VMware’s Unfair Advantage in Offering Private & Public CloudFoundry

By Atos Apprenda Support

Yesterday, I wrote a post that articulated how VMware’s foray into commercially offering public and private CloudFoundry would be bad for CloudFoundry ecosystem partners that are also offering CloudFoundry based products and services. The gist of the post was that by offering public CloudFoundry in Q4, 2012 and private CloudFoundry in 2013, VMware would be attempting to “be an arms dealer and fighting on one side of the war at the same time”. This will likely disenfranchise partners over time who will find they are competing with VMware in public and private cloud enterprise sales scenarios.

When an enterprise is presented a VMware CloudFoundry PaaS and a CloudFoundry PaaS from an ecosystem partner, they are going to have to pick one (after all, why pick two polyglot PaaS?) and pay for one, meaning the other loses the deal.

Additionally, it will likely confuse customers when it comes time to source public cloud services or private cloud technology since they have to decide between “CloudFoundry proper” from VMware, or the ecosystem offshoots from folks like the CloudFoundry community leads and other partners. At a minimum, most of this will boil into a drastic redesign of the ecosystem (making it MUCH less of an organic construct), or at a maximum, failure of the ecosystem. (Note: this doesn’t mean that companies that are part of the ecosystem will necessarily fail, just that the “from many, one” concept will not be of value)

Although this soon-to-be-evident ecosystem tension alone is a significant problem, VMware will likely exacerbate it by leveraging a structural advantage. What structural advantage, you say? The fact that VMware can leverage its proprietary technology core (the hypervisor, etc.), its management tools, and current market footprint in a way that is not reproducible by anyone else in the “ecosystem.”

Although PaaS need not be designed around VMs as a first-class architecture citizen, PaaS offerings can most certainly add value by directly integrating with the virtual infrastructure it lives on. VMware can integrate their commercial CloudFoundry offering with their proprietary hypervisor and tools through proprietary APIs, something that none of the ecosystem partners can do. This could create a significant product or service disadvantage for partners. On the private PaaS side, both VMware and partners will be going after the same customer base, most of whom use VMware as their hypervisor and private cloud of choice. VMware will be able to go into these accounts with a structural advantage that puts their value significantly higher than that of their co-opetition. VMware’s market share is large enough where this product advantage translates into a defensible structural advantage that they can exploit at the sales and marketing level. If you were an enterprise buyer, would you source your CloudFoundry technology from VMware, who is the keystone of CloudFoundry and supplier of your current set of major IT assets, or a smaller company on the outside that cannot achieve technical fidelity with VMware? This is a major dilemma.

Clearly, VMware will not acknowledge this fact among partners since it would scare the community, and community members are probably reluctant to make decisions that, at this point, they would perceive to be rash. Sometime in the future, however, both VMware and the CloudFoundry ecosystem will have to come to terms with this issue and figure out what the best way forward is. I smell a forked future…

Q: Do you think that VMware’s integrations just won’t matter in the scheme of things? Will the ecosystem find another way to achieve success in the face of potential competition from VMware? …….I’d love to read your thoughts.

Atos Apprenda Support

View Comments
  1. Stephen SpectorAugust 29, 2012

    I have read this blog and the post yesterday as well as AppFog’s blog response and thought I would add my opinion. As a fyi, I previously was the open source community manager for 3 years and the OpenStack community manager for 1 year so I have experience in open source community partners and ecosystems. is currently a community “owned” by Citrix with substantial efforts from Oracle, Citrix, Intel and other companies. Both Citrix and Oracle are competitors in the marketplace but are able to leverage joint efforts to create a platform to build their customized virtualization platforms. These two companies are the primary Xen competitors but their solutions are not necessarily targeting the same customer base so there is not much pressure between the companies in the project. This is very different from CloudFoundry with the various players in the community facing direct competition amongst themselves with little difference in solutions other than language platform. I agree with Sinclair that VMware’s market leadership and customer base gives them an extreme advantage with customers when comparing similar offerings.

    AppFog’s comment that all CloundFoundry sales are good for the community so VMware’s success is their success is an interesting comment to consider. A customer who chooses a PaaS solution is less likely to suddenly change that PaaS from VMware to AppFog for example as there is little benefit for the change. Without the solutions targeting different customer opportunities there is too much overlap and competition for a given set of customers for all the participants to target. With no reasons for customers to switch or be served by different versions of CloudFoundry there is no such thing as a VMware sale helping the community, it only helps VMware.

    Consider this example, Linux distros have a significant amount of options to add around the common kernel so Red Hat, Ubuntu, SUSE, etc offer customers a wide selection of choice.This differentiation between the solutions has driven Linux into various customer solutions with the distros focusing on different customer categories. Thus, the growth of Linux is good for all the distros as the more acceptance in the market, the more that other customers will choose Linux thus creating the rising tide effect.

    Finally, AppFog listed OpenStack as an example of a great community of companies coming together to create a platform that is shared. I would argue that OpenStack is not heading toward a complete platform shared by all participants but instead is offering companies various technical pieces to leverage along with proprietary technology which in the end will not create a global collection of IaaS clouds that customers can easily port their VMs. This is a significant goal for the project but we are already seeing such variety in the OpenStack cloud solutions that the portability goal is looking less likely as an outcome.

    Those are my thoughts and I believe this discussion is useful to help improve the open source communities and ensure that all the participants are keeping their eyes open when watching how the projects are run.

  2. Sinclair SchullerAugust 29, 2012

    Stephen, excellent thoughts! I think many of the parallels you have drawn are representative of the CloudFoundry scenario. Neither of my recent posts that dig into potential problems in the CloudFoundry ecosystem actually criticize any of the individual CloudFoundry ecosystem partners; they’re all interesting companies who have the potential to succeed on their own (much like any Linux distro). Linux, however, was/is open at it’s core, unfettered by any corporate sponsor that can skew community dynamics (like VMware in CloudFoundry).

    I think you’ve capture my point quite precisely; a VMware sale is good only for VMware, and hurts other partners who also sell CloudFoundry. VMware will likely integrate vFabric and Spring components, creating a flavor of CloudFoundry that is not reproducible by partners to ensure a high level of customer stickiness. This means that if you buy VMware’s CloudFoundry, you ain’t switchin’!

Comments are closed.