Watch It! Monitoring Lessons from Two Decades in Computing - Markus Krüger - NDC Oslo 2026

Value of the Information & Strength of the Argument

The presentation offers valuable, actionable insights rooted in hands-on experience. The argumentation is coherent and persuasive, building from fundamental concepts to practical implementation. The speaker supports claims with references to well-known industry resources and personal anecdotes, making the advice credible. He also addresses common pitfalls, such as high-cardinality labels and log noise, providing concrete guidance. However, the talk is more of an overview than a deep technical deep-dive, and some topics (e.g., AI monitoring) are covered superficially. The strength lies in the clarity of explanations and the breadth of coverage, which makes it useful for engineers at all levels.

Scientific Rigor, Source Quality, Title Accuracy

The speaker cites two authoritative books: ‘Accelerate’ and Google’s ‘Site Reliability Engineering’, both widely recognized in the field. He also references open-source tools and standards like Prometheus, Grafana, the ELK stack, and OpenTelemetry, which have well-documented communities. The title is a perfect fit for the content, as it accurately promises a wealth of monitoring experience. No external sources are directly linked in the video description beyond conference pages, but the oral references are credible. The talk is rigorous in its overall approach, though it does not include empirical validation or original research, which is expected for a conference talk.

The title accurately reflects the content: lessons on monitoring from two decades of experience.

8/10

The talk draws on two decades of practical experience and references established sources like the SRE book and Accelerate. It is well-structured and provides actionable advice, though it relies on anecdotal experience rather than empirical data.

Markers derived by PSI from the transcript: the creator did not define chapters.

• 02:00Introduction and speaker background at STIR
• 05:00Why monitoring is important and the concept of MTBF/MTTR
• 10:00Explanation of telemetry: metrics, logs, traces
• 15:00Four golden signals: latency, traffic, errors, saturation
• 20:00Metrics types and cardinality, introduction to Prometheus
• 25:00Log aggregation and structured logging, ELK stack
• 30:00Distributed tracing and OpenTelemetry, sampling strategies
• 40:00Dashboard design principles and Grafana
• 50:00Alerting, on-call practices, and AI in monitoring

• NDC Conferences — Organizer of the talk, provides information about the conference series.
• NDC Oslo — Specific conference where this talk was recorded.

• ndcoslo.com