Specifics of Data Analysis in Time Series Databases
summary
Compares VictoriaMetrics, Prometheus and ClickHouse with concrete examples: Grafana uses Prometheus API start/end/step so time filtering and resolution are automatic while ClickHouse needs a generated SQL blob via a columns macro. Shows electricity-meter data where calculating the rate/derivative reveals a spike that raw plots miss. Notes a real dashboard with 230 PromQL queries (about 120 selectors and 80 rate calls). Claims VictoriaMetrics uses a more memory‑efficient vector format, highlights counter reset detection, and calls out high‑cardinality index churn as unsolved.
AI-written summary · about
chapters (16)
- 0:00Intro
- 2:33What is time series data
- 5:35What kubernetes changed?
- 7:32Unexpected consequences
- 9:11What is special about time series data?
- 12:10TimeSeries specialized solution focuses on
- 13:55Meters and Counters
- 14:58CREATE TABLE
- 16:19ClickHouse datasource query
- 17:45Prometheus read API
- 22:35Speed (rate) of the metric change
- 25:10Counter reset detection - plotting rate
- 25:56Node Exporter: most popular Grafana dashboard
- 27:12Time Series analysis specifics
- 27:42Looking for more? Check VictoriaMetrics!
- 28:20Questions?
related talks
31mHow ClickHouse® Inspired Us to Build a High Performance Time Series Database
11h 43mPercona Community Live Conference - Day 1 - Open Source Databases Conference
50mHow To Improve PostgreSQL Monitoring and Observability With PMM
36mVictoriaMetrics: scaling to 100 million metrics per second (EN)
01h 00mHow do We Keep Metrics for a Long Time in VictoriaMetrics
40mScale Your Monitoring Solution With the VictoriaMetrics Ecosystem
46mPrometheus as a Time Series Database
49mThinking in time series, the new solution to your pro...