Hello, world. What is this MODEM podcast all about? Well, we saw a gap emerging in the technical podcast space. That gap was a deeply technical podcast, focused on the protocols, the details, the how-to-demistify-complex things, the unbiased, un-cola of networking that needed to be filled. We’re not interested in the marketing, or the fluff - we want to deconstruct the new shiny things and break them down into comsumable parts. Maybe also some witty banter.
Containerlab: Declarative Network Labbing with Roman Dodin
Building large-scale network labs can be tedious and error prone—More importantly, they can be notoriously hard to spin up automatically. Containerlab is a new tool that promises to "redefine the way you run networking labs" and I really think it hits that target. On this episode of the Modulate Demodulate podcast, Nick and Chris C. are joined by Roman Dodin, one of the brains behind Containerlab.
Find Containerlab Online
Networking Field Day 25
Nick's "Github 'O Configs"
Deep Dive into Thread with Jonathan Hui
Deep Dive into Thread with Jonathan Hui
Welcome back to the Modem podcast where we visit some of the dark and musty corners of the network and shine lights in pars of the network that many people haven’t touched before.
For many of us in main stream networking, we’re all about more. More bandwidth, more control. Bigger pipes. Faster switches etc… but there’s whole other spectrum that I’ve started to dive into over the last year that I wanted to bring to the table to discuss to prove that sometimes, more bandwidth isn’t the answer to everything, and that indeed, size matters, and sometimes, just sometimes, smaller is better.
Recently the smart home space has been abuzz when there suddenly appeared a new protocol on the scene that was all the new hotness called “Thread”.
Now like many protocols, it wasn’t really new at all, but was really just new to many of us who have never taken the time to dig into the interesting world of low-power low-data networking.
Luckily we have a special guest with us today to take by the hand and help us remember that everything in networking has it’s place. Even RIP.
In this episode, we welcome Jonathan Hui, CTO of the Thread Group to the show to take us down the rabbit whole into the history of this protocol, enlighten us to microverse of low-power/low-bandwidth networking, and help us remember that BGP isn’t the answer for everything.
Learn More about Thread
Thread Group Website
Thread Group Resources
Thread Networking Fundamentals
Connect on Social Media
Jonathan on Twitter
The Thread Group on Twitter
The Thread Group on LinkedIn
ARTEMIS: Making BGP Operations Suck Less
BGP is one of the most versatile routing protocols out there, but let's be honest—It kinda sucks when it comes to... Well, a lot of things. Specifically, there are a lot of issues with BGP's security (or lack thereof). BGP comes from a time when The Internet was much smaller and everyone knew everyone. Now that The Internet has exploded in growth, hijacks, both malicious and inadvertent, have required countless hours of pain-staking manual intervention and deep knowledge of BGP and the global routing table to diagnose—keeping your routes secure in this landscape almost seems like a sisyphean task.
Cue ARTEMIS. ARTEMIS (Automatic and Real-Time dEtection and MItigation System) does the hard work of detecting hijacks for you and it can even step in and attempt to mitigate them on your behalf. Okay, that sounds great, but how much does it cost? Nothing! ARTEMIS is free, open source, and even simple to install and use. Check out this episode of the Modulate Demodulate podcast as the lead developer for ARTEMIS and Co-founder/CTO of Code BGP, Vasileios Kotronis joins Chris C. and Nick to talk about the inner-workings of ARTEMIS.
Find ARTEMIS Online
ARTEMIS Slack Community
ARTEMIS Live Demo
Route Views Project
ARTEMIS IEEE Paper
ARTEMIS BGP Hijacking Survey with Network Operators Paper
Packet loss is good. Wait, what? Let's talk about Buffer Bloat with Dave Täht.
Buffer Bloat. Most folks in the networking industry have at least heard this term, and may have a vague idea of what it means. It's certain that all of us have experienced it at one time or another - and likely thought it was a different problem. Over the last few years a couple of queuing disciplines have emerged that have allowed the users of the internet to experience fewer and fewer of those odd symptoms, and we wanted to know more about how those problems are getting solved. Fortunately, we managed to pin down Dave Täht and get him to talk to us about fq_codl and cake. Or so we thought. Turns out, that's a really, really big topic. Luckily, we had the expert to take time and really get down to the root of the problem, how it is solved, and give us a fantastic bit of history about how it came to be. This one has it all, folks. Boats, guitars, stickers, Dave even plays us a song at the end. It's a fun one, and the deep knowledge does not disappoint. Join me, Chris Cummings, Dan Siemon, and Dave Täht as we wander through the complex forest of buffer bloat, queuing codecs, and queue theory.
Bufferbloat and Beyond Book
Netstat command for looking at fq_codl:
Linux tc -s qdisc show dev eth0 netstat -c fq_codel -vvv
netstat -c fq_codel -vvv
SFQ, DRR, SQM, other queuing disciplines
The Flow Queue CoDel Packet Scheduler and Active Queue Management Algorithm
Through the BGPStuff.net Looking Glass with Darren O'Connor
Tune in to this installment of Modulate Demodulate as Darren O'Connor joins Chris C., Nick, and Dave to discuss his side project—BGPStuff.net. This tool is a modern BGP looking glass built in Golang that anyone can use to gather a wealth of information on the BGP Routing table. Some of the things you can see are AS_Path, Origin AS, ROA, ASName, RPKI Invalids, DFZ RIB size, and more. In addition to a nice web interface, the latest version of BGPStuff introduces an updated RESTful API. Come check out our discussion of the architecture and technology behind BGPStuff.net!
Darren's VirtualNOG Project
BGPStuff Python Client
ASN Bogon Validation Python Library
[Non-Blocking] The DoD Brinks Truck of IPv4 Space
On this first episode of the MODEM Non-Blocking series, Nick and Chris C. wax ineloquent about the latest networking gossip straight from the DFZ. Recently, the DoD started announcing a lot of IPv4 address space that had been previously unannounced. How much is a lot? We're talking 13 /8s of IPv4. Tune in to hear our un-scripted thoughts on what this might mean for your network, The Internet as a whole, and the deployment of IPv6.
Hilco Blog Post
Team Cymru Bogons
NANOG Thread on DoD v4 Space Being Treated as Bogons
NANOG Thread on DoD v4 Space Being Announced
Cloudflare 188.8.131.52/8 Research
AMPRnet 44/8 Official Statement