IBM is asking that people judge the OpenDaylight SDN project by its accomplishments, not by its roster.
Seeking to quell lingering questions about the consortium's formation, intentions and goals, co-founder IBM addressed the issues at this week's Open Networking Summit SDN conference. OpenDaylight was founded by Cisco and IBM to, as the group says, define an open source SDN framework to stimulate SDN application development.
But the group, populated by vendors who have a lot to gain or lose through the SDN movement, is viewed skeptically by some who see it as an attempt to blunt SDN's momentum and impact. It is also viewed as a possible SDN standards alternative to the user-driven Open Networking Foundation, a strong proponent of the OpenFlow protocol which Cisco is particularly critical of.
"We've been deluged with questions, inquiries and misinformation," said Inder Gopal, IBM vice president of system technology, during his ONS address.
The OpenDaylight group is working to define a common industry-supported framework for SDN, Gopal says. And even though the consortium's work is proceeding under the auspices of the Linux Foundation, OpenDaylight "has nothing to do with Linux," he says.
"It is not intended to work exclusively with Linux," Gopal says. "The Linux Foundation has enormous experience and credibility with open source projects."
OpenDaylight will be available under the Eclipse public license in the third quarter, he said.
The scope of the project is to include an SDN controller, northbound and southbound APIs - including OpenFlow - proprietary extensions, and east-west protocols for federation between controllers. Members are contributing money, developers, intellectual property and code.
"This really is about code, contributing and writing open source code," Gopal says, adding that no other open source project started out with as much interest and participation from an eclectic group of vendors.
So what's the angle? There is no angle, honest! one of Gopal's presentation slides exclaimed.
"We think this will benefit the industry," he said. "Ultimately, you'll have to judge us by our actions. You'll see a very open, participatory process not driven by vendors' interests. It will be transparent. It will be the merits of the decision that drives the decision rather than politics or proponents."
That has been one of the biggest concerns about the OpenDaylight project: that it will propose a particular vendor's or vendors' technology as the "standard" for implementing SDNs.
The value of the OpenDaylight framework will be that it speeds up the adoption of SDN, Gopal says. He says that some potential customers have been less than eager to jump into SDNs "with both feet" because they feel it's unproven.
The OpenDaylight framework will enable faster innovation by vendors, interoperability across vendors, acceptance of APIs through a "broadly accepted" software platform, rapid access to new technologies and no vendor lock-in, Gopal asserts.
"Customers don't feel lock-in with open source," he said.
The goal is to produce code of the major common components required to build an SDN, Gopal says, deployed as is.
"It's about code," he says. "Vendors are to use OpenDaylight code as part of commercial products with differentiation above, below and around" the framework. It will also result in a community of code contributors and commercial vendors, he says.
"It is a meritocracy," Gopal said of the OpenDaylight decisions on which code to include in the framework. "Once it becomes Daylight, it is Daylight code. The best code wins."
Currently, an OpenDaylight technical steering committee is sorting through the contributions to determine what will make up the framework.
Doubters should give the process time, Gopal suggested.
"Come back in two years and grade us," he said, to which he concluded his address with a slide of open questions, such as why there were no users in the group; and how it might work with the Open Networking Foundation and other SDN standards organisations.