Is BGP Good Enough with Dinesh Dutt on Software Gone Wild

In recent Software Gone Wild episodes we explored emerging routing protocols trying to address the specific needs of highly-meshed data center fabrics – RIFT and OpenFabric. In Episode 92 with Dinesh Dutt we decided to revisit the basics trying to answer a seemingly simple question: do we really need new routing protocols?

We quickly agreed that the question is moot for data centers with few hundred switches (or less), and that engineers building larger data centers hopefully know enough that they don’t need our advice and use scale-out architectures anyway… but of course we couldn’t stop there and I hope you’ll enjoy the rest of our discussion where we desperately tried to figure out what problem we’re trying to solve with new routing protocols.

Latest blog posts in BGP in Data Center Fabrics series


  1. Thanks Ivan!

    It would be really slick if you could interview Randy Bush on similar topics:
  2. Another great podcast! Maybe relevant, but the podcast link below explains a bit about how and why a company chose the path of building a leaf/spine network with pods using Cumulus NOS, VXLAN/BGP/EVPN with puppet/ansible automation mixed in. I found this podcast to be a "poster child" of how and why today's modern networks should be build - assuming you have the right resources (most importantly people, policy and process/procedures - this cannot be done if any of the three are missing).

    Podcasts like these are worth their salt in gold :)

    Thanks Ivan and Dinesh!
  3. I had problems with understanding Dinesh's indian accent but I could catch that BGP and the existing link state protocols are good enough for every leaf/spine/super-spine topology. The real reason for a new routing protocol is as always a commercial one. Now I’m wondering how the advocates react after Dinesh peed on their parade.
  4. One of the best podcasts on routing in DC in recent times.

    Evil CCIE
Add comment