Synchronize Metabase datasets from Metriql datasets. The idea is to leverage Metriql datasets in your Metabase workflow without any additional modeling in Metabase.
The library is available in PyPI so you can install it via pip as follows:
pip install metriql-metabase
The library expects stdin for the Metriql metadata and interacts with Metabase via its API. Here is an example:
curl http://metriql-server.com/api/v0/metadata | metriql-metabase --metriql-url http://metriql-server.com --metabase-username USERNAME --metabase-password PASSWORD --metabase-database METABASE_DATABASE_NAME sync-database
You can use --file argument instead of reading the metadata from stdin as an alternative.
Available commands are list-databases, sync-database.
Yes!
How is this related to dbt-metabase?
While this metriql-metabase is heavily influenced by the dbt-metabase codebase, it integrates Metabase with Metriql, not directly to dbt. While you need to maintain Metriql as a separate service, here are advantages of Metriql over dbt-metabase:
- You can define the metrics as native SQL
- You can leverage Aggregates to speed up your queries
- Sync the datasets into various data tools, not just Metabase
- Native MQL experience when running ad-hoc queries on data.