Whether you’re overlay optimistic or overlay pessimistic about the way datacenter networks are going to evolve, there was a lot of proof in the virtual pudding at the recently concluded Open Networking User Group (ONUG) meeting in the Big Apple. The Overlay Working Group had issued their 2015 testplan challenge, and the results of their testing were published & discussed last week.
Admittedly and firmly among the Overlay optimistic, I see clearly now the leading indicators for a real shift toward open virtual networking. What happens when new things get real in networking? What happens when the “cool” gets beyond the lab and into the world? The challenges of operationalization quickly kick in, and the realities of scale conspire to push the cool back into the test environment. You simply can’t manage what you can’t see, and you can’t scale what you can’t manage. That’s often when the cool goes back on ice.
As SDN becomes real and pushes beyond test/dev environments and into production networks that span continents, and does so for financial institutions and hyper-scale enterprises for whom IT is a critical asset, these operational concerns have come front and center. The desire of agility with freedom of choice, the quest for “virtualized while open”, means that visibility and correlation between programmable SDN overlays and the array of network equipment (underlays) that form the physical network fabric is paramount. Then and only then can networks scale openly and run efficiently within and across seamlessly federated clouds.
This time, experience is on our side. The user communities at the leading edge of this change realize this well. How do we know? Simple. They told us. In a poll taken to prioritize the Top-10 overlay networking requirements, the three seen as most critical were:
– Monitoring (end-to-end)
– Correlation of Overlays & Underlays
– Control Plane scale (100K endpoints)
In fact, these top-3 got more votes than the other seven in the top-10 combined!
Good thing we let the cat out of the bag on our Virtualized Services Assurance Platform (VSAP) at the conference. In fact as Dimitri (@dstiliadis) and the Nuage virtuosos prepared to bring up 100K containers in less than ten minutes, the VSAP assurance engine was running alongside to deliver real-time correlation, monitoring & troubleshooting analytics. Watch the Tech Field Day crew as they discuss the topic & take a test-drive.
But that was last week… This week the gang is in Vancouver at the Openstack Summit, welcoming new partners including Arista Networks, Mirantis, and Avi Networks to our ecosystem and sharing reference architectures for Openstack deployment.
More reasons to remain overlay optimistic.