Martin Heinz

It'd be great to see Kubernetes make more extensive use of croups & especially nested croups, imo. The cpuset affinity should build into that layer nicely, imo. More broadly, Kubernetes' desire to schedule everything itself, to fit the workloads intelligent to insure successful running, feels like an anti-partern when the kernel has a much more aggressive way to let you trade off and define priorities and bound resources; it sucks having the ultra lo-fi kube take. I want the kernels "let it fail" version where nested cgroups get to fight it out according to their allocations. Really enjoyed this amazing write up on how Kube does use cgroups. Seems like the QoS controls do give some top level cgroups, that pods then nest inside of. That's something. At least! https://martinheinz.dev/blog/91

"shell history is your best productivity tool" : https://martinheinz.dev/blog/110

So first off, your gif looks great! I like your sizing and resolution--that's one of the things I was worried about. Video recording is new to me, so I'd love to know what you record with and resolution settings to pick? I stumbled on this nice blog post on automating these recordings. Maybe a partial solution? https://martinheinz.dev/blog/94 I was thinking of doing a manual process like yours for a start rather than automating, but hearing your thoughts makes me think I'll try automating sooner. I see what you mean about the multi-app aspect. I'll have to switch to browser windows for mine as well. Your android approach is nice!

Also related: bash_history is your best productivity tool; https://martinheinz.dev/blog/110 I have a weird story about that. When I was a junior sysadmin I worked for a company providing a SaaS solution for e-commerce. The environment I worked in was bare-metal (as was the style at the time), with a 2 tier network architecture, lots of internal tooling (mostly in Perl fwiw) and highly complex interactions between components (cache items recorded in MySQL, centralised job control through a pair of activeMQ servers that had a postgres proof of record etc). During this time, I fancied myself the growing expert, and thus I created the most gnarly one-liners you could imagine. But they did everything. I annotated them with a “function parameter” comment at the end $ if []; then x; fi ## FUNCTIONDOES: THING then I would control+R constantly from the bastion host to do my daily work. I had managed to automate a huge amount of common requests this way. The existing team was disgruntled and started leaving, eventually making me the most senior in the team. Of course, I also left. I kept good contact with the team that onboarded as I was leaving, and one day while drinking I was told (off-handedly but in no uncertain terms) that I was the author of “the bash_history”. After enquiring to what they meant, they indicated that after I resigned, someone found my bash_history file (because I had been onboarding them and they had seen my workflow) and it had been immortalised…

Steady now ... The features of Per-Interpreter GIL are - for now - only available using C-API, so there's no direct interface for Python developers. Such interface is expected to come with PEP 554, which - if accepted - is supposed to land in Python 3.13, until then we will have to hack our way to the sub-interpreter implementation. https://martinheinz.dev/blog/97

Cross-Origin Request Blocked: The Same Origin Policy disallows reading the remote resource at https://martinheinz.dev:8080/api/v1/posts/52. (Reason: CORS request did not succeed).

Why should my browser connect to a nonstandard port just to load a webpage? https://martinheinz.dev:1234/api/v1/posts/24

This is good, but why not just post the original URL? https://martinheinz.dev/blog/22

I've attended a local CNCF meetup where the chainguard folks presented Wolfi and their related tools to create container images and SBOMs. I was already skeptical of the product, having heard of it before. Unfortunately, attending that talk confirmed to me that they've just re-invented a severely limited version of Nix powered by yaml files that can output SBOMs in a standard format. Their software repository only has around 500 packages vs the 80k+ in nixpkgs. When I read "Alpine Linux", I was immediately reminded of its use of the musl libc which has endless DNS resolution issues[1][2] in and outside of Kubernetes, however, thankfully, it seems like they realized bundling musl was a bad idea and ship glibc instead. I'll stick to Nix and nixpkgs to build my reproducible container images with full dependency graphs back to the source code. 1: https://martinheinz.dev/blog/92 2: https://news.ycombinator.com/item?id=35058094