You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hey, external tables column changes aren't considered as a change when we run:
dbt ls --select "state:modified"
I believe that this happens because it's a source definition.
Is it possible to include column changes to state:modified for external tables?
Additional context
This is very similar to this issue. The difference is that columns' definition for regular sources can't really break downstream models, unlike columns' definition for external tables' columns.
Who will this benefit?
Anyone who wants to run CI tests, build downstream models each time external tables' columns were changed, and catch a wrong reference problem before production run.
The text was updated successfully, but these errors were encountered:
Poor man's solution
I ran into the same problem. As long as this feature is waiting to be added to DBT, you can use a python script and the module dictdiffer to detect changes in sources. The output of this script can be used for DBT's select arg. This script will also mark a source as modified if only the description changed. While that's fine for me, you may want to design something more elaborate.
You have to parse your current model before running this script.
Describe the feature
Hey, external tables column changes aren't considered as a change when we run:
dbt ls --select "state:modified"
I believe that this happens because it's a source definition.
Is it possible to include column changes to
state:modified
for external tables?Additional context
This is very similar to this issue. The difference is that columns' definition for regular sources can't really break downstream models, unlike columns' definition for external tables' columns.
Who will this benefit?
Anyone who wants to run CI tests, build downstream models each time external tables' columns were changed, and catch a wrong reference problem before production run.
The text was updated successfully, but these errors were encountered: