It is not unusual for vendor events to focus on the latest technological developments the company has introduced that make the jobs of the delegates easier, faster, more productive and/or provide their employers with richer, more comprehensive returns on the investments already made, or planned for the future.
The recent Nutanix.NEXT conference held in Nice was, in one way, no different, but in conversations with both CEO Dheeraj Pandey and Sunil Potti, Chief Product and Development Officer, it was again an opportunity to dig underneath.
This exposed some interesting nuggets, not least being the notion that the time for getting cloud into real production work, so much so that there really should no longer be thoughts of a ‘Plan B’. There was also talk of licencing technology out, though this might be a couple of years away at least to come to fruition.
But it started with Dheeraj Pandey talking about the company’s current audience, and the shift – for both them and the company – that is coming their way, and the reason that lie behind it:
It is not developers sitting there it is IT/Ops people and we are trying to convert them into cloud ops. Nutanix is removing the siloes found in traditional IT architectures and merging them into one environment, so the issue is to help these people retrain and reskill so that their careers can be elevated. That is where we are heading, but we are not really there yet.
Get grooving to the DevOps cadence
One reason for that is that he sees the company being still at the mercy of the on-premise customer, which have their own cadence of working with and accepting change. Now he sees the times changing to where there needs to be a different, DevOps cadence. This is one which takes over responsibility for many tasks that have been end user problems and part of the daily routine with on-premise customers.
So over the next three quarters the company expects to change a lot of the ways they have operated for on-premise customers, to the point where a swing away from on-prem as the core business target seems likely, though they will not be forgotten or dumped. Pandey said:
This is what hybrid really means: users need to operate both sides of the line, though they are very different, the cloud side wants to change all the time, while the on prem side is resistant to change. So maintaining the cadence both ways is going to be one of the key challenges ahead.
Part of that change of cadence was identified by company President, Sudheesh Nair, in the second day keynote. One of the key goals for Nutanix is to make as much of this on-premise management disappear from the users’ point of view. The conceptual model: whether tasks are local, dispersed across a large diffuse network or centralised in a bare metal private cloud or indeed on premise shouldn’t really matter any more to the user.
They should no longer be thinking about the issue of ‘where’, with what, or on what their data is processed, let alone fretting about the management and control of such issues. That does give Nutanix some complex issues to work with, such as the need for users to build architectures of multiple clouds – and arguably multiples of multiple clouds, against a background of staff still largely grounded in IT/Ops, as Pandey pointed out:
There are laws of physics to contend with, of course. The applications have to deal with data that is local to the application, otherwise the applications are not doing anything, not being used. The same thing happens with latency, because decision making has to be in real time. So computing needs to be local to the data.
Look at data sovereignty, eventually we have to make it run somewhere. The mundane now starts to matter, things like infrastructure, governance, what does that all mean in terms of exploiting the data effectively? The business we are in is the provision, deployment and continuous updating of the systems needed to make this all possible. And we are dealing with an audience that still wants to know about the details of how it is all going to work.
No Plan B
Sunil Potti highlighted why getting users (both as customers and as coal-face staff) through this transition phase is becoming important. He sees now as being the time to push cloud into production. If users understand cloud as a full construct they can see that it is no longer about testing the water, or building specialised services: all of that has already happened. Now they should start just using it, because in his view, it is now a good problem to have:
Users are entering the time when there is no plan B. It’s not going to happen next year but I think we’re going in that direction.
This is quite a sweeping, challenging statement that begs an obvious question: does this require new developments from Nutanix or is it some conceptual, philosophical understanding shift that we have yet to see? Potti said:
It is happening I would say because of the work that has been done by various vendors, Nutanix is only a small piece, probably one of the harder pieces was tackling the legacy part, which is like the core. There are obvious examples of where Microsoft, Google and Amazon are helping to move the bar, for different reasons, but we have said look, let me at least reduce the burden inside. But I do think it’s happening in the mainstream now, and folks are comfortable with not having a plan B. I think consumerisation of the end-point has helped accelerate the consumerisation of the data center in many ways.
This will require more movement yet in the direction of what Sudheesh Nair has called true hybridisation. This the step that does effectively make much of the cloud infrastructure disappear. Currently, according to Potti, though it is possible to run the same application both in the cloud and on-premise, there is still usually work to do around it to ensure it works properly in either environment. True hybridisation is about getting to the point where everything – the infrastructure requirements, the load balancing and optimisation – follows the app.
There is, however, an acknowledged downside to this: it has the potential to be a lock-in, for if it is to work well and seamlessly it has to run on a Nutanix-only environment. But Potti sees an important second option: a longer-term position that he suggests Nutanix is looking for. This is the potential to licence its technology to spread its use:
You have to get to the second option. You cannot be forcing people to use the physical implementation of the same stack, but you want them to use the logical implementation. But to convince the big guys that they have to align on the logical implementation you have to first show them that 10% of the model works. They have to understand that all the Implementations have to be similar. Then, over a period of time, they will retro-fit their current fabric to support that implementation construct. And I personally do agree that while we first teach them or show them the way as to why this is needed, in a year or two from now, if we are successful, the other option becomes achievable because that is the right option.
And we don’t have to worry about fifty clouds. There will only be three, maybe four, that occupy 80% of the market. And once they adopt it the remaining 45 cloud providers will follow.” He sees this taking another two to three years to come to any sort of fruition, as it will need to follow what he calls the `old fashioned way’ of showing the value of the two stacks, with the expectation that the cloud vendors will join in. And he already sees indications that they are getting the idea.
Nutanix events are developing a character of their own that is marked less by the technology developments, though they are certainly present, than by the way the company seems on a quest to find the best practical solutions that bridge the operational gaps between on-premise, private and public cloud services and moves users to a world where such issues are barely relevant to the task of achieving the best business results. And all the while it wants to take the coal face workers, many of whom could be left behind, with them on the journey.
Image credit - Free for commercial use