This week saw established analyst firm scrambling to recover after realising that OpenStack is a huge, unstoppable thing. Sean Kerner writes that 451 Research released some data :
451 Group now reports 2015 OpenStack ecosystem revenue at $1.2 billion and forecasts it will grow to $3.37 billion by 2018. From 2014 to 2018, 451 Group has forecast a compound annual growth rate (CAGR) of 31 percent.
Its becoming clear that a Gartner got its predictions on OpenStack quite wrong as recently as a year ago: 
A billion IT market is a business opportunity not a science project its a major revenue stream. Even more incomprehensible is that OpenStack was clearly building up critical momentum as a smart alternative to poor quality cloud products from VMware among enterprises.
OpenStack Pushers
But the overwhelming push behind OpenStack is not coming from Cloud Providers or Enterprise but from telecommunications service providers / carriers. They have thousands of “data centers” (called a central office) located near to the network edge. The market for bandwidth is reducing in price and margin which means that the platform they use must be low cost. Thats Openstack.
The Etherealmind View
I don’t understand how Gartner can get this so obviously wrong. OpenStack momentum has been clear for three years as it moved through the messy sausage-making product development phase.
Oh, and its not just one analyst. There are dozens of analysts that contribute to the group of official opinions.
Thanks to Stu Miniman for the tip:
Last year Gartner said #OpenStack is a science project – this year they are part of the opening keynote https://t.co/8f6snC4Tnk
— Stuart Miniman (@stu) April 25, 2016
EMA saw significant OpenStack activity 2 years ago among enterprises that were pursing so-called SDDC initiatives. And this year we surveyed network infrastructure teams about whether they were supporting hybrid cloud initiatives. Among those who were doing hybrid cloud projects, 28% had implemented OpenStack Neutron and 32% had implemented commercial OpenStack networking plug-in and APIs.
Also, a few months ago we surveyed 76 communications service providers who were early adopters of SDN. 61% of them said they were using SDN and NFV to create virtual private clouds with their COs and overall network infrastructure. And 46% of them said they had adopted OpenStack.
Gartner completely predicted that success….said no-one ever! Gartner don’t have a great track record on predicting open-source related success until it actually happens. It’s one of their blind spots although it normally takes them somewhat longer than a year to revisit their predictions and revise them to make them look like they were right all the time…
The idea of sticking compute/storage into VSOs has been around since forever.
It’s pretty much a non-starter. First-difficult to monetize, second-difficult to implement, third-most VSOs would require capacity augmentation which costs even more $$$.
Add the unions and this idea is DOA.
Separate comment–the problem again is that the business case for a private cloud (whether OpenStack or some other) is extremely weak.
Companies can make massive staff & CapEx investments into private clouds or they can just shed all that baggage and load up into AWS/Azure and call it a day.
The financials for private clouds simply don’t work out. Telcos are some of the very few enterprises where this makes sense and even then, it’s simply a COGS-based, IT-run project. Not a strategic initiative that is seen as an R&D investment by the finance folks.