According to Matt Asay, it's time for Red Hat to step up and make OpenStack work for the enterprise. Do you agree?
OpenStack has tremendous community traction. Now, if only the community of OpenStack cheerleaders could be translated into a similar scale of OpenStack deployments.
While OpenStack user surveys show plenty of commitment to the cloud platform, they also reveal a paucity of significant deployments. It's therefore not surprising to hear Basho CTO (and former cloud executive) Dave McCrory lament that he hasn't "seen an enterprise running production OpenStack at scale."
Red Hat, however, is betting its business on OpenStack. Based on its last earnings call, its faith in OpenStack may be paying off. It remains to be seen, however, whether it will pay off for anyone else.
OpenStack taking off... at Red Hat
While OpenStack remains a bit moribund, it's hitting its stride at Red Hat, according to comments from various Red Hat executives on its latest earnings call.
For example, CEO Jim Whitehurst explained that half of its OpenStack sales were six-figure deals (though how he answered the question implied that a significant chunk of every deal involved professional services to make OpenStack work), and CFO Charlie Peters called out that several of Red Hat's top-30 deals included an OpenStack component.
Finally, Whitehurst revealed that "the number of times [a] top-30 deal included OpenStack or OpenShift this quarter tripled from Q4 a year ago."
All of this is great, but insufficient.
After all, Amazon Web Services (AWS) is already a $5 billion business, and it's still growing at 51% year-over-year, according to Synergy Research. Indeed, AWS may well be the fastest-growing software business in history.
Microsoft Azure, for its part, is smaller, but it's growing at nearly 100% year-over-year. Their growth reflects a hearty appetite for public cloud computing, while Red Hat peddles private clouds to CIOs that still believe they must run behind their firewalls. It's a growth business for Red Hat, but given private clouds' dismal success rate, it's unclear for how long.
Hold the cheerleading, please
The immediate runway seems sufficient for Red Hat's growth needs--or will be, if the open source leader can make it work.
After all, this is what OpenStack dearly needs: leadership, vision, and management. As Gartner analyst Lydia Leong recently noted, "The relentless cheerleading for OpenStack seems to have faded into practicality with a touch of resignation."
The reason is clear: OpenStack remains a work in progress, and that progress has been slow.
The problem, ironically, may be the very community that makes OpenStack so popular. Or maybe it's more accurately described as an open-source project ruled by committee, one where project leadership seems as much tuned to corporate contributions as merit.
One OpenStack user described a tortuous experience of "monkey patching and janking around" to get the networking model to work. As he said, "You are set for pain the moment you don't want to run everything on a single Ethernet segment."
Another contributor complained that "OpenStack never had a solid, centralized architectural vision. Anyone that attempted to contribute architecturally was essentially ejected." He and others bemoan the politics (and cash) that govern contributions, rather than the more typical open-source model that holds that good code is the best currency.
And still another OpenStack participant suggested, "One often gets the impression with OpenStack that there's so much attention being paid to new stuff that none of the existing stuff (even recently new) is ever brought to a state of stability and usefulness."
This is no way to build a product.
Red Hat to the rescue
Red Hat seems to appreciate this. On the Red Hat earnings call, EVP of Products and Technology Paul Cormier was asked about "the biggest pain points that customers are facing right now" with OpenStack, to which he responded:
"[T]he thing we're concentrating on is just as we did with Linux moving to RHEL [Red Hat Enterprise Linux] is installation configuration [and] management around the code itself. So, making it more consumable to the masses. That's frankly what we do with open source."
Let's just say OpenStack needs a lot more of this "making it consumable" action from Red Hat... and a lot less cheerleading.