ONS Accelerate Workshop: Amazingly Refreshing

Sometimes the stars do align: Open Networking Summit organized their Service Provider Accelerate Workshop just a day prior to Network Field Day, so I had the fantastic opportunity to attend both.

I didn’t know what to expect from an event full of SDN/NFV thought leaders, and was extremely pleasantly surprised by the amount of realistic down-to-earth information I got.

As always, there was the usual group of people reinventing old wheels, and the occasional vendor pitches, but there was also occasional pure gold, including:

  • Amin Vahdat from Google telling us how the server/OS industry still doesn’t have a large-scale cluster OS management solution 30 years after they started talking about replacing mainframes with workstation clusters – a refreshing reminder amid the usual yammering about networking being stuck in the past. Some hard-to-do things take time;
  • Al Blackburn from AT&T explaining how they started their Domain 2.0 project by focusing on transformation on their internal habits, processes and culture, and major retraining efforts – obviously at least some people agree with my “the technology won’t save you” mantra;
  • Yukio Ito from NTT describing the OpenFlow-related problems they encountered in their deployment (not surprisingly, mostly aligned with what I was telling you in the OpenFlow Deep Dive webinar, although they did discover a few additional glitches);
  • Damascene Joachimpillai from Verizon explaining their reservations about scalability of OpenFlow and the viability of service chaining favored by most NFV proponents;
  • Albert Greenberg from Microsoft Azure talking about their implementation of Network-State Management Service;
  • Multiple speakers focusing on the lack of NFV orchestration, and lack of integration with OSS/BSS systems;

Was it worth attending the workshop? Absolutely. ONS team did a fantastic job. Thank you!!!


  1. Amin Vahdat may have said that there is no "large-scale cluster OS management solution", but the vast majority of workloads in enterprise datacenters is managed by vCenter. vCenter is not the equivalent of Google's Borg, but enterprises don't run Google-style applications in their datacenters. vCenter does what enterprises need to do.

    Networking could really benefit from even having comparable capabilities to vCenter (with all of its limitations and issues) ... vCenter is light-years ahead of anything that is available commercially or in open source in networking.
Add comment