Little experiment: Make percent for pyrra.dev responsive to show as many decimal places as possible. Trailing .000 are trimmed and decimal places are only rounded if not enough space is left.
Matthias Loibl
@metalmatze.de
Director of @PolarSignals.com Cloud. Building cool products! 🧊 Maintainer of Parca, Pyrra, @Prometheus.io, PromOperator and more. PromCon organizer. @cncf.io Ambassador.
Super fun to see all sorts of things with. Like, you can see a @prometheus.io rule evaluation followed by GC and other interesting things.
Polar Signals@polarsignals.com · last wk.🔥Flame charts are now GA for all CPU profiles on Polar Signals Cloud. X-axis = wall-clock time, so periodic GC, phase shifts, and lock contention show up as patterns instead of getting averaged out. www.polarsignals.com/blog/posts/2...
The Polar Signals profiler got a major upgrade. New dashboard, and everything is faster and smoother 🚀 We can’t wait for you to try it out! Here's what's new and how we did it: www.polarsignals.com/blog/posts/2...
Our new and faster profiler dashboard | Polar Signals
Learn more about our new and improved profiler dashboard.
polarsignals.com
Working at Polar Signals, I always come at things with a performance mindset. Running evals on our MCP turned out to be exactly that: performance engineering but for the model's context. Measuring the profiling skill's efficiency and effectiveness has been the real fun these past weeks. 😎
Polar Signals@polarsignals.com · last mo.Less context, better answers. 🔥 The new @polarsignals.com MCP makes smaller, more effective tool calls and iterates on what's actually worth querying. www.polarsignals.com/blog/posts/2...
Two of our favorite things on the blog today: 1) Experimenting with cutting-edge GPU-accelerated columnar storage. 🚀 2) Dogfooding our GPU Profiler. 📊 Let us know what you think! www.polarsignals.com/blog/posts/2...
My Journey Optimizing a CUDA Kernel with Polar Signals | Polar Signals
A storage engineer's descent into warp stalls, memory throttles, and aligned stores
polarsignals.com
Ever wondered how a profiler knows which JavaScript function is hot when V8 compiles your code three different ways? Brennan went deep on unwinding and symbolizing V8 stacks across Sparkplug, Maglev, and TurboFan. The detective work behind every flame graph. 🔥 www.polarsignals.com/blog/posts/2...
Profiling Internals: JavaScript on V8 | Polar Signals
How Profilers Work: Part 3
polarsignals.com
🚀 Launching today: GPU PC sampling in production. Instruction-level profiling telling you why and where warps stall. Whether it is memory latency, scoreboard waits, or barriers. 🧊 Previously NSight-only territory. Now running in prod with low overhead. www.polarsignals.com/blog/posts/2...
Open-Source Low-Overhead NVIDIA CUDA PC Sampling | Polar Signals
And how we made it happen
polarsignals.com
Forgejo has been building federation so I can use my instance to talk to your instance to work on software together. These things exist. Just no one knows. forgejo.org/2023-01-10-a...
Answering Forgejo federation questions
forgejo.org
Thor@thorfour.dev · 3mo agoWhy can't we have bluesky but for git repos? Stop relying on GitHub to build a reliable service but have a distributed but discoverable service.
We've been working on a major schema migration at @polarsignals.com that has unlocked some pretty wild improvements. Apart from a 50% (!!) reduction in storage size thanks to logical compression in vortex.dev, there has been a dramatic improvement in query performance.
OpenTelemetry ❤️ Prometheus! The #OTel and #Prometheus projects are strengthening interoperability. 🛠️ Building with both? Check out our client library comparison, plus updates on Prometheus compatibility and exporter spec stabilization. Links in 🧵! 🗳️ Running both? Take our survey! buff.ly/XhVm0Fg
Survey: 🔭OpenTelemetry ↔ 🔥Prometheus Interoperability
Privacy notice: The data collected in this survey will be anonymised and publicly shared in the GitHub repositories prometheus-community/ux-research and opentelemetry/sig-end-user under the CC0 1.0…
docs.google.com
Your Node.js profiles on @polarsignals.com just got personal!
