OpenStack is heading for a massive trauma that will see the Big Tent frame collapse.
Under the Big Tent model, which was introduced in 2015 and saw the end of the integrated release model, more projects are included under the definition of OpenStack. The benefit of this is that more capabilities are provided to users.
While it is only a year old, that model is doomed to fail according to Mark Shuttleworth, founder of Canonical.
Talking to CBR, Shuttleworth said: "If you look at OpenStack it has this frame big tent, well the truth is the tent will collapse and that’s going to be traumatic for everybody.
"Much like 1999/2000 was traumatic for a lot of dotcoms. The dotcom bubble collapse left the good stuff, it was IP, internet connectivity and that just continued to grow."
This is what he expects to happen in OpenStack, not that OpenStack itself will fail. Shuttleworth said that when the Big Tent collapses, OpenStack will keep going. "So don’t confuse this piece doesn’t work with OpenStack is going to fail. That piece will fail, the core is going to win."
Shuttleworth said that a lot of the projects that people are excited about will fail, but OpenStack itself, the Infrastructure as a Service piece will continue to grow and get better.
Essentially Shuttleworth believes that the reason for this is that customers want the virtual network, virtual compute and virtual components are what people want and what users benefit from, as long as OpenStack can deliver this at a cost which is competitive with Amazon Web Service, Microsoft Azure and Google Compute Engine.
"Then we are giving CIOs a real choice. A lot of the other stuff in the Big Tent is kind of bullshit as a service, bubble as a service, hot air as a service," he said.
During the OpenStack Summit in Austin Texas there has been a lot of talk about standardisation, whether that is making OpenStack the standard for private cloud deployment, or standardising around certain technologies within the stack.
This is a symptom of how much vendor aspirations are tied up with the open source cloud, as the reality is that components such as Nova, Cinder and Neutron are all complete, while the community could argue about it, there is no real point, said Shuttleworth.
Shuttleworth, who is the founder of Canonical, the company behind Ubuntu an open source software platform clearly has a strong feeling towards standardisation and removing what is seen as unnecessary complexity.
To this end, Canonical has an open source service orchestration management tool called Juju, which basically allows for the fast deployment of software through Charms – so the company does have a product to push.
In addition to the end of the Big Tent, users should be on a look out for a change of messaging from developer focused to looking at the business use cases. While the keynote on the second day of the summit went some way to starting this, there is a long way to do.
The Canonical founder said that really it’s all about the apps that can be run, not spending money on making discs spin and lights blink as he put it.
"Most of the conversation here is kind of dominated by features and extensions to OpenStack, which is the stuff that vendors want to sell, it’s not the stuff customers want to buy. Customers want to buy is network, compute and discs and then they’ll need to be able to run my apps on top of it."
So the shift in conversation is something that is healthy from the community, shifting away from yet another project inside OpenStack to how to be successful running apps on top of it.
During the conference a lot of focus has been on OpenStack being deployed as a private cloud, while there are also public cloud deployments, it’s safe to say that the majority are as private – hence the calls for OpenStack to become the standard for these deployments.
Shuttleworth said that originally it was felt that there needed to be a big OpenStack public cloud for the technology to be a relevant API. However, a CIO would want to tap into the economics of a public cloud without being locked into one single provider. This effectively leads to a decision to adopt at least two vendors when moving to the public cloud.
The hope for OpenStack is that it is simple for it to become the on-premises choice for businesses looking for a hybrid solution, but it isn’t necessarily a problem if there is not a public cloud that speaks the OpenStack API.
"Look at the Google announcement that you can back-up your OpenStack storage to Google – what a great story. You’re seeing Google, which is a great public cloud, linking into OpenStack as a private cloud – that’s good for both parties.
"This is kind of what I mean about really it boils down to OpenStack being competent, reliable and cost effective for compute, network, discs – get that right and everything else works."
Although changes are in the air for OpenStack, the users shouldn’t necessarily be worried but perhaps the vendors in the Big Tent should be.