Superset use multiple databases to build a single chart

Question

I am new to using superset, and faced with very difficult issue. Is there is some way to use multiple databases to build a single chart in Superset?

What I need to do is merge a results of 2 queries from different databases. Where the result of second query depends on the result of previous. More exactly, I need to post-process data received from both databases, and I'm trying to found any straight solution of this.

If there is no straight way to do so, how much it will be necessary to redefine the superset logic? How much of this goes beyond how superset works?

Answer 1

in my experience there is no "easy" way... but there is a known path...

as far i know... charts in superset does not make preprocessing, so, the preprocess is job for a separated module o query...

i use modules created in superset (flask) to reach that goal... but in simpler cases you can work with triggrers or materialized views (if you use postgresql).

even you can include simple scripts at the end of superset_config.py, but is not the most recomendable way to make customizations.

most cases depend on your developer experience, in the end the chart uses only a single table or view to consult data...

Some useful links

• https://www.digitalocean.com/community/tutorials/how-to-structure-large-flask-applications

• https://flask.palletsprojects.com/en/2.0.x/blueprints/

• https://www.w3schools.com/sql/sql_join.asp

• https://learnsql.com/blog/introduction-using-aggregate-functions-joins/