[Bug] Changing Incremental Model Partitions on Bigquery

[jennlw]

Is this a new bug in dbt-bigquery?

• I believe this is a new bug in dbt-bigquery
• I have searched the existing issues, and I could not find an existing issue for this bug

Current Behavior

Recently ran into an issue with Bigquery and incremental model partitions.

When the partition of an existing incremental model needs to be changed, the dbt full refresh command runs a create or replace statement against the model, which fails with the error:

Incompatible table partitioning specification. Expected partitioning specification interval(type:,field:<field_1>), but input partitioning specification is interval(type:,field:<field_2>)

This indicates that despite the replace statement, Bigquery expects the same partitioning specification as the original table.
We found that running a drop statement followed by a create statement succeeds with no errors.

Ideally dbt would be able to generate these statements on full refresh.

Please let us know your thoughts!

Expected Behavior

Full refresh should be able to generate a drop/create on full refresh.

Steps To Reproduce

Have incremental model partitions
When the partition of this model needs to be changed, get error when dbt full refresh command runs create/replace against the model.

Relevant log output

This is reported by the customer. I will ask them to subscribe to this Bug so that they can add comments if additional questions/input are needed.

Environment

- OS:
- Python:
- dbt-core:
- dbt-bigquery:

Additional Context

No response