SDN Needs a Charles De Gaulle

There were a pair of stories in the trades last week in which VMware finally took off the gloves and effectively proclaimed themselves the antithesis of Cisco in their approach to—well, not SDN, since both companies say they have gone beyond it. But to the next generation of networking, in any case.

(That’s rather a mouthful, though, so for the rest of this post I’ll refer to their efforts as SDN just as the rest of the industry does; please forgive the shorthand. I’ll get back to the other current discussion about whether SDN 1.0 is dead in another post.)

Any good journalist knows that a good controversy attracts eyeballs, and the best controversy has two—and only two—diametrically opposed antagonists. More than two would complicate the storyline. So we got stories that suggested not only that VMware is now officially at odds with Cisco, but also that SDN architectural options are themselves both binary and comprehensively represented by Cisco and VMware.

I mentally protested as I read that “the battle to watch is between hardware-based data center strategies that focus on ASICs and proprietary systems, and the wholly software-defined paradigm VMware pushes,” as one writer put it.

After all, VMware knows as well as anyone that performance relies on hardware advances (thanks, Intel!). There are plenty of networking vendors who support OpenFlow and other “SDN” protocols on platforms that are neither ASIC-based nor “white box”. Many of these vendors are also VMware NSX partners, providing encap/decap and varying degrees of visibility to overlay traffic.


For these partners (and Cisco too), VMware’s overlay approach is one option among several that they offer their customers. They collectively represent a range of philosophies about how smart or dumb a network should be, and they have differing approaches to the role overlays might play in managing traffic. Some use ASICs in some products; others abhor the very idea. Some also offer their own controllers.

In other words, there are <gasp> a multitude of decision points across a range of parameters, which is as it should be in a very immature and fast-evolving market. It’s also not exactly clear what proportion non-REN organizations with production SDNs are using products from any of these companies, making the presumptive close on the two architectures/product lines with the most recent birthdates all the more preposterous.

How, then, to dispose of this mythical dichotomy? It occurred to me that another time there was a major tussle for the souls and resources of the great unwashed—popularly known as the Cold War—things were actually a bit less neatly divided than was and is often portrayed, especially in English.

I hereby propose that for the sake of the vast majority of the IT world currently trying to figure out if or when SDN might realistically enter into their daily existence, and only secondarily what form(s) it might take over time—networking needs a vigorous champion to promote a modern form of SDN tiers-modernisme.

A brief history of what again?

Tiers-mondisme is French for “Third-Worldism”, a catchall term that arose in the Cold War era to describe everything from national liberation movements in colonial Africa and Asia to efforts at organizing a relatively united front to represent “third world” interests in international organizations such the UN. Although “Third-Worldism” is often portrayed—especially in English-language writings—as being a leftist philosophy due to being associated with socialist-leaning liberation fronts, in many cases it was a declaration of independence from both the US and USSR spheres of influence. The Non-Aligned Movement was one manifestation of this strand of Third-Worldism.

In a very elegant sleight-of-hand, two formerly imperial powers—France and China—both co-opted the Third-Worldist mantra to describe their own foreign policies. France under Pres. de Gaulle would be the benevolent paternal figure of world capitalism; Maoist China offered itself as the more relevant socialist model for smaller agrarian economies. Material aid as well as political support via France and China’s major roles in the UN were naturally part of the benefits to be enjoyed by the “non-aligned” ecosystem each country hoped to construct for itself.

Due to this history, tiers-mondisme sometimes connotes a tighter organizational framework than the English term “Third-Worldism”.

Unfortunately, the traction of these efforts was undermined by other small matters taking place in the same timeframe, such as the very ugly Algerian war for independence from France and the fact that China was busy lurching from the famine years of the “Great Leap Forward” to the chaos of the Cultural Revolution. (Nobody’s perfect.) And then the end of the Cold War made it all rather irrelevant.

And now, back to our story

The national-liberation style of opposition to network imperialists, as embodied by the array of SDN startups funded in the last two years, has provided a good foundation for getting the SDN movement underway. It’s hard to ignore that much VC funding going to an alternative concept. But in order to keep things going, I do think that a more established, but not hegemonic, organization will need to provide a strong voice speaking out consistently for a Third Way. The startups are necessarily focused on survival and on differentiating their particular approaches vs all the others. A more established entity to some degree has the luxury to take a broader stance that may pragmatically support the gamut of SDNish options, while providing market education on the pros and cons of each.

This entity could certainly be a networking vendor. It could be an NGO-equivalent with the backing of significant corporate sponsors with sufficiently diverse interests to ensure a rough sort of honor, so long as that group is able to attract and be relevant to a solid group of users. A vendor working hand-in-glove with such a group might be the most interesting possibility; the individual vendor would have a clear interest in focusing its efforts on offerings with the strongest market relevance, while necessarily staying connected to the broader interests of the community vs getting excessively religious about a single approach.

So, who might be our latter-day De Gaulle? The role calls for outspokenness, a certain pugnaciousness, a bit of an outsider mentality, and comfort with high-flown rhetoric and realpolitik in equal measures. A potentially hubristic desire to leave his or her mark by providing the world with a clear alternative to the obvious paths would certainly help.


Maybe not quite like this, though

I think it’s a shame Scott McNealy is retired. Who are our other candidates?

2 comments on “SDN Needs a Charles De Gaulle

  1. Pingback: SDN 2014 – Make Our Garden Grow – Part 1

  2. Pingback: SDN Needs a Charles De Gaulle

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s