I would like to know if it is possible to include a specific commit identifier when updating a repo in a workspace via the Databricks CLI.
Why? Currently we use the repos CLI to push updates to code throughout dev, test and prod (testing along the way). We use repos as we also need available non-notebook files available in the repo, and this set up allows our teams to work more independently knowing they can deploy the files they need contained within a single repo. When an update to say the prod branch is made, an ADO Release pipeline runs which executes on command line something like
databricks repos update --path "<path_to_repo_on_workspace>" --branch "prod"
Usually when deploying based on updates to branches (to say copy files to a network drive) we would easily be able to re-release a previous release if any issues were encountered with the latest know working release as each release artifact (new git branch commit) has a commit (or Build.SourceVersion variable in ADO release pipeline variables) which can be redeployed with a single click should it be required.
However, the databricks CLI command presumably has no exposure to the commit linked with a release, and therefore if an earlier release is redeployed it will simply update the repo on the workspace to the latest commit of the branch despite the release being linked to an earlier commit.
Does anyone know if there is a way to link the databricks repos update cli command with a specific commit (i.e. is it possible to link the commit identifier with the command somehow)?
