[nullindividual]

This is a significantly better technical presentation on how traceroute works[0]; for example, unlike the illustrations in the linked article, traceroute does not necessarily take a symmetrical return path; the return path is hidden from the client -- the client only sees the forward path.

[0] https://archive.nanog.org/sites/default/files/traceroute-201...

[wang_li]

Traceroute doesn't even show you a path. It shows you a bunch of devices that happened to have a packet when its TTL expired. Every item listed in traceroute's output is a different packet and can take a different path towards the destination.

On a different subject, why are people writing blogs about topics that are in the "literature" already?

[FujiApple]

It’s not guaranteed to be accurate, but tracing using the UDP/dublin strategy with a fixed dest port and varying src port per round can help to identify and visualize valid ECMP flows. I recently wrote some guidance [1] on using Trippy in this way.

[1] https://github.com/fujiapple852/trippy?tab=readme-ov-file#ud...

[Hikikomori]

Once I used iperf3 with 100 different udp streams/srcports to troubleshoot an issue, a small % of connections had >90% packet loss and this caused the connection pool of this service to fill up until it had only failed connections waiting to time out or going extremely slowly. ISP told me it was a broken linecard in a router, so packets were being dropped/corrupted on the backplane between the linecards.

Traceroute and mtr didn't use enough ports to show the issue clearly.

[ta1243]

I used this to identify some intermittent loss in one direction from I think Kyrgyzstan to the UK, There was a claim that sometime flows were fine and sometimes they weren't. Problem was an ECMP in Pakistan or Iran which was using udp port numbers to balance the flow. Could consistently get 25% loss with certain numbers and not with others.

[rudasn]

Hey! What literature would you recommend for learning more about this stuff (routing, firewalls, etc)?

I'm learning as I'm going and my goal is to come up with a way to build secure/beyondcorp/zerotrust-style networks using only existing tools and established tech, unless absolutely necessary.

Tools I'm using include wireguard, iptables, nftables, nginx, ping, traceroute, and nmap.

Which articles, books, authors, or repos must I read and which concepts must I understand?

Trying to find real knowledge about this stuff online is a nightmare!

[cpach]

Related threads with recs:

Ask HN: Good book to learn modern networking? - https://news.ycombinator.com/item?id=38918418 - Jan 2024

Ask HN: Best Books for Modern Computer Networks - https://news.ycombinator.com/item?id=39501933 - Feb 2024

Ask HN: What is a good book to learn about the network stack? - https://news.ycombinator.com/item?id=18506651 - Nov 2018

[nullindividual]

In English, route is synonymous with 'path'.

Feel free to submit a PR if you disagree:

https://man.openbsd.org/traceroute

[wang_li]

This isn't an English language question. The technology in question doesn't guarantee a singular route or a path to a destination. Subsequent packets from an originating host to a destination host can follow different paths. Even during an activity such as a traceroute. A packet that reached its TTL at what is presented as hop 12 didn't necessarily traverse previous hops displayed by traceroute.

[nullindividual]

You'll learn pedantry gets tiring when you get older. How traceroute is understood and displayed to the end user constitutes a route. If you have an issue with how the developers of traceroute framed it, that's on you.

[lostemptations5]

OP was the pedant in this case, unfortunately (for you).

[kccqzy]

Nothing can be assumed to be take a symmetrical return path. Even if I send a normal packet to some server on the Internet, the server's reply may or may not come back using the same path. As a former Google employee, I remember vaguely that this is something Google occasionally does on purpose.