Hi what’s the best way to change meltano orchestrator to LocalExecutor rather than Sequential? I have a postgres instance connected to the airflow orchestrator already. I have tried changing it in airflow.cfg before running initdb but it kept changing it back to SequentialExecutor as well as having

AIRFLOW___CORE___EXECUTOR

env var

You can either directly set the

AIRFLOW__CORE__EXECUTOR

env var that Airflow is reading from, or run

meltano config airflow set core.executor LocalExecutor

to write

core.executor: LocalExecutor

in

meltano.yml

and have it show up in

meltano config airflow

, which is used to populate

airflow.cfg

!

oh weird, I set the env var

AIRFLOW__CORE__EXECUTOR

so many times and it never worked…

meltano config airflow set core.executor LocalExecutor

while this worked however

@keat Hmm, that's odd! When I run

AIRFLOW__CORE__EXECUTOR=foo meltano --log-level=debug invoke airflow --help

, I see it getting passed along and picked up by Airflow (which understandably fails because that executor doesn't exist), and it works the same way when I put

AIRFLOW__CORE__EXECUTOR=foo

in

.env

. Can you verify that this is not working for you locally? Are you on the latest version?

@monika_rajput, @connor_flynn, and @jo_pearson here is a good thread on setting up executor