Hi all! I am working on a DevOps project to automate the creation of UC objects through different environments (dev-test-prod). Each time we deploy our code to a different environment (using a Github workflow, not really relevant) we want to also create/update the UC object that goes with that (dev_<catalog_name>, test_<catalog_name>, <catalog_name>). During this automation we are going to give certain groups certain grants based on how much access we want them to have.
My initial plan is to copy the dev catalog into a new catalog with a different name (test_<catalog_name>) and then adjust permission after the copy (all via notebook). Is there an easy way to do this initial copy via databricks/azure CLI? I couldn't find anything online. My initial though is to run a notebook to recreate an empty test catalog if it doesn't exists. Then iterate through the dev catalog to fetch the schemas and creating empty schemas in the test catalog. Then iterate through the dev schemas to fetch the tables, view, and functions and recreate them in the test catalog.
Has anyone done anything like this? Is this the best approach?
I originally created this post to ask a specific question about the recreation of functions outlined above but got carried away with explaining the entire project. The function question is below:
I have a function (multiple at this location but describing as 1 for simplicity's sake) at source_catalog.schema1.udf_function I want to copy this function to a new catalog (dest_catalog.schema1.udf_functoin). The CREATE FUNCTION SQL command does not seem capable of this unless I missed something obvious in the documentation. I cannot find anything else online. What is the best way to copy/recreate a function from one UC location to another?
Sorry if this post is all over the place but if anyone has any advice, please let me know. It would be extremely helpful.
Thanks!
