OpenData Timeseries is a Prometheus-compatible time series database built on SlateDB, an object-store-native LSM tree. It supports PromQL, Prometheus scraping, Prometheus remote write, OTLP metrics ingest, and Grafana through the Prometheus data source.
Most Prometheus-compatible systems still rely on stateful storage nodes and the operational work that comes with them. Timeseries takes the object-store-native route instead: durable state lives in object storage, local disks are caches, and you can separate the writer from the reader without building around a replicated TSDB cluster.
For development, that means you can run a single server locally and work against the same API surface you would use in production.
Use the docs quickstart for the fastest path from install to querying real metrics:
That guide installs Timeseries locally, scrapes metrics, walks through the built-in query UI, and shows how to add an external target.
Run the server directly from source:
cargo run -p opendata-timeseries --features remote-write -- --port 9090To enable OTLP/HTTP metrics ingest (POST /v1/metrics), start the server with
the otel feature and configure OTEL label handling in prometheus.yaml:
cargo run -p opendata-timeseries --features "http-server,otel" -- --config ./prometheus.yaml --port 9090otel:
include_resource_attrs: true
include_scope_attrs: true- Full PromQL compatibility
- Full Prometheus API compatibility
- Retention policies and rollups
- Bucket compaction
- Recording rules
- Service discovery for targets
- OTLP export protocol ingestion
- Log-native ingestion
- Read-only clients