For most organisations, the preference for their new IT infrastructure has already been decided. Most studies show that companies prefer a multi-cloud infrastructure. 63% are already in multiple cloud environments. However, we also hear that companies are sometimes surprised by the costs and struggle to see all the benefits. This makes it difficult to calculate the total cost of ownership (TCO). This also means that any financial benefits of a multi-cloud infrastructure remain invisible.
We sat down with Nutanix to take a closer look at the TCO. The world of IT infrastructure has changed dramatically in recent years. Not only the way in which the infrastructure is used and composed, but also how many things have been automated. In addition, the way in which the infrastructure is paid for has become more flexible. These are all things that companies need to consider when making choices for their future IT infrastructure.
Doing less yourself
Looking at the innovations in IT infrastructure over the last decade, several things stand out. Many companies have made the step from virtualised servers to a hyper-converged infrastructure (HCI) with virtualised workloads on top. These can be virtual machines (VMs) but also a Kubernetes environment with containers.
The introduction of HCI also made a start in automating many management tasks. The simplification of management, or taking the management of the infrastructure completely out of your hands. Updates to the HCI environments are now almost automatic. The creation of clusters and new VMs is also done with a simple press of a button. The convenience of the cloud in this regard has also come to the on-premises world.
Many vendors, including Nutanix, have focused on developing a so-called single pane of glass, one control panel with which you can manage everything. This means that you no longer have to log in to all the different servers, hypervisors, storage arrays or applications to perform management tasks. Everything is done from the same control panel, if possible, largely automatically or with the press of a button.
Some companies will already have applied certain aspects of these developments, while others are still exploring. The observant reader may have noticed that most of these developments have been applied to on-premise environments. This while most companies now want to move to a multi-cloud infrastructure. The question is, what will this look like?
What advantages and innovations are there for multi-cloud?
Of course, you can opt for the standard solutions in multi-cloud. For example, by combining your on-premises environment with AWS and Azure, or perhaps with Oracle or Google Cloud. Each cloud has its own control plane and integrating or making these environments work together is not exactly easy. Oracle and Google are calling for more uniformity and an open cloud without vendor lock-in, but this is far from being the case. Unless you can run your entire infrastructure on Kubernetes, but we doubt that.
What you actually need is a platform that can run and integrate on all those clouds. Something parties like VMware and Nutanix are working on. Nutanix seems to have an advantage because it strives for simplicity in its solution and works with all kinds of different clouds. It is a one-stop-shop with the necessary flexibility, whereas VMware mainly focuses on AWS and on-premise, with and without Dell EMC. VMware does offer other clouds, but you have to deal with those other clouds yourself and buy licences from VMware for those other clouds.
Nutanix tries to make it simpler by offering the ability to run your environment on-premises or in one or more clouds of your choice, with AWS, Azure and Google among the possibilities. Because Nutanix also manages all those environments, it’s a one-stop-shop, but it’s not mandatory. You may also set up your own Nutanix environment within a datacenter or hyperscaler and link them together. In all cases, the big advantage is that you can manage all your environments from a single pane of glass and therefore also move workloads (VMs or containers) easily between clouds. This is possible because the environments are identical, even if they run at different hyperscalers. Furthermore, most management is automated, from updating the Nutanix software to scaling up and down the infrastructure. With Beam, Nutanix has a tool that can advise in which environments certain workloads will run the best for performance or costs. The man-hours that go into managing your infrastructure will undoubtedly go down.
What Nutanix itself is keen to emphasise are the partnerships it has established with both AWS and Azure. Many large companies have large cloud contracts with Amazon and Azure, including many cloud credits. Because of those partnerships, customers can also buy or fund Nutanix environments using their AWS or Azure credits.
OPEX vs CAPEX
Finally, it’s good to take a moment to look at the billing models because there’s been a lot of buzz about that lately as well. Most of our readers will have noticed the trend of replacing CAPEX with OPEX. This means that CAPEX, financing the entire infrastructure at once, is less common. Nowadays, infrastructure is often paid based on usage in a subscription model, the so-called OPEX model.
Because of this, it sometimes seems that companies only offer OPEX. Even the big hardware vendors are increasingly working with OPEX models. For example, we recently spoke to Dell Financial Services, which now offers complete server racks based on consumption. We also asked Nutanix how they view this.
Nutanix let us know that customers basically always choose the OPEX model for Nutanix, although CAPEX is still a possibility. However, it is often more beneficial for customers to pay based on consumption, especially if they commit to a long-term contract of several years. The latter is fairly common for infrastructure, by the way.
Nutanix itself no longer sells hardware, but partners such as Dell, Lenovo and HPE can all deliver certified servers for Nutanix. We asked how that works. Is that also OPEX? According to Nutanix, it can actually be both. Customers can purchase server hardware for on-premises on a usage basis, but if the customer wants to pay for the hardware all at once, that’s also possible. The advantage of paying based on usage is that the supplier often delivers more, so there is overcapacity, and the customer can quickly scale up if needed. There is no need to ship extra hardware first.
Finally, Nutanix says that they have financial experts who can look with large customers at their budgets, their needs for IT infrastructure, and which form of OPEX, CAPEX or a combined is best or has the most advantageous for them. Ultimately, companies often have a separate budget for investments and one for structural, recurring costs. These are often not yet linked.
The total cost of ownership
The total cost of ownership has thus become much more complex. It used to be a relatively simple addition: the price at the bottom of the quotation form, plus all man-hours for maintenance and the average hiring of external specialists for migrations and complex problems. It has become more of a deductible, with standard consumption, possible peak consumption and the automation of many maintenance tasks, freeing up IT staff for innovative tasks. All in all, the TCO is more difficult to calculate, but the infrastructure is a lot easier and more manageable. So our advice is, always make sure you are well informed, talk to different parties and get their experts to look at your infrastructure. What do you need, where do you need it, and how does it fit your budget? Nowadays, this has become a lot more important than it used to be.