Urgent.News

What's breaking now, across thousands of outlets.

AI

Linnix Linux observability: eBPF and AI for incidents

A Linux node slows down, top shows 80% CPU, and nobody knows which process is getting in the way of the others. That is where Linnix starts, and it is why Linnix Linux observability deserves a look: the project tries to find who causes the stalls, not just who consumes the most. It relies on eBPF, on the kernel’s pressure counters (PSI) and, optionally, on a language model that explains…

Linnix Linux observability leverages eBPF technology and AI to diagnose performance issues in Linux environments by identifying which processes cause stalls rather than simply measuring CPU usage. The project utilizes eBPF to monitor specific kernel events such as fork, exec, exit, and scheduler activities. It also incorporates PSI (Pressure Scheduling Information) counters to measure real waiting times instead of just activity levels.

An optional layer utilizes a language model for explaining incidents, although this component is disabled by default.

Key decisions within the project include selecting which signals to trust when a pod slows down, what level of automation to accept, the choice of license, and monitoring the agent itself. The attribution feature identifies noisy neighbors, fork storms, PSI saturation, and memory leaks, with detection metrics expressed as Prometheus metrics. PromQL queries are provided to rank offenders against victims based on stall times.

Linnix offers a configurable AI analysis layer via an OpenAI-compatible server with a default configuration that can be toggled off. The model's confidence is presented as an estimate rather than a calibrated figure, emphasizing the importance of verifying findings through manual investigation. The software operates under the AGPL license, which has specific implications for distribution and modification, and a commercial license is available for organizations with licensing restrictions.

The project includes mechanisms for safe deployment and monitoring, with guidance on installing it in Kubernetes or locally via Docker.

Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at dev.to →

More in AI

Let two AI agents on different machines exchange text - no ACQ package or local server to install

I'm Instinct, acting on behalf of Talchaim Damri, who built ACQ, the service this article walks through. The copy-paste relay When two coding agents on different machines need to exchange text - say…

  • Two AI agents exchange text via HTTPS POST requests on separate machines.
  • ACQ service by Talchaim Damri powers the relay without ACQ package or local server.
  • Rooms self-destruct after 2 hours inactivity or 24 hours existence.

AI for Logistics: Why 81% of AI Projects Fail

AI tech promises huge returns for logistics, but many companies are seeing disappointing results. Lean Solutions Group CEO Jack Freker reveals a critical flaw: 81% of firms neglect the organizational transformation needed for AI to truly deliver.

More from Wednesday 18 November →