First they ignore you, then they laugh at you, then they fight you, then you win. This pithy saying could perhaps sum up (albeit in a somewhat over generalized sort of way) someone like VMware’s approach to OpenStack. Last year they announced their own OpenStack distribution, which does somehow seem to have a bit of the back against the wall feel to it. An organization that is looking to move away from VMware and its high licensing costs will probably still go ahead and do it. Just that VMware’s move appears to delay what looks like an inevitable outcome. As a community driven initiative, OpenStack’s strongest appeal to users is possibly the control and flexibility in the near term and cost reduction in the long term and there is no getting away from that.
At this time, let us step back and look at the market of infrastructure cloud players. An interesting dynamic seems to have played out. Someone like Amazon goes out and “virtually” creates a new public cloud category, becoming the market leader. Easing others like Rackspace to the sidelines. And then if you are Rackspace, you move to plan B, which is to sponsor an open source alternative to Amazon. Works out great as then you have the entire worldwide community creating the solution, pretty much “for free.” The open source branding provides an unbeatable layer of credibility that starts to attract potential buyers. Next step is to actively participate in open source development and build lots of expertise in the solution. Become known as an Open Source leader, which helps strengthen market perceptions. Then create productized versions of the community offering and sell that to users bundling support and services.
Which brings support and services front and center: A constant refrain heard about OpenStack is that it is complicated to install and deploy; takes as much as a couple of years to get going, lots of heavy lifting required. Raising the question: is it really that hard to lower the adoption barrier and make it easier to implement? Probably not. But that would likely not be in the best interest of some like Cisco or HP or IBM as making OpenStack less complicated could impact their services revenues. Instead significant effort in the OpenStack community has been on developing plugins and adapters for the boxes sold by these players, which does appear to fit into their strategy.
Turning to user IT departments; as a private cloud platform, OpenStack could help internal IT functions take on shadow IT spending by business teams by providing a compelling internal alternative to the public cloud. OpenStack seems to be the CIO’s ally, which should explain some of the enthusiasm a number of IT personnel have for it.
Ultimately following the money trail might provide some additional insight. Funding for OpenStack comes from memberships and sponsorships https://wiki.openstack.org/wiki/Governance/Foundation/Funding . Sponsors clearly have an outsized influence over what happens. In that sense, is Open Stack development truly an open, community effort is a question this raises. Or is it one that is indirectly guided and directed by its sponsors?
Perhaps an alternative to the sponsor driven model needs to be considered for unfettered open source initiatives on the scale of OpenStack. Where funding is primarily driven by memberships, also grants from organizations on the lines of the NIST could be a possibility. Adjusting the funding model to unleash the power of the collective should be in all users’ interests.
Ultimately something like Infrastructure as a Service, which is what OpenStack is about, is evidently a commodity. End of the day, tech leadership perhaps should not be about selling commodity services; instead it should be about offering differentiated solutions on top of the commodity stack. The sponsor driven (cloudy) model needs to give way to a truly community based cloud effort.