Broadcom - The Rise of OpenStack?
There's an unintended consequence to Broadcom's approach to private cloud.
Private cloud is hard and you should think twice before undertaking the burden. This may sound odd coming from someone who has preached a hybrid operating model for the past 10 years. The key is that I’ve promoted a hybrid operating model. The private cloud is adopting a cloud operating model across the board.
Let’s get something straight. All of this stuff is hard. However, the private cloud operating model has been problematic from a technology, people, and process perspective. I’ve dabbled in building private clouds the first eight years of the trend. The technology is complicated, but the people and process parts are even more difficult.
It’s difficult to find people who want the nonstop momentum of private cloud operations, and it’s even more difficult to create processes at scale to support that technology momentum.
VMware Cloud Foundation is more than adopting technology. It’s adopting a cloud operating model. VCF may indeed simplify the technical installation of a private cloud. But the level of lift is still great.
In more conversations, customers are re-evaluating the technology foundation of a private cloud journey. Yes, that means considering OpenStack again. If customers have to go through the disruption of adopting a private cloud, re-evaluating the technical go-to-market partner makes sense.
VMware has undertaken the challenge of communicating the value of a private cloud and convincing customers that their technology ecosystem is the correct choice.
What have you found? Are you following the Broadcom message? Are you considering a private cloud, and have you considered alternatives such as OpenStack?
Reply to this email or comment below. They go directly to my inbox.
It is going to be a very interesting next 3-5 years and I'd be curious about your take. VMware liked to point out the massive adoption base, like 99 of the Fortune 100. Hock Tan arrived and stated that vSphere etc was massively overpriced. After some thought this seems obviously true, in hindsight. If _everyone_ among a group of very sophisticated buyers and negotiators is buying your product, they probably are confident they are getting a good deal.
And where are the viable competitors? Microsoft gave up their hypervisor offering. There are some things like OpenStack, and a handful of smaller players. Where's the near-peer competitor? The one aiming to take out the slightly-more-price-sensitive companies?
There isn't one. (At least I don't think there is -- I'd be interested in you correcting me.) Why not?
Building a vSphere / VCF competitor of quality would be very expensive. Let's call it a billion dollars - I don't know if it's that number. But nothing insurmountable by well capitalized large tech companies. In order to do the project, they would have to look at the total addressable market, the value captured by VMware, and what excess value they could go after. Understanding that a successful product generates economic value, and then that value is divided between the customer and the producer, new entrants would have to look for excessive producer value that they could go after.
But if the customer is already capturing a large amount of the value, there is no incentive for them to move and any value a new entrant could capture to offset the R&D cost is simply not enough. VMware's pricing kept competitors from entering.
Now things have / are changing! I can only imagine there will now be many customers out there looking for alternatives to VCF, with a lot of money in their hands. They're willing to give up more customer value to the producer, up to where VCF is now priced! That's a lot of economic incentive to enter the market.
I am thinking there will soon be a lot of new changes and dare I call it excitement in on prem virtualization. And not necessarily a direct replacement of a new type-2 hypervisor. Maybe Bare Metal Containers? New base OS that includes hypervisor features? I don't know but it will be interesting.
As a long-time advocate of OpenStack, I’m noticing a significant increase in large organizations reevaluating this technology. Even smaller companies looking to avoid the high costs of Broadcom's new licensing model are exploring OpenStack as a viable option for migrating from VMware.
Migrating doesn’t have to be complicated. There are plenty of free and open-source tools available to help transfer VMware images, and a growing number of third-party providers ready to assist with the process.
Historically, OpenStack posed challenges: installations required highly skilled teams, and once you got it running, upgrading could be a nightmare. Older versions were often buggy and lacked the features that VMware users needed.
That landscape has shifted dramatically. Now, there are multiple well-supported deployment tools available, catering to everything from simple one-node test systems to fully supported, multi-site production clusters. Companies like Red Hat, IBM, and Canonical offer excellent solutions, even for those who may not have deep technical expertise.
With the new SLURP methodology adopted by OpenStack, upgrades are not only feasible—they can be relatively smooth (as smooth as any software upgrade can be, at least).
The OpenInfrastructure Foundation's VMware Migration Working Group recently released a whitepaper revealing that over 90% of VMware features now have OpenStack equivalents. As for the remaining 10%? It’s worth questioning how many users really depend on those features and for those, sticking with VMware might still be necessary even with the Broadcom tax.