Probably, I am getting old and sentimental, but again I couldn’t to resist the fundamental pages.
I had to take a closer look at FIB scaling issues and came across some papers.
A Tree-Based Packet_Routing Table for Berkeley Unix
Constructing Optimal IP Routing Tables
There was an interesting challenge these days. I had to emulate BGP FlowSpec RFC 5575 peering and some announces to the DUT.
Unfortunelly, my testing and assurance software version doesn’t support this ability, but I have found out this suitable cozy project. Link below, thanks to developers.
How Claude Shannon Helped Kick-start Machine Learning The “father of information theory” also paved the way for AI.
I'd pin up the Scriptures A Mathematical Theory of Communication
B-Jaxx is one of my favorites, no doubts.
And another one CLOS design.
This stuff is not so modern, but quite interesting and illustrates some useful design approaches.
Evaluation of an IP Fabric network architecture for CERN’s data center
Unfortunately, I often note that there is a huge gap in understanding of network performance testing methodology.
Almost every related case goes down to “It seems it was quite better earlier”. No metrics, no anamnesis, so it makes me sad each time.
Just to be more confident I’ve gone through these canonical documents. Will keep it in mind for reference.
ITU-T Y.1564: Ethernet service activation test methodology
RFC:2544: Benchmarking Methodology for Network Interconnect Devices
Finally, I’ve got a chance to take a dive into Segment Routing concepts, operations, and other aspects.
I was labbing with the migration scenario which is described in “DAY ONE: MIGRATING TO SEGMENT ROUTING”.
Have got a lot of fun, recommended.
I’m going to setup SRv6 topology as well, once I get the time.
Juniper DAY ONE: MIGRATING TO SEGMENT ROUTING
Despite the remote and punishing environment, we already have a plethora of potential eyes on the bottom of the sea. There are around 400 fiber optic cables spanning the oceans.
These cables are already the backbone of the global Internet - what if they could be the backbone of deep-sea monitoring and research efforts as well?
Quote: But RRAM does have a drawback. Like Flash memory, it wears out after being overwritten too many times.
I’ve recently taken a closer look at RIFT(Routing in Fat Trees) protocol.
Some details below.