Having recently spun out its developer-focused technology into Pivotal Labs, VMware is back at the DevOps table with a $30 million investment in Puppet Labs, a leading open-source vendor of cloud automation tools. The investment also comes with a partnership whereby the two companies will "work together towards more extensive automation and orchestration across infrastructure and application elements for VMware-based private and public clouds, physical infrastructures, OpenStack and Amazon Web Services." But given that VMware's $30 million investment represents two-thirds of Puppet Labs' total capital raised, the question must be asked: does it also come with serious strings attached?
Knowing Puppet Labs founder and CEO Luke Kanies, probably not.
No One's Puppet
It's not surprising that VMware would want to invest in Puppet Labs. After all, Puppet has become an enterprise standard for automating IT infrastructure. There are scads of big companies using Puppet, including Google, Harvard, and 24/7 Real Media. While the IT automation is increasingly crowded, with Opscode and SaltStack competing for open-source eyeballs and big vendors like IBM and BMC also in the ring, Puppet Labs has become the go-to technology today for companies with serious infrastructure automation needs. With reference to Opscode, specifically, I've heard it said that companies with existing infrastructure use Puppet, whereas startups born in the cloud tend to gravitate toward Chef (Opscode).
All of which fits well into VMware's customer profile. But does VMware fit into Puppet Labs' plans going forward? And how does Puppet Labs retain its independence given that so much of its funding now comes from one vendor, adding to the funding it already raised from Cisco and Google, as well as a smaller amount from VMware in an earlier round?
For those who know Kanies, this is a bit of a silly question.
I've been friends for years with Kanies, and advised the company early on before it had taken funding. Back then it was just two guys (Kanies and co-founder Andrew Shafer, who has since gone on to work at Cloudscaling and OpenStack and is currently working on a stealth startup), and I was fortunate to spend time with them, working through likely business models. Kanies was somewhat reluctant to take funding then, as he wanted his independence. If you know Kanies, you know that independence is a big deal with him. Raised on a hippie commune in Tennessee (yes, really), Kanies isn't the sort of person to sell his soul, or independence, for $30 million.
Not that doing so for VMware dollars would be a terrible outcome. It seems likely that $30M was second prize in VMware's mind; that it would have preferred an outright acquisition. But with the investment, which Kanies says leaves VMware as a minority shareholder, Kanies gets the money to scale Puppet Labs without the strings attached (sorry, couldn't resist). As Shafer related to me over IM:
The cloud transition is creating growth opportunities for next generation systems management products. The alliance will be great for Puppet Labs and great for VMware. It's important to remember that VMware is building a portfolio of core infrastructure pieces that are rooted in open source, and has given them some autonomy even after acquisition. Puppet Labs has a preexistent relationship with a number of these open-source communities, like OpenStack, which should be complemented by this investment.
Or as Kanies tells me, when asked whether this investment means he's locked into VMware-only infrastructure: "Puppet's strength is in supporting heterogeneous environments, whether physical or virtual servers running on-premise or in the cloud, we support a broad array of infrastructure and are agnostic as to whether it comes from Red Hat, Microsoft, or VMware."
Translation? Kanies is no tool. Or a puppet.