metadata-ingestion/docs/sources/timescaledb/README.md
TimescaleDB is an open-source time-series database packaged as a PostgreSQL extension. See the official TimescaleDB documentation.
This connector extends the Postgres source with TimescaleDB-specific metadata: hypertables, continuous aggregates, compression and retention policies, and optionally TimescaleDB background jobs. Both self-hosted TimescaleDB and Tiger Cloud are supported.
| TimescaleDB Concept | DataHub Entity (Subtype) | Notes |
|---|---|---|
| Database | Container (Database) | |
| Schema | Container (Schema) | |
| Table | Dataset (Table) | |
| Hypertable | Dataset (Table) + hypertable tag | Dimensions, chunk count, compression flag, and retention policy are emitted as custom properties. |
| Continuous Aggregate | Dataset (View) + continuous_aggregate tag | Refresh policy and source hypertable are emitted as custom properties. Upstream lineage to the source hypertable is emitted. |
| View / Materialized View | Dataset (View) | |
| Stored Procedure | DataJob (Stored Procedure) | Grouped into a Procedures Container DataFlow per schema. TimescaleDB policy procedures are filtered out. |
| Background Job (optional) | DataJob (Background Job) | Grouped into a Background Jobs DataFlow per schema. Requires include_background_jobs: true. |
| Job execution | DataProcessInstance | Recent runs of a background job. |
| Column | SchemaField |