VictoriaMetrics
lastmod 2024-12-01created 2024-12-011 edits
A fast, cost-effective monitoring solution for time series data
Features
- Optimized for time series data, and offers a range of features even when old time series are continuously and rapidly replaced by new ones
- Long-term storage for Prometheus: an alternative to Prometheus and Graphite that can be used directly in Grafana (drop-in replacement)
- Powerful stream aggregation: can be used as an alternative to StatsD
- Suited for large-scale data: supports a variety of enterprise workloads such as APM, Kubernetes, IoT sensors, connected cars, industrial telemetry, and financial data
- Queries: supports both PromQL and the better-performing MetricsQL
- Easy setup: no dependencies, a small single binary, configurable with command-line flags. Defaults are fine-tuned. Backup & restore with instant snapshots
- Global query view: unified queries across multiple Prometheus instances or multiple data sources
- Supports many protocols:
- Prometheus exporter remote write API, exposition format
- InfluxDB line protocol (HTTP,TCP,UDP)
- Graphite protocol with tags
- OpenTSDB put messages, HTTP OpenTSDB /api/put requests
- JSON line format, arbitrary CSV
- Native binary format
- DataDog agent, DogStatsD, NewRelic agent, OpenTelemetry, etc.
Additional features in the enterprise version
- Anomaly Detection: automatically detects complex anomalies to simplify alerting rules
- Automated regular backup procedures
- Reduced storage costs with multiple retention periods
- Downsampling: performance optimization for old data
- Stable releases: long-term support (LTS) and custom feature development through technical support