

It is optimized for time series with high churn rate.It uses 10x less RAM than InfluxDB and up to 7x less RAM than Prometheus, Thanos or Cortex when dealing with millions of unique time series (aka high cardinality).It outperforms InfluxDB and TimescaleDB by up to 20x. It provides high performance and good vertical and horizontal scalability for both data ingestion and data querying.Later this data may be queried via a single query. Multiple Prometheus instances or any other data sources may ingest data into VictoriaMetrics. It implements PromQL-based query language - MetricsQL, which provides improved functionality on top of PromQL.Easy and fast backups from instant snapshots to S3 or GCS can be done with vmbackup / vmrestore tools.All the data is stored in a single directory pointed by -storageDataPath command-line flag.All the configuration is done via explicit command-line flags with reasonable defaults.VictoriaMetrics consists of a single small executable without external dependencies.



The cluster version of VictoriaMetrics is available here. Just download the latest version of VictoriaMetrics and follow these instructions. VictoriaMetrics is available in binary releases, Docker images, Snap packages and source code. VictoriaMetrics is a fast, cost-effective and scalable monitoring solution and time series database.
