This is the last post in a series of three articles dedicated to VMware cloud strategy following the VMworld Europe 2016 announcements. After reviewing the VMware Cloud on AWS offering and VMware Cross-Cloud Architecture, it’s time to draw a line and have an overall look at VMware Cloud strategy. Let’s get started!
The Case for VMware Cloud on AWS
As stated on my previous article, the core of VMware Cloud on AWS is to allow customers to extend (or in some cases even replace) some of their datacenter to the cloud by leveraging the extensive Amazon Web Services infrastructure. To be clear, VMware Cloud on AWS is a full VMware SDDC implementation built on bare-metal AWS servers, it is not nested virtualisation (Oracle Ravello is an example of nested virtualisation). You can therefore expect identical performance as you would have if you were to run the same SDDC stack on-premises. VMware claims that besides the advantage of running an SDDC stack/datacenter on a pay-as-you go (OPEX) model, you also get to leverage the full scale of AWS services. I am not an expert but still trying to understand what is the advantage of it. My belief is that you either build cloud-native apps or you don’t, I don’t get the point of trying to bridge the gap.
Looking at the cold facts, Gartner (drink!) said in early January 2016 that in the year 2016, the total worldwide cloud computing market forecast is at cca 203 billion USD, with a forecasted IaaS (Infrastructure-as-a-Service) share of 22.4 billion USD. Considering that VMware Cloud on AWS is an IaaS service that will be available sometime in 2017, and that in 2016 the growth of the IaaS market was said to be cca 38%, at a similar growth rate the forecast for IaaS in 2017 would be cca 30.9 billion USD (I suck at math therefore feel free to correct me, castigate etc etc.).
According to this report, the VMware Cloud on AWS offering will join a market that is largely dominated by Amazon AWS (31% market share), Microsoft Azure (11%), IBM (8%) and Google Cloud (5%), with the next 20 contenders lagging far behind at cca 27% combined worldwide market share.
VMware Cloud on AWS offering, while it finally offers a clear definition of what VMware intends to do in the IaaS space is great but it’s not doubleplusgood. It allows for the extension of one’s datacenter in the public cloud (thus creating a hybrid cloud environment) and it leverages VMware best of breed technologies (vSAN, NSX..) but it’s not going to “kill the cloud market” if I was to use Oracle terminology.
Don’t get me wrong, the technical teams at AWS and VMware certainly did a hell of a fantastic engineering job, but it is not bringing something entirely new or visionary to the market. We should look at it as being targeted directly against Microsoft Azure. For VMware and its customers it presents a credible alternative to running legacy workloads on a single Microsoft stack, thus potentially slowing or halting Azure Stack adoption in the enterprise world. For AWS, it’s potentially extra revenues and yet another way to stay ahead of their main competitor, Microsoft Azure.
The Case for VMware Cross-Cloud Architecture
VMware Cross-Cloud Architecture is the marrow, the quintessence of VMware Cloud strategy. Where VMware Cloud on AWS is a valid solution that will help enterprises to adapt and “tame” the public cloud, VMware Cross-Cloud Architecture (abbreviated as CCA for laziness/ease of typing) might become the One Cloud Management Platform to rule them all.
It seems that when VMware realized that they just can’t catch up with other competitors, they decided that sitting above them and controlling them -while they fight their own wars and support the costs of personnel and infrastructure- might be a better alternative.
CCA has the following advantages among others:
- Abstraction of cloud provider technological stacks for compute, storage and security
- Cross-Cloud workload migration
- Single pane of glass for workload management
- Single pane of glass for billing
- Central console for identity and access management
I do not want to recapitulate here all that I wrote in my previous post about CCA but we can shorten it by saying that it provides an abstraction layer above cloud service providers. We can envision a future where companies may decide to massively and seamlessly move workloads from one cloud provider to another to mitigate attacks or performance issues, to cut down costs, or even to spin up instances at another provider in another geo zone just for the sake of DR purposes. The use cases seem unlimited.
Is VMware CCA the one and only solution available to the market? Not really. Competitors such as RightScale and Scalr seem to be well entrenched in this market, and since I’m not a clouderati I may have very well missed half a good dozen of other valid solutions. The “power” or differentiator of CCA comes with the fact that it leverages VMware massive customer base, that it appeals to VMware traditional allies (the sysadmins & IT Operations) with the proverbial ease of use of vCenter, that it encapsulates VMware NSX into the security layer and it extends VMware classical features such as vMotion and DRS to the cloud.
Max’s Opinion
Many VMware aficionados and likely investors too must’ve had a sigh of satisfaction when VMware finally announced, after so much time, a seemingly coherent cloud strategy. We have to look at VMware Cloud announcements into two axles: the tactic one with VMware Cloud on AWS, and the strategic one with VMware Cross-Cloud Architecture.
VMware Cloud on AWS is not such a deal breaker, but it allows enterprises to slowly extend into the cloud, buying themselves a good 3 to 5 years of time before public cloud adoption becomes mainstream. The future however may not be entirely made of fluffy white clouds. Darker, heavier clouds menacingly loom at the horizon as, after a week of truce and mutual patting in the back, Amazon announced the release of AWS Connector, a tool which has a clear intent of migrating your existing VMware workloads to Amazon EC2.
For Cross-Cloud Architecture, I applaud VMware. They finally understood that they are no match compared to cloud heavyweights like AWS, Azure and Google. Burning money into even trying to catch up on terms of infrastructure investments would be a waste of time, money, energy and no one would buy into such a story.
Cloud providers do not care about interoperability, to some extent. They want your cash, your workloads and if they can keep you captive the better for them. By inserting their cloud abstraction layer between the customer and the CSPs, VMware will allow customers to consume cloud as a commodity, just like they did years ago with x86 hardware. Win for VMware, and win for customers? Hopefully!
Disclosure
I received a blogger pass from VMware, which allows me to access the entire event at no cost. I did cover all my expenses here including travel, airplane fare, taxi and hotel. Yes, taxi because I hate metro. I am blogging out of my own will, without any obligation, and without any request from VMware. This article is independently written and represents only my personal opinions.