SQL Database Source Control: 10 practice questions

Question 1 of 10

A developer adds a property by hand to the SQL project file that Fabric generated for a SQL database. What happens to that edit?

1. ASqlPackage reads it and then skips the project build.
2. BFabric keeps it and applies it on the next update.
3. CFabric overwrites it on the next commit from Fabric.
4. DFabric moves it into a post-deployment script.

Question 2 of 10

A team selects Branch out to new workspace from a workspace that holds a SQL database with committed tables. What does the database in the new workspace contain?

1. ANo objects until someone selects Update in that workspace
2. BThe committed objects, with new and empty tables
3. CThe committed objects, with a copy of every table row
4. DThe table rows only, while the definitions stay in Git

Question 3 of 10

How does the SQL project that Fabric generates treat the `.sharedQueries` folder?

1. AIt builds the scripts into the dacpac as database objects.
2. BIt turns each shared query into a stored procedure on commit.
3. CIt tracks the scripts but excludes them from the project build.
4. DIt keeps the folder in Fabric only, so Git never receives it.

Question 4 of 10

A developer pushes a `.sql` file that uses a feature that SQL database in Fabric does not support. Update all then fails in the Source control panel. What must happen before the workspace can be updated again?

1. AThe change is reverted manually in source control.
2. BThe publish adds a smart default in place of the feature.
3. CUpdate all is run again so that Fabric skips the file.
4. DThe generated project file is edited to exclude the file.

Question 5 of 10

A SQL database must run with a specific compatibility level in every stage. The team delivers changes with Git integration and deployment pipelines. How should the setting reach each stage?

1. AWith a T-SQL script that runs against the database after deployment
2. BWith a commit, because the project file stores database settings
3. CWith the publish step, which scripts the database options for you
4. DWith the deployment pipeline, which copies it from the source stage

Question 6 of 10

A developer writes table definitions as `.sql` files in Visual Studio Code. Team policy says a definition must be a file in the repository before the table exists in any database. Twice, a pushed file with invalid syntax made Update all fail for the whole team. Which workflow fits the policy and avoids that failure?

1. ACreate the tables in the Fabric SQL editor, then commit the database.
2. BPush the files, then let the build inside the Fabric update check them.
3. CSave the files in `.sharedQueries` so the project build checks them.
4. DBuild the SQL project locally, push the files, then update from Fabric.

Question 7 of 10

A lookup table must hold the same five rows in three databases. The development database is updated from Git, and the test and production databases are filled by a deployment pipeline. Changes to the rows must be reviewed as code, with no manual step after a release. What should the team use?

1. AA `MERGE` query in Shared Queries, set as a post-deployment script
2. BA `MERGE` query in Shared Queries, run by hand after each release
3. CA `MERGE` statement saved by hand in the generated project file
4. DA branch workspace, because branching out copies the lookup rows

Question 8 of 10

Two developers change stored procedures in the same SQL database. Each change needs its own environment and a review before it is shared. The release manager must choose the moment when the primary database changes. Which sequence meets these needs?

1. ABranch out to a workspace and complete the pull request, which updates the primary database.
2. BChange the primary database, commit it, then open a pull request to review the change.
3. CBranch out to a workspace, complete the pull request, then update the primary workspace.
4. DBranch out to a workspace, then publish its dacpac to the primary database with SqlPackage.

Question 9 of 10

Fabric publishes updates from source control with `/p:GenerateSmartDefaults = true`. A developer adds a column that does not allow null values to a table that already holds rows, then updates the database. What does this option do?

1. AIt empties the table so that the column can be added.
2. BIt provides a default value for the column automatically.
3. CIt creates the column as nullable and reports a warning.
4. DIt stops the update because the rows have no value for it.

Question 10 of 10

A developer pushes new table files and selects Update all in Fabric. The update succeeds and nobody edits the database afterward, yet the SQL database shows Uncommitted. What should the developer do?

1. ASelect Update all again until the status clears.
2. BDelete the generated project file and build it locally.
3. CRevert the pushed files, because the update did not apply.
4. DCommit the database to source control from Fabric.