How hard could it be to debug a network issue where pod connections time out?
It could take weeks if you are (un)fortunate like Alex.
But Alex and his team didn't despair and found strength in adversity while learning several Kubernetes networking and kubespray lessons.
In this KubeFM episode, you'll follow their journey and learn:
How a simple connection refused led to debugging the kernel syscalls.
How MetalLB works and uses Dynamic Admission webhooks.
How Calico works and assigns a range of IP addresses to pods (and what you should watch out for).
How to use tcpdump
and strace
to debug network traffic.
And as a bonus, Alex shared his knowledge of onboarding engineers and how to perfect the process.
Spoiler alert: this episode goes into a great level of (networking) detail, but the solution turned out to be very simple.
Sponsor
This episode is sponsored by Learnk8s — become an expert in Kubernetes
More info
Find all the links and info for this episode here: https://ku.bz/P5Y-NrSW5
Interested in sponsoring an episode? Learn more.