Cisco is investing heavily in the open source software movement as it makes the transition from being primarily a networking company to an IT and services company.
Lauren Cooney, the company’s recently appointed senior director, open source strategy, told ComputerWorld Australia at the recent OpenStack Summit in Austin Texas: “We are making a very considerable bet on open source… This is absolutely critical for Cisco over the next couple of years as we shift Cisco to begin more of a software and services company.
“Open source is transformational because it allows people to break out of company boundaries and work with people across different disciplines and different industries,” Cooney said. “It is [a] transformation for Cisco, our partners, our users and for the industry as a whole.”
According to Cooney, Cisco has to date not communicated its commitment to open source very well, but this is changing. “One of the little known facts that Cisco has not put out there very well and that Cisco needs to do better is just where we are at with open source,” she said
To this end the company has set up a dedicated open source website (opensource.cisco.com). It has also established a Twitter account at @CiscoOpenSource, though it currently only has a small number of followers.
Cooney, who has been working in the open source movement for 15 years at IBM, Microsoft and other companies, joined Cisco about four years ago, but not specifically to work on open source. “I told my boss I wanted to do it and he gave me the job,” she said. She reports to David Ward, chief architect & CTO of engineering.
Cooney said Cisco had ramped up its involvement in the open source community significantly, increasing its membership of open source projects from three to more than 20 in only two years.
“This is not just about joining a foundation or putting money somewhere, it is about code and people and resources, and there is a strategy behind this, beyond just our customers asking us for open source solutions,” Cooney said. “This is very different from the Cisco of five years ago. This is the transition we are making.”
She said that the extent of Cisco’s direct contribution to open source projects could be judged from the number of contributions to Github linked to Cisco email addresses. “Per month we are on average making 1500 contributions per month to GitHub, and a lot of people would be using their personal email addresses, so that probably represents about half of the total,” Cooney said.
However she added that the company’s involvement in open source organisations was “not about Cisco putting in the largest amount of code”. “We want to see diverse participation; we want to see things tested or deployed or the likelihood of getting there quickly and we want to see the framework for a vibrant ecosystem,” she said.
Cooney said Cisco would shortly have more to say about how it expects to generate revenues from its open source involvement. “That will be something we will be discussing in the coming weeks but I would expect it to be more around services and support than having a [software] distribution,” she said.
“There are so many distributions out there already, so why not just fix the bugs, release a stable version of the code, upstream the bug fixes back into the community. Developers are smart enough and the code is good enough to just download a package and run it.”
Asked when Cisco would be making some statements to financial markets about its open source strategy Cooney said: “I think that will happen somewhere in FY17, late 2016 [Cisco’s financial year starts on 1 August].
“Looking to the future we will be making a strong investment in open source. We will be looking a lot at IoT and at PaNDA a platform for data analytics that we will look at releasing later this summer.”
Cisco has already created a website for PaNDA (panda.cisco.com) where it describes PaNDA as “a collaborative open source development project dedicated to provide an extensible and scalable 'big-data’ platform for network data analytics, including for operational intelligence (OSS) and business intelligence (BSS) functions,” with “a strong dependency on the Apache Kafka and Hadoop projects.”