With the advent of 100G, a wide range of different transmission standards and form factors were published by IEEE and MSA groups. This produced a plethora of transceivers and created a difficult choice for the best solution for networks and datacentres, for todays and also future needs.
For 400G and 800G, IEEE and multiple MSA groups have published more than 30 transmission standards, with...
One of the biggest costs in any access network is the need to adapt lower access data rates into higher data rate backbone transport. Historically this has been achieved using either OTN switching in legacy systems or packet switching for modern applications. All-optical communication would obviously remove the need for expensive Optical-Electronic-Optical interconnects, but until now it has...
There has been no good Python library to manage the network elements using the gNMI. To overcome this limitation, we’ve decided to create our own. Being non-professional software developers, it is also an interesting learning curve.
As such, the talk will share the status of the pygnmi library development, examples of usage and integration with existing automation frameworks and story of...
Documentation is an important part of the life of a network engineer, whether we like it or not.
Unfortunately, not all documentation is created equal, and there's well documented cases out there where bad documentation does more harm than no documentation.
In this short talk, I'll talk about the deployment of 1.1.1.1, and how that relates to documentation.
IPv6 turned 25 in December. While global adoption of IPv6 might not be where its architects thought, IPv6 has come of age and inserted itself firmly into the Internet. This talk takes a look at how far has come.
In this presentation we will take a look at a new and novel routing protocol for the datacenter underlay; Routing in Fat Trees. Or in short, RIFT.
RIFT solves some of the challenges current datacenter underlays are facing; large IGP domains, zero touch provisioning, 5 or even 7 stage specific fabrics challenges, slow BGP convergence, etc.
An IP Flexible Algorithm allows an IGP to compute constraint-based paths. Therefore, it allows an IP network to achieve some TE capabilities without deploying MPLS or Segment Routing.
