https://linen.dev logo
Join Slack
Powered by
Is there a way to specify the `target` directory l...
# plugins-dbt
i
Is there a way to specify the 
target
directory location for dbt-snowflake to use for the manifest.json it creates? It's making it in .meltano/transformers/dbt/target/manifest.json right now. Should I be using that dir? Or make one in $MELTANO_PROJECT_ROOT/transform/target along with my dbt_project.yml, models, etc...
j
It's controlled by 
DBT_TARGET_PATH
I believe. I use the default location that it creates. I've been using dbt-snowflake/core 1.7.0 and I had to make sure I specify the 
DBT_CLEAN_PROJECT_FILES_ONLY
to be false
As an aside; If you'r ever working in docker and need to populate the json manually you can define manual commands in the meltano.yml file for the transformer
Copy code
commands:
      parse_manifest:
        args: parse --quiet
        executable: dbt_invoker
🙌 1
And then in my docker file i was running 
RUN meltano invoke dbt-snowflake:parse_manifest
i
I ended up specifying the target dir in the DBT_TARGET_PATH to transform/target since my dagster repository.py looks there for the manifest. That second part is so you can copy over the current manifest for your image to use, right?
j
yeah; for Dagster
it needed the manifest.json to run the dbt jobs, but i couldn't get the manifest to populate without a connection to the snowflake (like a chicken and egg thing)
🙌 1
that parsing command broke me out of the cycle
i
Ik you mentioned using windows earlier, did you also have an issue running jobs directly from your machine when using 
meltano invoke dagster:start
localhost? I ran into that yesterday and was wondering if that's something that resolves itself (like the snowflake batch files) when using a linux distro
j
i dont use meltano to run dagster per say
i
gotcha
8 Views
Open in Slack
PreviousNext
Powered by