Observability with Prometheus and beyond

on YouTube ↗27m

" summary

Overview of a concrete stack: Prometheus for metrics with pull-based scraping, counters/gauges/histograms and PromQL for queries; single instances have handled 125 million active time series; Thanos/Cortex for long-term, horizontally scalable storage. Loki stores logs using the same label model and indexes only timestamp and labels to keep costs down; Promtail ships logs. Tempo stores 100% of traces (exemplar-driven access to samples). Also covers SLOs/error budgets and the rule: alert only on things impacting user-facing service.

AI-written summary · about

" chapters (4)

1. 0:00Introduction
2. 3:00Monitoring vs Observability
3. 11:28Prometheus
4. 17:47Loki

" related talks

• 36mObservability with Prometheus and Beyond: Leveraging Cloud-native Technologies Outside of CloudRichard Hartmann · Open Source Summit + Embedded Linux Conference + OSPOCon 2021
• 29mParanoid Observability –Success Stories and Misfortunes observing an observability stackErik Sommer · Open Source Monitoring Conference 2024
• 33mIntro: PrometheusGanesh Vernekar, Matt Layher · KubeCon + CloudNativeCon North America 2019
• 01h 19mIntro + Deep Dive: PrometheusRichard Hartmann, Julius Volz · KubeCon + CloudNativeCon Europe 2019
• 30mIntro: PrometheusRichard Hartmann, Ben "SuperQ" Kochie · KubeCon + CloudNativeCon China 2018
• 31mObservability: Dust Off Prometheus with VictoriaMetricsJulien Briault · Devoxx France 2024
• 10mUpstream-First, High Scale Prometheus EcosystemKemal Akkoyun · PromCon Online 2021
• 33mPrometheus Project IntroJulius Volz · KubeCon + CloudNativeCon Europe 2018