In databricks deployment .py files getting converted to notebooks

[amit_jbs]

A critical issue has arisen that is impacting our deployment planning for our client. We have encountered a challenge with our Azure CI/CD pipeline integration, specifically concerning the deployment of Python files (.py). Despite our best efforts, when deploying these files through the pipeline, they are being converted into normal notebooks. As a result, we are forced to manually recreate the .py files and copy-paste the code to run the notebooks.

We have extensively searched for a solution to this issue across various resources, but unfortunately, we have not found clear and comprehensive documentation that provides a step-by-step guide to address this specific situation.

If someone can provide guidance or a solution it would be greatly appreciated.

[-werners-]

What is your pipeline?  We propagate notebooks using Azure Devops Repos with PRs and merges. like that files do not get converted.

[Dazza]

Experiencing a similar issue that we are looking to resolve except the files are .sql. We have a process that has 1 orchestration notebook , calling multiple .sql files. These .sql files are being converted to regular databricks notebooks when deploying so we are having to promote them to production in a manual manner right now - which is not ideal. Any insight/support here on this same issue would be appreciated

[amarantevitor94]

Did you manage to find the solution? If so, could you share it? 

I have the same problem, and if I find the solution, I'll share it here.

[AGivenUser]

We had the same problem but you can deploy the python files as 'RAW' file types with Databricks CLI.
Ref on usage from ADO Pipelines: https://learn.microsoft.com/en-us/azure/databricks/dev-tools/ci-cd/auth-with-azure-devops

[AGivenUser]

Another option is Databricks Asset Bundles.