Gartner: OpenStack in the enterprise? Ha ha ha, you must be joking
'Don't believe the hype'
The OpenStack world has come in for criticism from a Gartner analyst because the claims made by companies backing the open-source project frequently don't line up with reality.
In a forthright post published on Tuesday Gartner analyst and research director Alessandro Perilli chided the OpenStack community for a lack of clarity, lack of transparency, lack of vision, and lack of pragmatism.
OpenStack is touted as being an open-source infrastructure deployment and management engine that lets companies load up a free software suite to help them take on proprietary cloud giants such as Amazon, Microsoft, and Google.
But besides Rackspace and HP, few public implementations of the product exist, and it doesn't seem to have accrued the level of cash necessary for it to slow the rise of the mega-clouds.
Perilli put some of the blame for this slow growth at vendor marketing (and credulous tech press reports) that have claimed OpenStack is making huge strides into the enterprise, when in fact it's barely there at all.
"Don't believe the hype generated by press and vendor marketing: OpenStack penetration in the large enterprise market is minimal," Perilli said. "Yes, there are some ongoing deployments that can't be disclosed yet, but for one promising or successful deployment there are several that fail and that will forever remain undocumented."
The criticisms line up with a recent wave of discussion that has percolated through the cloud industry as the community starts to wonder why after three years of major development the OpenStack technology is still lacking key features at its core, and why so few people seem to be making much money from it.
Some of OpenStack's problems lie with its modularity, as this means the tech can be sliced and resold in a variety of ways, which leads to the development of confused strategies from vendors. It is often said that the project needs a Linus Torvalds-like figure to drive central development of the project and guard against gratuitous feature-creep.
On the same day as Perilli's post, a Red Hat cloud evangelist published a personal post lamenting the way the press has criticized the OpenStack community. The post called out El Reg's own reporting on the half-formed "Solum" platform project.
"Amazon’s ambition and their rate of innovation beyond their core market is not seen as something that will sink them, instead it is lauded as one of the key reason why Amazon will be difficult to compete with in the future," Diane Mueller said, while lamenting the criticism OpenStack has come in for for its feature creep.
The difference between Amazon and OpenStack, though, is that Amazon's core services such as EC2, S3, and others are stable, while some of OpenStack's core tech such as its Neutron networking layer are very, very weak.
Some of its new projects are troublesome as well, including the aforementioned Solum, and the Ceilometer monitoring solution. "Ceilometer is a tragedy masquerading as a farce," said Andrew Clay Schafer who used to be involved in OpenStack when at CloudScaling. "In my opinion, this project should not exist and as it exists should not be relied upon for anything, much less billing customers."
The main issue facing OpenStack, aside from the tendency for the community to overstate its enterprise success, is the level of feature-creep that has caused the creation of so many shonky products, Perilli said.
"For way too many, [in the OpenStack community] it's all about number of code contributors and simplifying the installation process, in this exact order," Perelli wrote. The "number of code contributors doesn't tell anything about vendors’ vision and long term differentiation."
OpenStack's success matters, as the better it gets the more Amazon and Google and Microsoft will be forced to compete with an open-code solution. If successful, it will bring more transparency and competition to the cloud market, which is a great thing for buyers. But at the moment the community needs to sort itself out and work out not only what it is trying to do with OpenStack, but how to fix the core components so that it can sell to serious enterprises. ®
Sponsored: Hyper-scale data management