I’ve been watching the open source cloud initiative OpenStack since its inception around five years ago. I’ve taken part in some events around the initiative culminating recently in my attendance at the Vancouver OpenStack summit (disclosure: the OpenStack Foundation covered my travel and expenses to attend the event).
At Vancouver, I was joined by around 6,000 attendees, the majority of whom had invested significant time, not a trifling amount of money and a dose of jet lag to attend the event. Sitting in the main hall for the day one keynote, I turned around from my analyst seat in the front row and marveled at the huge audience who were there to hear about the latest moves in the project.
I reflected with my analyst and media colleagues about the event and discussed the irony that exists. Maybe it’s because OpenStack is an open source project with more than its fair share of weirdly colored hair, piercings, and tattoos. Maybe we find it harder to contextualize something that is “by the people, for the people” rather than a rabidly commercial proposition.
Either way, OpenStack has, since day one, been plagued by questions as to its validity. Every time I attend a summit, I’m amazed by the number of people who ask me whether OpenStack is actually a thing. I’ve never been asked that question when attending Salesforce‘s DreamForce conference for example.
True, it has been frustrating in the past to have a dearth of really solid case studies to talk about. Not the proof of concept stories, but the real solid production ones. This year we heard those – FICO is reinventing its business with the help of OpenStack. NASA’s Jet Propulsion Lab is making very pragmatic use of a hybrid strategy alongside OpenStack to run its missions.
Another difficulty that I have had in the past is in really determining what OpenStack is, and what it wants to be. My thoughts are simple – OpenStack should offer a rock-solid alternative cloud platform that covers the basics: compute, storage and networking. OpenStack should not be dragged by vested interests into peripheral areas – it strikes me that left-field projects like Trove and Solum (a database and platform adjunct to OpenStack respectively) do nothing to increase focus.
I expressed this view to OpenStack Foundation executive director Jonathan Bryce. He didn’t disagree outright and commented that the Foundation is being stricter in terms of interoperability, but in my mind that doesn’t go far enough. Indeed, when expressing this view to other attendees – both analysts, customers, and vendors – there was a general agreement with my viewpoint.
The news from the summit was solid – there were lots of partnerships announced that aimed at helping customers chose and deploy consistent and coherent solutions. In particular Mirantis, the high-flying pure-play OpenStack vendor had pretty much every vendor on the planet issuing press releases about them. If anyone can think of a better validation by the industry I’d like to know what it is.
The community app catalog, while a new project, is also beneficial for customers. It is aimed at enabling users to share templates for setting up different applications and services on the OpenStack platform. Again a customer-focused project that should help with adoption.
There was, as can be expected, much love for containerization and half the conference seemed to be about showing how OpenStack is the natural place for containers to reside. OpenStack is a developer-led project so it is unsurprising that containers, a developer-friendly tool, should have so much prominence at the event.
In the early days of OpenStack, conversations centered around a dogmatic battle of those who support the private cloud, and those who see it as a false cloud. Thankfully most of us have moved on from those days and accept that, for a large number of businesses, some private infrastructure is still necessary. For those organizations, OpenStack offers solidity and stability.
It has to be admitted there are some vendors who will fail outright, or whose OpenStack initiatives will fail. Nebula, the well-funded and well-regarded OpenStack vendor, is no longer. HP, the legacy vendor with so much promise, has had an almost comical series of missteps and blunders. While these blunders have centered around OpenStack, they haven’t been caused by OpenStack itself. Sure there is still work to be done, but OpenStack is, I believe, here to stay.
This article was written by Ben Kepes from Forbes and was legally licensed through the NewsCred publisher network.