Polar Signals@polarsignals.com · 4mo agoTired of Node.js flamecharts full of `a`, `m`, and `_0x3f2`? We just shipped source maps support for Node.js profiles in Polar Signals Cloud. Just add a plugin to your build pipeline, and your profiles show the real code. Know more: www.polarsignals.com/blog/posts/2...
The future! This will be a thing, for sure. Pay for more MCP calls.
"I'm really excited about access control for agents." We caught up with @pst.blue at #KubeCon EU. He's diving into how the cloud native ecosystem is shaping infrastructure for the agentic age. #CloudNative #Agents www.youtube.com/shorts/ETc8j...
Philipp Strube at KubeCon EU 2026 in Amsterdam
YouTube video by Polar Signals
youtube.com
99.5% is slightly better than 3.5 hours in 28 days if we’re talking uptime. Really not great.
Pier@pierbover.com · 4mo agowoah damrnelson.github.io/github-histo...
We caught up with @mauilion.dev at #KubeCon EU. He was fired up about @rejekts.io pulling 220 people when sponsorship dried up, and noticing AI hype tempering with more energy going into infra, like virtualization. #Kubernetes #CloudNative youtube.com/shorts/zhz0X...
Duffie Cooley at KubeCon EU 2026 in Amsterdam
YouTube video by Polar Signals
youtube.com
We caught up with @bwplotka.dev at KubeCon EU. He's excited about semantic conventions across OTel and Prometheus finally becoming practical, and AI tools changing how open source is built. #KubeCon #OpenTelemetry #Prometheus youtube.com/shorts/QF3Yk...
Bartek Plotka at KubeCon EU 2026 in Amsterdam
YouTube video by Polar Signals
youtube.com
We caught up with Felix earlier this week at KubeCon EU, and TODAY he's presenting on the OpenTelemetry profiling signal reaching alpha, together with Florian. 📍 15:15, Hall 8, Room F Profiling as a first-class signal in OTel is huge. Don't miss it!
The @Prometheus.io Contributor Fest is underway! Let's learn how to contribute and review PRs for Prometheus!
For the next hour and a half (until 16:30) I'll be at the Prometheus project booth! Come around to talk Prometheus, SLOs and more or just say hi. See you!
For the next hour and a half (until 16:30) I'll be at the @Prometheus.io project booth! Come around to talk Prometheus, SLOs and more or just say hi. See you!
For local development you don't need a dedicated MCP server for CockroachDB, Prometheus, or Jaeger. Just exec into the container and run commands directly. SQL in cockroach, wget the Prometheus HTTP API, same for Jaeger. 😏
🤖 Using the Polar Signals MCP just got even easier. Check out how to use it with your favorite AI agent. 🧊 www.polarsignals.com/blog/posts/2...
Introducing Oauth for AI Agents
Use your favorite AI agent securely with Polar Signals Profiling
polarsignals.com
Polar Signals at #FOSDEM 2026 day 2! Both talks in the Rust room. Profiling Rust applications with Parca Brennan at 12:30 @ UB2.252A (Rust) fosdem.org/2026/schedul... Frederic at 16:00 @ UB2.252A (Rust) Random seeds and state machines: An approach to DST in Rust fosdem.org/2026/schedul...
OCI image support coming to systemd. I'm getting strong rkt flashbacks. 🥲
Polar Signals at #FOSDEM 2026 day 1! Tommy at 12:30 @ H.1308 (eBPF) OOMProf: Profiling Go heap memory at OOM time fosdem.org/2026/schedul... Thor at 17:25 @ UB2.252A (Databases) From Disks to Distributed: Our Journey of Database Evolution in the Cloud fosdem.org/2026/schedul...
FOSDEM 2026 - OOMProf: profiling Go heap memory at OOM time
OOMProf is a Go library that installs a eBPF programs that listen to Linux kernel tracepoints involved in OOM killing and records a memory profile before your Go program is dead and gone. The memory…
fosdem.org