A while back I wrote a blog post that hypothesized why VMware might pursue a spin-out of CloudFoundry (this was prior to the official spinout announcement into the ‘Pivotal Initiative’). My hypothesis was that there was a natural conflict of interest between VMware’s commercial desires and CloudFoundry’s direction. It would be impossible for VMware to make real strategic moves in PaaS and Cloud since it would likely lean toward its legacy hypervisor business rather than lean toward CloudFoundry. Additionally, the ‘ecosystem’ that CloudFoundry has was in direct competition with VMware; any PaaS service or product that VMware would bring to market would be in direct competition with one or more “partners” (the double quotes are to indicate a severe misuse of the word – if you didn’t catch that)
Now that CloudFoundry is officially part of the ‘Pivotal Initiative’ and in the EMC-proper family, VMware shed its strategic anchor. Its first major move? Invest $30M in PuppetLabs.
Why is this interesting?
At the last VMworld, VMware announced that they would offer a PaaS service at the end of 2012 and commercial enterprise private PaaS “sometime in 2013.” Clearly, they are still interested in CloudFoundry – and Pivotal would be silly to not leverage the VMware sales force. At the same time, members of the CloudFoundry team positioned CloudFoundry as a better approach of managing infrastructure and environments when compared to Puppet. This leaves VMware in a weird position – will field sales push the “better” CloudFoundry, or the newly minted vested interest in Puppet Labs? When a customer asks “should I use Puppet or CloudFoundry?” what will sales say? This is an awkward position that will lead sales to choose the path of least resistance, and confuse the customer.
At a high level, this is likely a hedge by VMware that insures that they (and sales) can answer cloud automation and heterogeneity questions at a commercial level, regardless of the workload. On the surface, this feels like a lack of faith in CloudFoundry or a lack of faith in PaaS being able to impact their business (which would provide even further explanation to the spin-out move)
Now, if the CloudFoundry PaaS was more of an application server where the purpose was to be the foundation for next generation application development, rather than depoy and scale automation, it likely couldn’t be compared to Puppet at any value level. The fact that you can pit the two against each other not only conversationally, but at a workflow level, is the crux of this whole issue.