Nutanix versus AWS?

I hadn’t heard about this previously:

Nutanix trying to provide an alternative to AWS.

https://www.bloomberg.com/news/articles/2018-04-27/nutanix-s…

This is a concern for me, since Agile engineering has been around for a long time. This is pretty fundamental stuff.


It’s unclear if the launch will slip into 2019. But Pandey said adopting the agile engineering needed for cloud computing has been particularly difficult.

“It’s been a harder problem than we thought,” he said in an interview. “It was less to do with products, but it’s the ‘how’ of engineering. How do you break things down, how do you do smaller releases, how do you have multiple parallel streams of engineering going on at the same time?”

anyone have any thoughts on this?

cheers
Greg

4 Likes

Wow - I am extremely unimpressed with Pandey’s comments in this article.

I have personally been involved in Agile processes since the mid to late 90’s, though we didn’t call them Agile at the time (called Extreme Programming and test-driven development back then).

While many large enterprises are still late to the party - I helped transition the bank I’m currently working for to start an Agile product group 2 years ago - Agile has been around and started getting adopted well over 20 years.

Now Nutanix has some special problems in their hyperconverged systems, combining server, storage, and virtualization. But a lot of work on this has been done a number of years ago. Microsoft Azure has had stateful App Services for a while now, which also combine server, storage, and some types of virtualization. Not exactly the same thing, but a similar problem for Agile processes.

Let’s face it - a company which can’t figure out its basic processes to produce its product is very questionable.

I would also be very concerned about the first couple of versions now. Products that miss critical deadlines like this don’t use the extra time to come out looking perfect. People are going crazy at Nutanix right now - new teams are being hired who don’t have any knowledge of existing system architecture and being thrown in the water to sink or swim. Existing employees are working 60-80 hour weeks. There is a really good chance that the first version that comes out will be buggy. Considering they don’t have their agile processes figured out yet, and it takes time to adopt new processes, it will take time for them to get the processes to deliver new updates seamlessly working.

Now maybe I’m doing a lot of speculating on an article from the CEO, but I would be very careful about putting money in Nutanix until they prove their Xi Cloud services work and are actually adopted by customers.

13 Likes

I’m still reeling from the excitement of Amazon’s AWS earnings and dominance in the cloud. Then I read Nutanix CEO Pandey’s statement, “…So I don’t have to worry about competition.” I’m not well versed on the Nutanix product offering that is so different from AWS that it will “disrupt Amazon”. These bold statements from a company that is just learning about agile process make me far less confident in their direction and allocation of resources.

1 Like

My concern and I need to learn a lot more about it is whether they can both supply end customers as well as compete at the same time.

I might have it totally wrong but it seems like operating like booking.com play where they dis-intermediate the hotels from the customer and sell them rooms but then use hotel capacity to fulfil it but on a more distributed basis (you need 1 physical room at a hotel for a person for a night but storage could be load balanced and converged across data center availability). So they make use of other’s capacity rather than build their own physical data centers.

Like I said I might have it completely wrong but need to go and figure this out more. It could be genius or it could be high risk.
A

Wow - I am extremely unimpressed with Pandey’s comments in this article… Let’s face it - a company which can’t figure out its basic processes to produce its product is very questionable.

Hi Steppenwulf,
Let me preface by saying that I don’t know what any part of the technology is all about.

However (!), Nutanix has a Net Promoter Score of 90 !!! Now NOBODY gets a NPS of 90! It means ALL your customers think you are the greatest. Add to that that Gartner rated Nutanix first on completeness of vision, and first on the ability to execute.

If you think that a company like that doesn’t know what they are doing, I’d suggest that it’s more likely that you’ve missed what’s going on. (That’s my totally ignorant and common sense take).

Best,

Saul

16 Likes

Net Promoter Score

https://www.netpromoter.com/know/

I am not a fan of Nutanix.

We looked at their solution for a couple of things. Our IT group adopted it for some thing, the engineering groups I work with passed on it. It’s a nice idea, but they did a poor job of selling it, and couldn’t justify their price to us. It was far cheaper for us to go with VMWare ESX clustering on bare metal than for the price/performance we were after compared to the premium we needed to pay Nutanix for their “hyperconverged” solution which left us wanting more CPU and memory than they offered.

It does not at all surprise me they’re having problems rolling out a public cloud solution. It’s not at all their core competency. They are focused on data centers and helping them work in a more cloud-like manner. This comes at a huge cost, but there aren’t too many companies out there attempting this. There are two names I’m aware of catering to the private cloud market; Nutanix and OpenStack. And since OpenStack is free/open-source software, it’s a “bring your own expert” solution. Most large enterprises making a business driven decision (i.e. top-down, vs. bottom-up, engineering driven decisions) will opt for Nutanix’s solution.

But helping an enterprise transform their data center in to a private cloud is a very different technological problem to solve compared to building your own public, multi-tenant cloud, complete with all the bells & whistles with APIs to boot.

