said Ben Flast, director of product management at MongoDB, an open source data platform company that works on best practice deployments with Nutanix and others in the cloud infrastructure business.

Flast follows up by saying that multi-cloud not only enables the creation of application architectures that exploit the best services from multiple cloud providers simultaneously, it also delivers resiliency. With cross-cloud failover, in the event of an outage, the MongoDB product leader says that data can be “automatically switched” to another cloud provider in the same geographic region, ensuring uninterrupted service. Finally, multi-cloud allows companies to meet regional and cloud provider preferences making it easy to meet customer demands or comply with local regulations using a single database.

Competitive Market Analysis

There has been a lot of talk surrounding Nutanix and its apparently favorable position in relation to VMware customers in the wake of VMware’s acquisition by Broadcom. As detailed here on Forbes and as noted above, this space is rich in competing technologies, although as Nutanix CEO Rajiv Ramaswami has stated, not all customers will feel the need to move away from VMware, initially at least.

Since its inception in 2009, Nutanix has attracted software engineering teams that want software-defined storage and virtualization-centric cloud management tools. Long associated with hyperconverged infrastructure services (a technology methodology that brings compute power alongside network and storage into a single space), the large installed base of VMware vSAN and vSphere deployments isn’t going away this decade. Dell also works at the level with its VxRail services, software designed to offer pre-configured and pre-tested hyperconverged infrastructure services, which some users will find simpler to use. Comparable technology also exists with the now-discontinued Cisco HyperFlex.

Looking at the vendors Nutanix sits up against, it’s fairly clear which brands are bigger. While that shouldn’t be taken as a direct measure of competency or scope, it would often suggest that the larger firms have more expansive R&D budgets. Let’s also remember that Nutanix has sought to reposition itself as a wider multi-cloud player and not just a hyperconverged infrastructure specialist, but the company’s HCI legacy will likely hang on to its coattails for a while yet.

As Oracle Cloud Infrastructure also gains some traction on the periphery of this space, market watchers will need to start their analysis of cloud repatriation services with Nutanix alongside those public cloud offerings from AWS, Microsoft Azure, Google Cloud Platform. Where the cloud hyperscalers naturally have the ability to offer deeper native integration depth, they will generally also narrow themselves to their own cloud when compared to Nutanix’s inherent portable cloud message and platform agnosticism, which it currently gravitates its corporate message set and platform capabilities around.

As AI-driven operations functions now start to service across all vendors in the multi-cloud infrastructure market, the winners in this arena will arguably need to take on a portion of the cloud portability ethos that Nutanix now champions. In the composable, containerized, compartmentalized and contextualized world of cloud computing, the need to get a “carry out” and move is real.