Do not use entry_point to get workspace_id, job_id, run_id, and other metadata. There is a ready, stable solution to do that.
More good/bad practices on:
https://www.sunnydata.ai/blog/databricks-multi-statement-transactions
https://databrickster.medium.com/just-because-you-can-do-it-in-databricks-doesnt-mean-you-should-my-...

My blog: https://databrickster.medium.com/