In my opinion, which, along with $3.00 will get you a cup of coffee at my favorite small coffee shop, this will be Nutanix’s undoing. Their hyper-converged solution requires too much overhead and cost to be successful. They won’t disrupt AWS, Google, or even Azure. Heck, they won’t even disrupt Rackspace’s offering with what they have, and it’ll be another 6-12 months before it’s ready?

Thanks, but I’ll pass.


Paul

28 Likes

Based on Saul’s most recent post (#40718), which I had not read yet, I take back my claim of “this will be Nutanix’s undoing.”.

I’m still not a huge fan of Nutanix, but if there is not delay, and they’re not getting distracted by the Xi roll-out being more complicated, then perhaps it won’t be their undoing.


Paul

2 Likes

Wow - I am extremely unimpressed with Pandey’s comments in this article… Let’s face it - a company which can’t figure out its basic processes to produce its product is very questionable.

Hi Steppenwulf,
Let me preface by saying that I don’t know what any part of the technology is all about.

However (!), Nutanix has a Net Promoter Score of 90 !!! Now NOBODY gets a NPS of 90! It means ALL your customers think you are the greatest. Add to that that Gartner rated Nutanix first on completeness of vision, and first on the ability to execute.

If you think that a company like that doesn’t know what they are doing, I’d suggest that it’s more likely that you’ve missed what’s going on. (That’s my totally ignorant and common sense take).

You may have a point - I may well have missed what’s going on. But it seems like the problem they have fixed is to virtualize physical resources in a new way that can be more efficient. This is what they presumably got their Net Promoter Score on.

The problem they are trying to fix is to create a hybrid cloud management system that can use their data centre resources, along with other private and public clouds. This is where they don’t seem to understand the basic processes to implement an agile cloud-native solution.

These two solutions are completely different - the first is extremely close to the hardware, where waterfall lifecycles are still the norm, I believe. The other is so far from the hardware, you almost can’t get there from here - and the norm is continuous deployment in an agile process.

This second problem has already been solved very elegantly by Pivotal, by the way, in which I have a 30% profit this month. It’s not exactly apples to apples, but is a solution in the same problem space. I still would say Pivotal is a huge winner people are ignoring, by the way.

7 Likes

This second problem has already been solved very elegantly by Pivotal, by the way, in which I have a 30% profit this month. It’s not exactly apples to apples, but is a solution in the same problem space. I still would say Pivotal is a huge winner people are ignoring, by the way.

Hi Steppenwulf, Your comments interested me, but this is what I found out:

Dell still owns 70% of the company post-IPO and 96% of the voting power.

Pivotal is thus actually a wholly controlled subsidiary with a small slice of its shares trading in the open market.

Dell itself is a core piece of its business, providing Pivotal with a large source of its revenues.

That wasn’t encouraging.

Saul

16 Likes

Greetings board. Long time lurker/reader, first time poster. I followed this Nuantix discussion following some of the confusion in the market regarding their Xi Services. It seems the RBC memo discussed in this thread has confirmed the availability dates, but I just wanted to add some additional flavor based on some product assessments and publicly available information.

I appreciate Saul’s (and the board’s) research and feedback on Nuantix’s performance, customer satisfaction, and other extremely relevant factors. It is this thorough and clearly presented info that brings me back to this forum nearly every day. This being said, I do also believe it is important to understand some of the technical details of their current and future product offerings, as this can give us great insight into potential revenue growth and the timing of this potential growth.

I cannot speak to the potential delays of the Xi Services availability, but I can add some feedback based upon my initial evaluation and research of this product. First, the initial marketing of Xi is as an enabler of Disaster Recovery services (DR) for a customer’s private Data Center hosted applications. I would expect that over time, they will layer more features onto this foundation, but initially they have selected a very commonsense and valuable service for many of their existing customers, and likely many non-existing customers.

You can see from publicly available info that to provide this DR service from a customer’s primary Data Center location, Nuantix has partnered with Google, in order to host this DR service via Xi, in the Google Cloud Platform (GCP.) This was announced publicly in June 2017.
https://www.blog.google/topics/google-cloud/nutanix-and-goog…
https://www.nutanix.com/press-releases/2017/06/28/nutanix-te…

The key info gleaned here is that Nuantix is not competing with Google in the public cloud space, but rather partnering with them. As a sidenote, this will likely drive some competition in the Public Cloud hosting space, where I believe roughly AWS has a 60% market share, to MS’s 20% and Google’s 12%. Private to Public cloud migration, and the subsequent “Hybrid cloud management” is still a tricky area for many enterprise customers, and I believe this would be a differentiator for Google, with Nuantix serving as the main service provider.

Another key point, is that right now it seems Nuantix is setting up these services in GCP’s locations in the US only. This is key to note, as it gives us insight into the Global Xi Service availability, at least in 2018. Even though a large amount of customers would likely be in the NVA/West Coast availability zones, this would mean no customers hosting data/apps in Europe or Asia would likely be a 2018 target, until Nuantix and Google expand this service to other GCP Zones.

