3 Comments

Summary:

Juniper, one of the big technology vendors in the crosshairs of the revolution occurring in the networking industry has finally decided to tell its story around the rise of software defined networking and OpenFlow. It’s going to have to walk a fine line.

ixia-junipertest-setup

Juniper, one of the big technology vendors in the crosshairs of the revolution occurring in the networking industry, has finally decided to tell its story around the rise of software-defined networking and OpenFlow. Basically, Juniper’s willing to offer everything to everyone. It understands that some of its business will get commoditized, but it still believes there are customers and markets willing to pay for specialized hardware.

It’s a pretty realpolitik view of the world, which is smart considering that Juniper in many ways saw the bottleneck that the network would become a few years back. To address some of the issues of scaling out virtualized networks, it built its QFabric product that allowed customers running an all-Juniper system to flatten out the network and address some of the scaling problems. But now that OpenFlow and even open source controllers such as Big Switch’s Floodlight are on the scene, Juniper is opening up as well.

It has been hinting at its “be all things to all people” plan for some time. Back in February, when Nicira launched, Michael Bushong at Juniper penned a blog post in which he wrote:

Network Virtualization shines a light on the deep, dark underbelly of network infrastructure today. For too many years, we’ve treated it as plumbing: a service that needs to be ultimately reliable, infinitely scalable, and as close to zero-cost as possible. But the more we’ve driven cost out of the system, the less attention we’ve paid to our customers needs – flexibility and dynamic control which is where the real value lies.

Sure, Juniper is spinning its software-defined network story to downplay the threat of commodifying its switches and routing gear, but Mike Marcellin, VP of Platform Systems Division marketing at Juniper, acknowledged that in some areas, such as the top-of-rack data center switches, this was already happening. So Juniper will open up SDKs and support OpenFlow, Floodlight and other protocols those customers might want, while still offering pricier and more high-margin gear for service providers inside their core and edge networks.

Meanwhile, it’s also spending a lot of time trying to placate its customers by understanding what they want and how they need see software-defined networks offering them value. If Juniper does this well, the shift here may be less about it’s looming obsolescence and more about how it turns from churning out mass-produced gear that tries to solve most customer needs to trying to creating boxes that can be customized through software to serve individual customers. Instead of being the Holiday Inn, it has to become the Ritz-Carlton. Software defined networking and the threat and opportunity it poses is a topic we’ll discuss more fully at our Structure 2012 conference next week.

You’re subscribed! If you like, you can update your settings

  1. Brent Salisbury Tuesday, June 12, 2012

    Amazes me there is still no commitments to a viable controller out of the industry perennials. I gotta think, even if this all crashes and burns it would probably pay for the 10 software guys to roll it from POCs. Good for the VCs I reckon. If a fraction of this adopts I would imagine switch hardware margins to shrink closer to x86 with the profit moving north of the controller. I didn’t say app store though, never ever ever ever ever ever ever.

  2. JNPR Insider Tuesday, June 12, 2012

    Juniper talks a good game around SDN, but I’d encourage you to inquire what’s behind these initiatives. They have lost the best SDN talent they had when investment in the programs was sliced late last year. Where did these people end up? Competitors. Each of which is more serious about SDN than Juniper.

    So when talking to them, ask about the resources actually working on SDN – it’s between little to zero. Juniper makes great proprietary network systems. Juniper’s foray into software has been a fiasco thus far and is degrading.

    1. Sorry but if you think that by losing “the best SDN talent” within Juniper has stopped the progression of SDN within I think you are a bit naive. Just as much as people leave from one company to another so it works the other way around.

      Companies of the likes, size and complexity of Juniper don’t move very fast to begin with and often spend a good amount of time on the sidelines listening, interacting and talking to customers to see where the market and its demands may be calling for before making and investment to move forward.

      SDN is a very young industry and most vendors are still up on their whiteboards talking and discussing where their their products align and integrate with “SDNs”. But, and not to rehash any previous threads on this topic, Juniper has already delivered on SDN. QFabric is SDN and if I assume (based on your name handler) that you are tracking Juniper then you would have read on the recent announcement where Juniper just released a mini version of its QFabric, so development on its SDN is moving just fine and right along.

      I actually happen to think and have seen Juniper other “SDN” development, including its OF integration and would tell you that not only do I think Juniper has one of the better stories around SDN but their plans on how to integrate OF and use their SDK is one of the better ones I have seen to date. There are some great ideas and work going on to include OF/SDN and QFabric and how they will co-mingle and come together. I would love to see Juniper integrate or spin off their “IC” (Infranet Controller) and make use of the user telemetry and metadata collected as part of an OF controller.

      -JG

Comments have been disabled for this post