And another question: what would be the best way to use DBT with custom DB connector like Athena? Clearly

dbt-bigquery

or

dbt-snowflake

do not apply, Would it be better to install obsolete

dbt

package with extra dependencies or do something along the lines of forking

dbt-bigquery

? How does it work?

I can speak a bit to the second question. A few versions ago, dbt stopped shipping the most common adapters (snowflake, bigquery, redshift, postgres) with their dbt package and so it became necessary to install them as you need them. That's why you'll see the

pip_url

for the legacy dbt plugin has all of those packages, for backwards compatibility. Given that they're independent packages, with different settings and different dependencies, it makes sense to keep them as separate plugins.

@andrey_tatarinov - I think you should be able to start with

meltano add transformer

using the

--custom

option. I got Meltano and dbt work with Athena a while back so I know it is possible.

Thanks!