I know there were some good messages exchanged in this and other threads on Nuantix’s development methodologies and the rumors about product delivery slippage. I cannot comment with insight or research on if this is a current issue in the release of this product. I will however speculate that seeing that Nuantix will partner with Google to deliver this offering for Xi Services, that a large factor in the product readiness is likely the physical setup, testing and workload balancing and scaling across Google’s GCP Data Centers.

The fact that they must have Google support on this partnership in order to operate within Google’s DCs might be a positive factor for this product release. What remains to be seen, is whether Nuantix is going to deliver Xi exclusively through GCP, or via other DCs globally. This direction could be an additional level of insight into the potential service scope, competitive factors and overall revenue potential.

I hope this helps others with some additional fuel in their due diligence of Nuantix and in this tech space.

Notmalware

26 Likes

https://awsinsider.net/articles/2017/07/03/nutanix-extends-t…

While all three major public clouds are part of the Xi plan, most emphasis at .NEXT on Wednesday was on GCP because Nutanix and Alphabet, Google’s parent company, announced a strategic alliance.


It looks like they plan on having a path with AWS and Azure also, just that there is a specific partnership announced at their annual conf last July with Google in particular.

Offering a cloud-based DR service is really smart, because the concept lends to Multi-Cloud.
If I understand this correctly, you could have an on-prem DC, utilize Azure, and have GCP as the DR site via Xi services. (maybe I have this wrong, but that is just an example)

Since Nutanix wants to “be the O/S for the Multi-Cloud” this is basically their way of taking likely existing clients, probably offering Xi services at a discount to get logos on the Xi platform and prove it successful. Then once enough success stories exist, growth could rapidly rise from there, as clients who haven’t already used Nutanix for HCI may be looking for an enterprise multi-cloud mgmt solution, and would then be taking a look at Nutanix for the first time.

This is still pretty new to me, and it will be interesting to hear the next few ER’s to see how progress is going. I am hoping to learn more from my Nutanix channel technical contacts soon too.

Dreamer

4 Likes

“Offering a cloud-based DR service is really smart, because the concept lends to Multi-Cloud.
If I understand this correctly, you could have an on-prem DC, utilize Azure, and have GCP as the DR site via Xi services. (maybe I have this wrong, but that is just an example)”

DreamerDad - I think this is the intent, and what drew my initial interest. With some maturity, you could in theory shut down your secondary DC sites. At worst, optimize them for something other than waiting for a DR event - which I would imagine many IT Ops groups looking at tech like HCI have already done to some degree.

But I think this could also pave the way for services to provide temporary capacity bursting and many other services. Google supports SAP HANA in GCP, so you could theoretically spin up a HANA environment to run your quarterly reporting for two weeks, then shut it all down. This way you host and operate an expensive platform for 8 weeks a year instead of 52. Maybe you keep a small environment up for Dev/Test, or even monthly reporting. The big, resource intensive platforms are only bursted. BTW, AWS can do this too, but the trick here is the automated migration and data replication activity that Nutanix would provide, which to be honest is a PITA. Another topic is how the SAPs, Oracles etc of the world handle the licensing of these situations, but that is an entire other (painful) topic.

I also agree with your point on their expansion plan. Xi could easily be a loss leader to expand their HCI offering. I know other posters mentioned less than stellar experiences with Nutanix HCI, but in my experience, all of the main providers have their issues - and Cisco, Dell(VMW) are not exactly setting the world on fire with their software. Many of the things we are discussing could (should?) have been done already. In my opinion, Nutanix is getting in first with a more complete offering and the backing of a major public cloud provider. Their HW platform is fine, but their software sets them apart, which is why you see factors like net promoter score being stratospheric.

I think the plan for Nutanix to partner with AWS and Azure in addition to GCP makes absolute sense for them, but to me it makes less sense for Google. I am curious about the nature of this “strategic alliance” and if this gives Google a period of exclusivity, or other entitlements. The DR offering with Xi being limited to a small portion of Google’s DC scope hints to me that they are going to watch to see if this is successful. If it is, I really don’t see them excited at the prospect of the playing field being leveled with AWS and Azure for this tech.

Notmalware

13 Likes

This way you host and operate an expensive platform for 8 weeks a year instead of 52. Maybe you keep a small environment up for Dev/Test, or even monthly reporting. The big, resource intensive platforms are only bursted. BTW, AWS can do this too, but the trick here is the automated migration and data replication activity that Nutanix would provide, which to be honest is a PITA. Another topic is how the SAPs, Oracles etc of the world handle the licensing of these situations, but that is an entire other (painful) topic.


great post, thanks!

per your comment above, perhaps part of the Google angle was that Google itself needed a “hot” name in Nutanix to bring more attention to GCP. I don’t see any reason why GCP can’t keep growing alongside AWS and Azure as a Big 3. I don’t see Oracle catching up or IBM lasting.

Google has already shown capacity to provide TPU/Tensorflow for ML/AI.
Quick google (ha) search shows a presentation by SAP touting Tensorflow for HANA…might be exactly as you predicted above.

https://d.dam.sap.com/a/ynJJyP/06_Uncover_The_Power_Of_A_Big…

Dreamer

2 Likes