Attune: in-place pod right-sizing operator
https://github.com/attune-io/attune
Attune is a Kubernetes operator that automatically right-sizes pod resource requests and limits using In-Place Pod Resize.
https://github.com/attune-io/attune
Jarvis: web UI for Prometheus Alertmanager
https://github.com/kj187/jarvis
Jarvis is an open source web frontend for Prometheus Alertmanager — interactive, realtime, and self-hosted.
https://github.com/kj187/jarvis
oxynote
https://github.com/oxynote/oxynote
A self-hosted workspace that puts a live metric chart next to the paragraph explaining it, gives your internal APIs Stripe-style docs, and lets a change go through a draft, a diff and an approval.
https://github.com/oxynote/oxynote
GoNIX — a UNIX with a Go userspace
https://gonix.dev
A UNIX-like operating system on the Linux kernel, with the userspace — init, coreutils, networking, services and SSH — written mostly in Go.
https://gonix.dev
Workload-Aware Scheduling in Kubernetes 1.37
https://helayoty.substack.com/p/workload-aware-scheduling-in-kubernetes-284
Kubernetes 1.37 closes that gap with three things: a set of reusable building block APIs, a shared workloadbuilder Go library, and an updated, explicit Job integration built on both.
https://helayoty.substack.com/p/workload-aware-scheduling-in-kubernetes-284
automaxprocs now costs your Go binary a CPU
https://medium.com/@omlahore47/automaxprocs-now-costs-your-go-binary-a-cpu-61d63a12e3ec
Since the past eight years, if you ever found that your Go service was getting throttled inside Kubernetes, the way to fix that was using one import. And with the release of Go 1.25, that same job was absorbed into the runtime itself. But that import is still sitting in your go.mod file, and it is now taking a CPU away from you.
https://medium.com/@omlahore47/automaxprocs-now-costs-your-go-binary-a-cpu-61d63a12e3ec
Your gRPC Service Only Talks to One Pod. Here’s Why.
https://medium.com/@sundaradhi97/your-grpc-service-only-talks-to-one-pod-heres-why-da54ad78aa90
You scaled your gRPC service to five replicas. One of them is on fire. The other four are asleep. This isn’t a bug — it’s a direct consequence of how Kubernetes moves traffic between pods, and once you see the mechanism, the fix is obvious.
https://medium.com/@sundaradhi97/your-grpc-service-only-talks-to-one-pod-heres-why-da54ad78aa90