How to use a specific database on our mysql instance?

0-2 years experience

1. 1We have a MySQL server with several databases. How would you write a script to connect and run a query against the 'analytics' database?
2. 2If you run SELECT DATABASE(); after connecting, what will it return if you haven't selected a database?
3. 3What command would you use to switch from the 'test' database to the 'production' database in a MySQL session?

2-5 years experience

1. 1Our new feature needs to write logs to a dedicated 'logs' database, but the existing codebase always connects to the default schema. How would you modify the connection logic, and what pitfalls would you watch for?
2. 2During a deployment we saw errors like 'Unknown database "metrics"' after adding a new DB. Walk me through how you'd debug why the application couldn't see the new database.
3. 3We want to run a migration that creates tables in a specific database without affecting others. Which MySQL options or statements would you use, and why?

5-8 years experience

1. 1We are building a multi‑tenant SaaS where each tenant gets its own MySQL database on a shared instance. How would you design the connection handling to reliably select the correct database per request, considering connection pooling and failover?
2. 2Our read replica lag is causing occasional 'Database not selected' errors when the app switches databases mid‑transaction. How would you address this at the system level?
3. 3If we need to enforce that certain users can only access a specific database, what MySQL privileges and architecture would you put in place to prevent accidental cross‑database queries?

8+ years experience

1. 1We plan to migrate from a monolithic MySQL instance with many databases to a sharded architecture where each shard hosts a subset of databases. What high‑level strategy would you propose for routing traffic to the correct database, handling schema changes, and minimizing downtime across teams?
2. 2Our organization wants to standardize how services specify the target database in connection strings to avoid hard‑coded USE statements. How would you design a cross‑team convention and tooling to enforce this, and what migration path would you suggest?