The cloud stack market continues to go through waves and gyrations, but increasingly now the future is becoming more clear. As I have been writing about for a while, the number of competitors in the market for “cloud stacks” is totally unsustainable. There are really only four “camps” now in the cloud stack business that matter.
The graphic below shows only some of the more than 40 cloud stacks I know about (and there are many I surely am not aware of).
VMware is really on its own. Not only do they ship the hypervisor used by the vast majority of enterprises, but with vCloud Director and all of their tools, they are really encroaching on the traditional data center/systems management tools vendors. They have great technology, a huge lead in many ways, and will be a force to reckon with for many years. Many customers I talk with, however, are very uncomfortable with the lack of openness in the VMware stack, the lack of support for non-virtualized environments (or any other hypervisor), and a very rational fear of being monopolized by this machine.
Data Center Tools from the big systems management vendors have all been extended with cloud capability at use in both private and public clouds. Late to the party, they are investing heavily and have shown fairly significant innovation in recent releases. Given that the future of the data center is a cloud, this market is both a huge opportunity and an existential threat. Deep hooks into the data center with service desks, service catalogs, automation and orchestration capabilities provide a near term protection. There are just too many trained resources with too much invested for most IT organizations to just walk away.
Unlike the VMware approach, all of these vendors support a more heterogeneous environment – especially CA and BMC. Most support some combination of Xen, KVM and Hyper-V in addition to VMware hypervisors. They are also moving up-stack, supporting integration with public clouds such as Amazon and others, application-level functionality, and more.
OpenStack is the new 800-lb gorilla. In less than 18 months OpenStack has emerged as the most vibrant, innovative and fast-moving segment of this market. Evidence of progress includes contributed code from over 1,000 developers, more than 128 companies in the community, a growing list of commercial distributions from incredibly smart teams, and a maturing technology base that is starting to gain traction in the enterprise. It’s still very early days for OpenStack, but it very much feels like the counterweight to VMware’s controlling influence.
The froth in this market is coming from increasing number of very cool (and occasionally well-funded) OpenStack commercialization efforts. As with most markets, there will be winners and losers and some of these efforts will not make it. This market is so new that whatever shakeout may occur, it won’t happen for a few years.
Other solutions are going to find the going tougher and tougher. Some may be doing well and growing today, but ultimately the market will shake out as it always does and many of these current solutions will either find new use-cases and missions, or they will be shuttered. I have unconfirmed reports of at least two of the currently available stacks on my list being withdrawn from the market for lack of sales. Is this the start of a “great cloud stack shakeout?”
Where are we heading?
The majority of the market in 3 years will have coalesced into three big buckets, and it’s starting to happen now. vCloud, OpenStack and the big data center vendors will rule the roost at the core stack level going forward. The graphic below is not intended to show the size of these markets.
The guys in the “other” category reading this post are probably not ready to hear the bad news, but this is what I believe to be the ultimate state. There will be niche survivors, some who will migrate to the OpenStack island (rumors abound), and others who may pivot to new markets or solution designs. Some are just focusing on Asia, especially China, since it’s more of a wild west scenario and just showing up is guaranteed to generate deals. However, many of them will have gone out of business by 2015 or be barely scraping by. Such is the nature of new markets.
One key distinction with the “big four” data center/systems management tools vendors is that they are not going to be the same kind of open and vibrant ecosystems as OpenStack or vCloud. With their huge sales organizations and account presence, they don’t necessarily need the leverage that an ecosystem provides. Some in the #clouderati community might conclude that they are toast. I’ve heard several say that there will be only two choices in the coming years, but I disagree and do think that the DC tools guys get it now and have a lot of money to invest.
I have this opinion based on spending most of my days working with large enterprises and governments who have millions invested in these vendors, and I expect a fair bit of enterprise cloud infrastructure – especially for their more mission-critical applications – to be real long-term opportunities for the big guys. vCloud and OpenStack will certainly hurt them in their core markets, however, and there will be lots of pivots and new initiatives from these mega vendors to ensure their relevancy for a long time to come.
The market is starting to form up, and it looks like there will be three big segments going forward (and a small market of “other”). If you’re not in one of them, and solidly so, you’re doing something else in a few years. There just won’t be enough revenue to support 40+ profitable and viable vendors. How many will survive? That’s a tough question, but here’s my prediction for the market breakdown in 2018.
OpenStack commercial distributions: 4 viable, 1 or 2 that are clear leaders
DC Tools: 4 main and a couple smaller guys
Other: at most 3, mainly in niche markets
Total: 12 viable cloud stack businesses in 2018
What do you think?