When deploying notebooks across workspaces via Fabric deployment pipelines, the attached environment does not adapt to the target workspace. I have to manually reconfigure each notebook after every deployment, which is error-prone.
My current setup is simple with 2 workspaces:
- Dev workspace → small capacity, notebooks run on the Starter Pool
- Prod workspace → large capacity, notebooks run on custom pools
After deployment from Dev to Prod, every notebook falls back to the workspace default, which isn't what I want in Prod - I need notebooks pointed at specific environment. I have to click into each notebook to repoint it.
Can we have deployment rules for this? Or is there something I can already do to avoid manual work?
1 Comment
- You can use the magic command to set the capacity for the notebook and you can use a workspace variable as the value you set for the capacity setting. For example you can use this: %%configure { "conf": { "spark.sql.shuffle.partitions": "200", "spark.executor.memory": "28g", "spark.driver.memory": "28g" } }
