Update: Kireeti Kompella Leaves Juniper for an SDN Startup.

What SDN Start Up did Kireeit Kompella Leave Juniper to Join?

Light Reading reports this morning that Kireeti Kompella has left Juniper to join Contrail Systems a new software-defined networking (SDN) startup.

While we don’t have independent confirmation of Light Reading’s report — if it’s true — a couple of observations that maybe of interest to the community.(Update:  9/10/12 10PM.  We have confirmation that Kireeti has left Juniper)  First, a little background, I worked with Kireeti at Juniper and have tremendous respect for him.  Not only is he a brilliant networking CTO — he is a genuine and high quality person — who happens to be a lot of fun to be around.  Second, I was recently reconnected with Kireeti when I moderated the Hot Interconnects Panel: SDN Fad or Phenom.  (You can register to download his at the bottom of this post).

Kireeti’s Juniper’s departure and Contrail’s expected hire has a number of SDN ecosystem implications and questions, including:

  • Sheds light on Contrail’s strategic direction.  We’ve heard rumors for sometime that Contrail is taking an MPLS-like approach to SDN and network virtualization.  Many will argue that Kireeti is one of the fathers of MPLS — which lends credibility to the Contrail product rumors.
  • May imply that Kireeti’s vision for SDN is not aligned with Juniper’s planed approach.  Read what is public of Kireeti’s SDN vision — which he outlined in part at the Hot Interconnects Panel — by downloading his slides (register below)
  • Highlights Juniper’s SDN troubles.  If true, Kireeti would be is Juniper’s second high profile SDN leadership departure (David Ward leaving Juniper to return to Cisco was the first).  This just reinforces the questions we hear from service provider and enterprise customers in the SDN Community including:  What is Juniper’s SDN Strategy?  What does it mean for qFabric?  Do they have the talent to execute a winning SDN Strategy?

While Contrail may have a fantastic opportunity, to me, Kireeti’s move says more about the state of affairs at Juniper (note: my view on should you join an SDN startup).  Which can be summed up by questions I recently fielded from two F100 CIO’s who are long time Juniper customers: a) what is Juniper’s relevance in SDN and network virtualization? and b) can you give me a reason to short-list Juniper for a trial or POC?

Register now to download slides

  • This field is for validation purposes and should be left unchanged.

Like our Content? Join the SDNCentral Mailing List

  • This field is for validation purposes and should be left unchanged.

About the Author

.

Matt has 20+ years of software-defined networking (SDN), cloud computing, SaaS, & computer networking… More

  1. manass732
    says:

    What amazes me is that Juniper seems to be in perennial denial – the latest piece on Network World on Juniper/Bob Muglia is even more perplexing. On one hand he says that they are working very hard on the controller:

    “We’re actually working kind of hard on it,” Muglia says of the open source SDN controller effort. “It’s sensible for Juniper because we want to be disruptive in this space.”

    and then later on in the article it states that:

    Still, Muglia could not say whether Juniper will brand its own SDN controller or source another.

    If they are hard at work why will they be unable to say if they will brand their own crud. It seems like they are hardly working.

  2. I agree — it definitely muddy the waters about what they are doing. If they are going open source — there’s really only semi-viable option (Floodlight), so why not just say it. I’m skeptical about an open source market for controllers — though that’s a discussion for another blog post.

    Regardless of what Juniper does — the lack of clarity will not win the hearts of the app developers which is what any open source model requires.

    • manass732
      says:

      I have another thing that has been bothering me – Floodlight is an open-source controller and Bigswitch has plans for proprietary controller that provides additional functionality. What is the compelling value proposition for the proprietary version from the customer point of view – Availability/Reliability, Scale or what?

      In that case, if the proprietary controller has significant customer benefit, then how does the open-source ecosystem help them and why would any significant networking vendor want to build their application(s) on the open-source version.

      My view is that the Controller Platform Software is the next evolution for competitive differentiation for a networking company – basically, the networking hardware elements are going to become simplified and as a result, the software running on those elements will not be the strong differentiator going forward. If this line of thinking has any merit, then the value of JUNOS for Juniper and IOS for Cisco is significantly diminished and the new battle ground is the Controller Platform Software.

      How does Juniper evolves in the open-source controller software model – it either has to derive all the value from the application/services software (security, load balancing, etc) or professional services/integration which is not their strong point either. This seems challenging and new ground for them (and others in this space).

  3. [...] credible source of software-defined networking (SDN) and network virtualization — by continuing our reporting of people movements in the SDN space — wanted to share a post this morning from VentureWire that Jason Matlof from Battery [...]

Leave a Reply

You must be logged in to post a comment.

Login


MODAL