When OpenStack's board of directors knocks, it's time to listen.
Wow. With friends like these tweeting stuff like this, who needs enemies? Or competitors?
One thing I love about Mirantis co-founder and CMO Boris Renksi is that he is very candid about the industry, and his company's place in it. Maybe too candid.
For example, Renski recently positioned Mirantis' latest distribution of OpenStack by declaring its "most important feature" to be the fact that "it actually works." This wasn't a backhanded compliment of Mirantis, which has proudly offered a solid OpenStack product for some time.
No, Renski was commenting on the state of OpenStack. In other words, Mirantis' distribution works, but the rest...? Not so much.
Of course, he would say that
Renski isn't just outspoken, he can be an outspoken critic of his competition. So, perhaps it's not surprising for him to say that Mirantis' OpenStack distribution is fantastic, while everything else is dross.
Renski isn't alone in pointing out flaws in OpenStack. In the seminal State of the Stack keynote at the most recent OpenStack Summit, OpenStack board member and visionary Randy Bias proclaimed that "OpenStack is at risk of collapsing under its own weight."
OpenStack's greatest strengths—community and fast-moving development—are also its biggest challenges, as a lack of focus on interoperability has left a sometimes creaky mishmash of conflicting projects that carry the OpenStack brand but don't work well together.
This has resulted in OpenStack, despite its community enthusiasm, being adopted at a relative tortoise pace, especially when compared to simpler technologies like Docker and MongoDB.
As simple as the public cloud
I know the private cloud rearguard hates to hear it, but public cloud adoption continues to hurtle forward, even as private cloud adoption limps along. As Gartner data shows, private cloud VMs have grown 3X over the last year, which sounds good until you compare it to public cloud VMs surging by 20X in the same period.
Now, look at the companies supporting OpenStack vs. those going all in on public cloud, where most of the market growth is. For OpenStack dabblers, like IBM and Rackspace, their lack of commitment to the public cloud will almost certainly cost them.
As Gartner's Magic Quadrant for public cloud storage services indicates (published by Microsoft), the market is going to those that are committed:
As Microsoft notes, IDC is forecasting total digital data to be close to 40 Zetabytes (40,000,000PBs) by 2020, and the "majority of it will be stored in cloud." Not just any cloud: public cloud.
Yes, enterprises will continue to struggle with OpenStack and the private cloud in an effort to keep control of their data (as they suppose), even though the biggest security breaches have been within the data center, not public clouds.
But eventually, the need to have something "that actually works"—and helps drive development faster and more flexibly—will push more workloads to the "just works" environment of public cloud providers like Amazon Web Services (AWS) and Microsoft Azure.
OpenStack's uphill battle
Guess what? It's only going to get harder for OpenStack.
As Piper Jaffray analyst Gene Munster writes of AWS, "AWS remains the king in IaaS" because "many developers swear by AWS's functionality, flexibility, and ease of use." As such, "AWS's mindshare is securely in place to defend against what will continue to be a competitive market."
Meanwhile, "OpenStack can run a fine private cloud, if you have lots of people to throw at the project and are willing to do lots of coding," according to Gartner's Alan Waite. That's not a recipe for long-term success. Again, no matter how committed an enterprise may be to the idea that it can only protect its data within a private cloud (despite so much evidence to the contrary), eventually that commitment is going to wane if it's simply too hard.
Especially compared to alternatives.
In his keynote, Bias warned that "explosive growth drives complexity" and that "continued complexity slows adoption." Unfortunately, it's difficult to "simplify, kill, or re-architect 'broken' projects,' he continues, especially using OpenStack's existing, rigid technical governance model.
I don't think OpenStack can match the convenience of public cloud providers. Exceptionally capable and bright people like Bias disagree. If OpenStack wants to have a chance, however, it needs to heed self-critical appraisals from Renski and Bias and take appropriate action.
Otherwise, the public cloud will win by default, and not merely on merit. It's not enough to finally "actually work." The enterprise demands better.