Update: Checkout our entire OpenDaylight coverage:
We’ve quietly heard rumblings of a potential new SDN open-source consortium, current known as OpenDaylight, for the last month with limited details. Over the last few days the noise level of these rumors across the SDNCentral network have spiked (think google searches, emails, and spike in popularity of our posts about SDN and open-source).
As many of you may know – we’ve highlighted on SDNCentral our perspective on SDN and open-source such as how the lack of open-source reference implementation, the danger of VMware controlling the vSwitch, and how open-source is the biggest risk to SDN (as currently structured) — and are big advocates for viable approach to open-source SDN technologies.
Here’s what we’ve learned about Project OpenDaylight:
- Rumored to be co-sponsored by Cisco and IBM and possibly Citrix
- 100+ participants are current meeting in silicon valley this week of multi-day meetings
- Rumored to include Cisco OnePK and possibly open-sourcing of Cisco’s controller and perhaps include portions of other vendors controllers.
- Specification / software looking to include broad range of SDN technologies including OpenFlow, BGP, and OnePK
- Targeting announcement before the Open Networking Summit in April
If this is true — there are a number of open questions, like?
- Who else is participating? Is Arista, EMC, VMware in (we believe not)? What about Microsoft? HP? Dell?
- What does this mean to Big Switch and their attempt to drive Floodlight as an open-source alternative?
- Who owns what? How can people contribute? How can non-members (like independent software vendors (ISVs) participate?
- Who’s going to support it?
- WIll there be one controller or approach? or will there be multiple tracks for Telco and Datacenter (or other)?
- Who’s leads the consortium? How are decisions made?
Again, if true — there are potentially significant implications for the entire SDN ecosystem including:
- Remember Star Wars — where the Empire destroyed a collection of states — be wary of the consortium’s ability to make decisions and compete with a unified EMC / VMware approach to SDN. Focus on vendors who offer a whole solution to solve your problem and select the best solution for your specific business problem.
- Remember just because a solution is open-source does not mean you are protected from vendor lock-in
- Look for vendors / partner who can develop SDN solutions that are tailored for your specific needs and environment
- Challenge start-ups to answer why you should go with their solution that’s unlikely to be part of the consortium early on.
- Incumbent Vendors
- This potentially feels like other standards / foundations / consortiums — where you need to be present and accounted for — while carefully investigating where you can earn a ‘win’ and build differentiation before you make a major financial committment.
- This could make life very difficult for the start-ups – and makes me glad I don’t work for an SDN start-up. Between multiple vendor spin-in’s and a consortium offering open-source controller and development environments — life may get significantly harder for SDN start-ups to build real businesses. Many of the SDN capabilities starts-up are offering maybe open-sourced and supported by mainline vendors.
- Figure out how to differentiate via apps. Remember that apps today are a custom-development business not a pre-packaged software business.
- The path to success has just moved from great engineering talent and to teams who can have customer access to find repeatable sales.
- Expected contented SDN exits — though at more reasonable valuations.
If true — OpenDaylight could be a model for moving SDN forward — or a diversion to buy incumbent vendors time to further their SDN plans. We’ll share more as we gain more publicly available data.