Azure-Player / azure.datafactory.devops

Azure DevOps Deployment Tasks for Azure Data Factory objects
MIT License
39 stars 21 forks source link

Expose "IgnoreLackOfReferencedObject" option #72

Closed dbeswick-bupa closed 2 years ago

dbeswick-bupa commented 2 years ago

Hello,

Thanks for this nifty tool! This PR just exposes the existing "IgnoreLackOfReferencedObject" Publish Option.

We have scenarios where our Data Factories get quite large and have many different teams working on them, so we separate the Factory into "components." That means that a Factory might be made up of many Git repos, and pipelines might depend on pipelines in other repos. You might call this a "partial deployment" scenario.

In that case, it's not an error if a dependency isn't in the set of JSON files being deployed, as long as it does exist already in the Factory.

Deleting "missing" resources also doesn't make sense under current assumptions. In the future, we would also like to remove resources that were deleted from a component repo. I plan to do this by giving each "component" their own folder in the Factory, and so each folder can be considered a complete deployment set. This set can then be used to calculate which existing Factory resources should be deleted. Please let me know if you'd be interested in this work in future.

Regards, David