Task 1.1: Implement scalable Python project structures for Declarative Automation Bundles (formerly Databricks Asset Bundles / DABs) to support modular development and CI/CD integration.

• AAdd sync: include for the dist folder of built wheels

  Incorrect

  The artifacts build already uploads the built wheel on deploy; the file reaches the workspace, but with an unchanged version the serverless environment keeps using the package it already has.

• BSet dynamic_version: true on the whl artifact

  Correct

  Serverless jobs only pick up a changed custom package when its version changes; dynamic_version patches the wheel version from the file's timestamp on each deploy, so no manual version bump is needed.

• CRun databricks bundle deploy with the --force flag

  Incorrect

  --force overrides Git branch validation; re-uploading a wheel with the same version still does not make serverless install the new code.

• DSwitch the target to mode: development

  Incorrect

  Development mode changes names, schedules and concurrency for the target; it does not change how serverless decides to reinstall a package with the same version.

Serverless environments cache installed dependencies, and the serverless environment page notes that a changed custom package must also get a new version number for jobs to use it. The artifacts dynamic_version setting (or the artifacts_dynamic_version preset) makes the bundle patch the wheel version automatically on every deployment.

• AList them, or a glob such as resources/*.yml, in the include mapping

  Correct

  databricks.yml can reference other configuration files through the include mapping, so resource files can live in their own folder.

• BGive each resource file its own databricks.yml in a subfolder of resources/

  Incorrect

  A bundle has one and only one databricks.yml, at the root of the bundle project.

• CAdd the folder to the sync include list so that the files are uploaded

  Incorrect

  sync controls which files are uploaded to the workspace; it does not load configuration from them.

• DPass each file with a separate -t option to databricks bundle deploy

  Incorrect

  -t selects a target defined in the configuration; it does not add configuration files.

A bundle must contain exactly one databricks.yml at its root. That file can reference other configuration files, such as resource files, through the include mapping, which accepts paths and globs.

• AA notebook task that imports the resources package and calls that function after each deploy

  Incorrect

  Resources must be defined at deployment time by the CLI, not by a job that runs after deployment.

• BAn artifacts entry of type whl that builds the resources package on deploy

  Incorrect

  Artifacts build files to deploy; they do not make the CLI load resource definitions from Python.

• CA python section in databricks.yml naming the function, with databricks-bundles installed

  Correct

  The CLI calls the functions listed under python.resources during deployment, using the virtual environment that has the databricks-bundles package.

• DA rename of the Python files to .yml so that include can match them

  Incorrect

  Renaming the files does not turn Python code into valid YAML configuration.

Python support for bundles loads resources through functions listed under python.resources in databricks.yml. The specified virtual environment must contain the databricks-bundles package. Python and YAML resources can coexist.

• AKeep an artifacts entry of type whl with build set to the same build command

  Incorrect

  A build command runs locally before each deployment, which builds the wheel again instead of using the tested one.

• BOmit the artifacts mapping and reference ./dist/*.whl in the task's libraries

  Correct

  Without an artifacts mapping, the CLI assumes the wheel is already built and deploys the files named in the libraries whl entries.

• CSet dynamic_version to true so that the CLI skips the build step

  Incorrect

  dynamic_version does not skip a build; the artifacts entry still builds the wheel.

• DUpload the wheel to the DBFS root and point the task's libraries at the dbfs:/ path

  Incorrect

  Storing libraries in the DBFS root is deprecated and disabled by default on current runtimes.

If the wheel is already built, leave out the artifacts mapping. The CLI then deploys the wheel files listed in the libraries array's whl entries.

• Adatabricks bundle validate run in an empty project folder

  Incorrect

  validate checks an existing bundle configuration; it does not create one.

• Bdatabricks bundle generate job --existing-job-id for a job

  Incorrect

  generate writes configuration for an existing workspace resource; it does not scaffold a project.

• Cdatabricks bundle deploy with a new target name

  Incorrect

  deploy needs an existing bundle; it does not create project files.

• Ddatabricks bundle init with the default-python template

  Correct

  The default-python template generates databricks.yml, resource files, a src package with wheel build settings and sample unit tests.

databricks bundle init creates a bundle from a template. The default-python template lays out databricks.yml, resources, a src package and tests.