<@U088WCP3B4P> i run dbt run --select elementary. ...
# ask-elementary-ai
n
@kapa.ai i run dbt run --select elementary. elementary tables are created on my msql. since elementary cli can not use msql, what can i do with these tables on msql to create the dashboard
k
A message from kapa.ai
Based on the knowledge sources, it appears that Elementary Data currently doesn't support MySQL as a metadata database for the CLI functionality. According to the information provided, there are a few important limitations: 1. Elementary doesn't officially support MySQL for its metadata storage. This is evident from the open GitHub issue about MySQL support where a contributor confirmed "MySQL is currently not supported" [GitHub Issue #1651](https://github.com/elementary-data/elementary/issues/1651). 2. The Elementary CLI cannot connect to a different database than where your dbt models are running. As one contributor explained: "Elementary is a package in your dbt project and runs as a part of it, so it cannot have a different target" [Slack forum discussion](https://slack.com/archives/C02CTC89LAX/p1703853557268729). 3. When someone asked if they could configure Elementary to use a different database (like Redshift or BigQuery) for its tables while continuing to use another database for dbt transformations, the response indicated this isn't possible because "dbt itself does not support connecting to multiple warehouses in the same run" [GitHub Issue #1651](https://github.com/elementary-data/elementary/issues/1651). Unfortunately, the knowledge sources don't provide a direct solution for using Elementary tables created in MySQL to generate the dashboard. Your options appear limited since the Elementary CLI is designed to work with the tables it creates in supported databases. If you're determined to use Elementary with your MySQL setup, you might want to: 1. Follow the GitHub issue about MySQL support to see when/if support is added 2. Consider if it's feasible to migrate your dbt project to one of the officially supported databases like Snowflake, BigQuery, Redshift, or Databricks I apologize that there isn't a more direct solution available in the provided knowledge sources.