The idea behind OpenDaylight is simple. To move the industry forward toward next generation (software defined) networks. That sounds like the ONF a bit, but maybe their play is still to focus on standard APIs --- not sure, but look forward to an announcement from the ONF as well.
Open Source + SDN Controller + Standard Northbound and Southbound APIs + Networking + Industry Incumbents + Industry Startups + Developers = OpenDaylight.
For the details on all of the specifics, I’ll point you to the main website, Linux foundation article, and two other blog posts from Omar Sultan and David Ward that I happen to read in the past hour.
What’s my take?
- Open Source for networking is new, but not so new. Over the past few years, a few open source projects worth noting are Open vSwitch, Quantum (part of OpenStack), OpenFlow, and Floodlight. Most recently SwitchLight was announced. The companies that have been most interested in these have been cloud service providers and Enterprises with large data centers. I think that’ll change over time.
- OpenDaylight is great for emerging companies like Mirantis and Cloudscaling that support open source platforms such as OpenStack or create their own fork to distribute/support to their own customers. As companies like this emerge to support OpenDaylight, we will see more industry wide adoption of these technologies. Mid-market and Enterprises want support, and companies like this will give them that. I have been calling those companies a new type of VAR. They will sell to you and support you, but in reality, they could be providing you with their version of a product making it not only a VAR, but a VAR + Vendor. I like the approach. Would you buy from a company like that?
- On the technical overview page at the ODL website, SDN is defined as “Software Defined Networking (SDN) separates the control plan from the data plane within the network, allowing the intelligence and state of the network to be managed centrally while abstracting the complexity of the underlying physical network.” Can the industry finally agree SDN is the separation of the control and data planes or is it still “more than that?”
- On the Linux Foundation website, it goes on to describe what each vendor contributed to the ODL project. For example, going down the list, I see the following: “contribute experience (Arista)”, “planning to contribute (Big Switch)”, “has submitted proposals (Brocade)”, “has contributed (Cisco)”, “pleased to contribute (Citrix)”, “intends to contribute (Ericsson)”, “is committed to participate (HP)”, “intends to submit (IBM)”, “has proposed (Juniper)”, “intends to propose (NEC)”, “will contribute (Nuage)”, “will contribute (PLUMgrid)”, and “is excited to participate (Red Hat).” - Maybe I’m too focused on verb-tense, but it looks like only Cisco, and maybe Citrix have contributed code thus far.Interesting. The core of the ODL controller is based off Cisco’s ONE controller as can be read in the Cisco blogs. I do look forward to seeing how things could change once all manufacturers, 3rd parties, and independent developers start contributing and getting code approved. AND why wasn’t VMware listed? What are they contributing? They are on the vendor ecosystem in the Cisco blog.
- How much will an individual be able to contribute and get approved to an open source project like this that is backed by so many large companies that are paying up to $500,000 to be a member of OpenDaylight? Why do companies join and pay those big fees? To buy their hardware and maybe in the future, their software since now, after this announcement, many more will see the value of software in the network community. Check out this post from Ivan.
- This is a huge leap forward that can foster an entire ecosystem of companies and individuals and can really have an immense impact on the industry. A few months ago, there was an article that talked about the network industry settling in on 2 or 3 controllers or network operating systems. I can’t find the article, but I think it had 2 from big vendors like Juniper and Cisco and the 3rd being open source, naming Floodlight. The analogy used was comparing the number of OSs used to the number we see on smart phones today, 2 – 3, from Google, Apple, and Microsoft. So, how does ODL impact this? What happens to Floodlight? I’m assuming Cisco is officially the leader in code contributions to OpenDaylight, but will it be the de-facto open source over Floodlight. Maybe we’ll see OpenDaylight, Floodlight, and another be the three main controllers to be standardized on. - More importantly, these controllers will help the industry standardize on northbound APIs that will lead to greater application development across the industry just like we saw with iOS and Android. But cell phones don’t have open source versions – you want an app, you get it through Apple, for example. Will Cisco and BigSwitch end up having competing “app stores” in the long run? We’ll see --- plenty to get done before that.
That’s it for now – I’ll try to keep posts short over the next two weeks since there will be many things to keep up with since ONS is just around the corner.
Thanks,
Jason
Twitter: @jedelman8