PaaS is dead, long live PaaS

I love it when the people you accuse of being example of a sector being dead agree with you vehemently.

reflection of clouds on high rise building
Thinkstock

A few weeks ago I opined, as is my want, about what I saw happening in the technology space with regard to Platform as a Service (PaaS).

As I saw it, PaaS was pretty much a concept that had been superseded by newer approaches to application and infrastructure creation and management. You’d be forgiven for thinking that two of the best known PaaS offerings, Cloud Foundry and OpenShift, would be pretty antsy about such a bold claim.

Surprisingly, that doesn't seem to be the case. In fact, Abby Kearns, the chief executive of the Cloud Foundry Foundation, asked to jump on a call with me to opine exactly why she, who runs a foundation which itself shepherds an initiative that most concur is PaaS, agrees wholeheartedly with my view.

Firstly, it is worth taking a look backwards and seeing where the Cloud Foundry Foundation, and its eponymously named initiative, started. Cloud Foundry was originally a project within VMware. VMware, like other large technology vendors, often incubates projects which are peripheral (and, sometimes, completely orthogonal) to its core business. Cloud Foundry was one of those -- and the PaaS project quickly got widespread support from some big name vendors such as IBM and HP.

Fast forward a few years and commentators, myself included, began to clamor for VMware to jettison Cloud Foundry into its own foundation. I questioned the ability for VMware to contemporaneously support an open source initiative with a wide and diverse membership, while still commercializing the product itself. The only solution I could see was to let Cloud Foundry be managed by an independent foundation, separate from VMware’s own business. Of course, at the same time, many things changed for VMware -- its own Cloud Foundry business was spun out into Pivotal while VMware itself became a small part of the monstrous Dell/EMC merger. Machinations indeed!

Anyway, fast forward to today and the Cloud Foundry Foundation has grown as the project has matured. Some key statistics from the year:

  • The Foundation grew to 70 members, including new participants Allstate, Google and Volkswagen -- at a pace since inception of adding three new members per month
  • The Cloud Foundry project saw a 36% increase in week-over-week code commits
  • The Cloud Foundry developer contributor base expanded to more than 2,100 software engineers with more than 2,300 pull requests accepted from non-dedicated committers
  • More than 185 discrete modular projects are incubating or active within the Cloud Foundry Foundation
  • The Cloud Foundry user community grew to more than 55,000 members from 181 self-organized local groups

So, despite my protestations that PaaS is dead, it seems that Cloud Foundry is alive and kicking. So what gives? Well, according to Kearns:

“PaaS, as a name, as a concept and as a solution is dead.”

That is some fighting talk from an organization that has always, as far as I can tell, supported a project that is PaaS. The statement also puts her slightly at-odds with her technology chief of staff, Chip Childers who seemed to be “all in” on the PaaS moniker when he wrote in a recent post that:

“The PaaS framework is about enabling developers to focus on writing code, or the business logic, while it takes care of details like plumbing. You don’t have to worry about creating your container or how to do log aggregation. Cloud Foundry simplifies much of the setup and management of your software once it’s in the environment.”

Not so, says Kearns, suggesting that for two years the foundation has moved aside from the PaaS acronym, being dubious that it actually describes what the Cloud Foundry platform does. According to Kearns:

"What Cloud Foundry is about is a set of open source software projects that solve problems for enterprises, mostly in run-time. There is a platform. There is no service.”

Which is weird given that the Cloud Foundry Foundation website has the acronym PaaS plastered all over it. Kearns states that this is simply a historical aberration caused because people still use the PaaS term when searching. Notwithstanding the nomenclature, according to Kearns, PaaS is an interim solution and too heavy of a solution.

Kearns points to her view on organizational innovation for driving her thoughts around what Cloud Foundry is, and isn’t:

“Organizations change at a linear rate; technology changes exponentially (see also Moore’s Law). The challenge is how do you manage relatively slow-changing organizations in an exponentially changing technological environment? That’s why developers/DevOps are leading this wave of digital transformation.”

Kearns points to the huge uptake that Cloud Foundry has seen across the various commercial operations selling it and the open source product itself. Apparently the “non-PaaS” provider powers cloud initiatives at public companies with a combined market capitalization greater than $3.2 trillion. Of course that is a marginal sort of a statement since it doesn’t cover how much Cloud Foundry is actually being used. Perhaps more important, a huge variety of organizations have chosen Cloud Foundry-based platforms, from Telstra to Swisscom, from Hitachi to Huawei, from Lockheed Martin to Mercedes-Benz. There is some massive support behind Cloud Foundry, but the question that remains for me is whether these organizations are choosing the platform as an interim step before an (arguably) more flexible, composable and lighter-weight option eventuates.

Questions remain about how much “heavy lifting” organizations will want to be provided by a platform, and how much of this stuff they bake themselves. The reality is that not one size will fit al --l and some shops will want to build every part of their infrastructure while others will be happy to use a less flexible platform that does much of the work for them.

Either way, for the followers of PaaS, it will be interesting to see where that moniker ends up.

This article is published as part of the IDG Contributor Network. Want to Join?

6 tips for scaling up team collaboration tools