Task 5.2: Understand environment-specific configuration using Automation Bundle (formerly Databricks Asset Bundles) variables and overrides while promoting the same codebase across dev, test, and prod targets.

• Aabc

  Incorrect

  The top-level default is the last place the CLI looks; any other source of a value wins over it.

• Bdef

  Incorrect

  The target's variables mapping is checked after --var, BUNDLE_VAR_ environment variables and variable-overrides.json.

• Cghi

  Incorrect

  A BUNDLE_VAR_ environment variable ranks below a --var option given on the command line.

• Dxyz

  Correct

  The CLI checks --var options first, so the value passed on the command line is used.

The CLI resolves a variable in order: --var, then BUNDLE_VAR_ environment variables, then variable-overrides.json, then the target's variables mapping, then the default.

• Aspark_version 15.4.x-scala2.12, node_type_id i3.xlarge and num_workers 2

  Incorrect

  The target's value for num_workers overrides the top-level value for the same job_cluster_key.

• Bnum_workers 8 alone, with no spark_version or node_type_id

  Incorrect

  Target settings are joined with the top-level settings; unspecified fields keep their top-level values.

• Cspark_version 15.4.x-scala2.12, node_type_id i3.xlarge and num_workers 8

  Correct

  Settings for the same job_cluster_key are merged, and for conflicts the targets mapping takes precedence.

• DThe deployment fails because job_cluster_key main is defined twice

  Incorrect

  Defining the same key under a target is the intended way to override or join settings.

Target overrides join with top-level resource settings by key, such as job_cluster_key for job clusters or task_key for tasks, with target values winning conflicts.

• AA lookup by name: lookup: warehouse: 'Shared Analytics'

  Correct

  A lookup resolves a named object's ID, so the correct ID is used in each workspace.

• BThe warehouse ID hard-coded under each target's variables mapping

  Incorrect

  This works but must be updated by hand whenever a warehouse is recreated, which the team wants to avoid.

• CThe ${workspace.host} substitution in each task's settings

  Incorrect

  This substitution returns the workspace URL, not a warehouse ID.

• Dtype: complex with the warehouse name set as the variable's default

  Incorrect

  A complex variable holds only the value it is given; it does not resolve a name to an ID.

Variable lookups resolve IDs for objects such as clusters, warehouses, jobs, pipelines and service principals by name. A missing or duplicate name causes an error.

• Asales_${bundle.environment}

  Incorrect

  The documentation says to use ${bundle.target} instead of ${bundle.environment}.

• Bsales_${workspace.current_user.short_name}

  Incorrect

  This varies by the deploying user, not by the target.

• Csales_${var.schema}, without declaring var.schema

  Incorrect

  A variable must be declared, and given a value, before it can be referenced.

• Dsales_${bundle.target}

  Correct

  ${bundle.target} resolves to the target name at deployment, giving one definition that adapts to each target.

Substitutions such as ${bundle.target}, ${bundle.name} and ${workspace.current_user.userName} insert deployment context into resource settings.

• ARedeploy the bundle with the new --var value before each run

  Incorrect

  This works but turns every run into a deployment, which is not how per-run values are meant to be passed.

• BDefine a job parameter and pass it with bundle run --params

  Correct

  Bundle variables are resolved at deployment; job parameters carry run-time values into a run.

• CSet BUNDLE_VAR_run_date before calling bundle run

  Incorrect

  Environment variables also set a deployment-time variable value, with the same limitation as --var.

• DPut the date in .databricks/bundle/<target>/variable-overrides.json

  Incorrect

  This file is another source of deployment-time variable values.

Bundle variables shape what is deployed; job parameters shape each run. Use bundle run --params for values that change between runs.