Hi folks,
I would like to ask for best practises concerning the topic of parametrizing queries in Databricks Asset Bundle deployments.This topic is relevant to differentiate between deployments on different environments as well as [dev]-deployments vs. production deployments.
A well working example are notebook tasks or DLT jobs: I can pass all sorts of parameters. I can not only parametrize the environment (dev/test/prod) but also database objects (e.g. tables) with an identifier for [dev]-deployments (to not interfere with the regular database objects).
The problem now arises with Mat Views. For those I must use a SQL warehouse, so I cant use the existing tricks. Neither can I use a (parametrizable) Databricks Query object, as those are not commitable into repos. I am currently using a sql_task on a flat file which contains a hard coded query (s. below).
Using the power of sed in the deployment process, I am indeed able to handle the environment. But [dev] deployments are not handled by that and I secretly hope for a better solution than using sed.
Is there?
Thanks and best regards
```