Question 780 A production job has a known defect. An engineer fixes its deployed notebook directly, but the version-controlled bundle still contains the old code. What long-term problem is created? The production Delta data automatically becomes read-only All branch protections are automatically removed Git creates a permanent correct patch automatically Configuration and code drift that may be overwritten on the next controlled deployment Explanation Correct Answer D:The deployable source of truth no longer matches production, so the hotfix may be lost. Incorrect Answers A:Direct notebook edits do not inherently change table access mode. B:A manual notebook edit does not by itself delete repository branch rules. C:Direct remote edits are not automatically recorded in the approved Git repository.
Question 781 A CI service principal can create a Lakeflow Job in a target workspace, but the job's configured run-as principal cannot SELECT its source table. What distinction was overlooked? If a job exists it always has permission to read every table Assume a service principal that can deploy jobs automatically receives SELECT on every governed table Grant the CI deploy principal more workspace permissions but leave the job run-as table grants unchanged Deployment privileges and runtime data-access privileges are separate Explanation Correct Answer D:Creating a job resource does not grant its execution identity the table permissions needed to process data. Incorrect Answers A:Job resource creation alone does not authorize all catalog data. B:Workspace deployment permissions and Unity Catalog table grants are checked independently for the configured run-as identity. C:The job accesses data as its run-as identity, not as the CI deployer. Extra deployer rights do not fix the runtime SELECT denial.
Question 782 A common bundle has dev run as an engineer and prod run as a governed service principal. Which configuration pattern makes this explicit? Use target-specific run_as settings in the bundle Disable production auditing to let the job run Hard-code a personal access token inside the notebook Force all environments to use the engineer's personal account Explanation Correct Answer A:Each target can specify an appropriate job execution principal without altering the shared task code. Incorrect Answers B:Audit settings do not replace a valid execution identity. C:Embedding a personal credential creates unnecessary security and lifecycle risks. D:A production service principal is a valid way to avoid personal-account dependencies.
Question 783 A release bundle needs a database password for a production workflow. The team wants configuration in Git without exposing that password. Which approach is appropriate? Encrypt the password locally but commit both ciphertext and decryption key in bundle YAML Reference an approved secret facility or environment binding rather than committing the password Append the password to the Git commit message Put the plain-text password in databricks.yml Explanation Correct Answer B:External secret management separates sensitive values from version-controlled declarative configuration. Incorrect Answers A:Keeping the decryption key beside the encrypted secret in Git does not protect the actual credential. C:Commit history permanently exposes the value and does not secure runtime access. D:A tracked YAML file can expose the secret to reviewers, clones and logs.
Question 784 A bundle declares a required `catalog` variable with no default. Production deployment fails during validation because the variable is unset. What is the correct remediation? Supply the production target variable through a supported CLI, environment, or target setting Delete the required variable and every use of it blindly Replace the missing catalog value with an arbitrary development catalog Disable validation and assume deployment fills the variable Explanation Correct Answer A:Required bundle substitutions must resolve before the target can be validated and deployed. Incorrect Answers B:Removing references without reviewing their purpose can break the resource definitions. C:An invented fallback can deploy production jobs to the wrong data namespace. D:A missing required input should not be ignored during promotion.
Question 785 A team changed a bundle's name and target identity and now sees what appears to be a second set of deployed resources. Why? Assume that deploying to the same workspace always reuses an existing bundle deployment regardless of name and target Match only displayed job names and assume identical resource identity across bundle targets The bundle identity includes its name, target and workspace Rename only the displayed job titles and assume the bundle deployment identity is unchanged Explanation Correct Answer C:Changing identity coordinates can cause a distinct deployment state rather than updating the original resources. Incorrect Answers A:A workspace can host multiple bundles; changing bundle identity coordinates can result in separate managed resource state. B:Display-name equality cannot prove deployments share a bundle identity or the same managed resource tracking state. D:Resource display names are distinct from bundle identity; the configured bundle name, target and workspace govern deployment tracking.
Question 786 Two release pipelines attempt to deploy the same bundle target simultaneously. What control should normally be preserved to avoid conflicting updates? Always bypass the lock with force-lock The bundle deployment lock or serialized deployment coordination Run both deployments against the same target without coordination Disable version control for one of the pipelines Explanation Correct Answer B:A deployment lock protects a target from concurrent state mutation by multiple release processes. Incorrect Answers A:Unconditionally bypassing safety locks risks conflicting resource updates. C:Two uncoordinated state changes can overwrite or interfere with one another. D:Removing version control does not solve concurrent remote deployment conflicts.
Question 787 A developer uses `databricks bundle sync` and expects remote notebook edits to be automatically copied back into the local Git checkout. Why is the expectation wrong? Use `databricks bundle deploy` to pull remote notebook edits into the local Git working tree Run bundle sync twice, assuming the second invocation reverses transfer direction Bundle sync sends local file changes to the remote workspace, not remote changes back to local Use `databricks bundle sync --full` expecting it to make the remote workspace the authoritative local source Explanation Correct Answer C:The documented sync direction is local to remote, so remote edits require a different recovery path. Incorrect Answers A:Bundle deployment applies local configuration remotely; it does not reverse sync remote notebook source into local files. B:Repeated sync applies the same local-to-workspace direction; it is not a bidirectional Git reconciliation operation. D:A full bundle sync is still local-to-remote; the flag does not reverse its transfer direction.
Question 788 A developer wants to inspect which workspace files a bundle sync would change without yet modifying the workspace. Which CLI option applies? Use `databricks bundle destroy` Use `databricks bundle run` Use `databricks bundle sync --full` without dry-run to preview changes Use `databricks bundle sync --dry-run` Explanation Correct Answer D:The dry-run option previews synchronization actions rather than applying them. Incorrect Answers A:Destroy is a destructive remote resource operation, not a sync preview. B:Run executes a configured job or other resource, not the file synchronization preview. C:A full sync performs synchronization; it is not a no-change preview of the proposed file operations.
Question 789 A developer switches from feature-A to feature-B in a Databricks Git folder but the job still points to a previously deployed feature-A asset. What is required before testing feature-B as the deployed workflow? Change all data tables to external managed status Rerun the old deployed job without updating its source revision Expect a workspace job to follow whichever local Git branch the developer currently views Validate and deploy the bundle from the intended feature-B source and target Explanation Correct Answer D:Checking out another Git branch alone does not update the remote bundle-managed job resource. Incorrect Answers A:Table ownership mode is unrelated to selecting the deployed code revision. B:The deployed job still uses the prior asset until the new source has been validated and deployed to the intended target. C:A live job definition remains tied to its deployed code/configuration, not the user's presently selected branch by default.
Question 790 A transformation unit test fails in a staging pipeline, but the production bundle is syntactically valid. What should the release gate do? Block production promotion until the failed behavioral test is resolved Delete the failing test and deploy immediately Promote because syntactic validation returned success Change the production target name to hide the failed test Explanation Correct Answer A:Bundle syntax validity cannot override a demonstrated data-transformation regression in tests. Incorrect Answers B:Removing evidence without fixing a real regression breaks the test gate. C:Syntactic validation and functional correctness are different release criteria. D:Renaming does not repair the transformation behavior.