Knowing what your resources are doing, and what changed and when, is the difference between a quick fix and a guess.
Server and application logs
Your own logs live inside your servers: the operating system logs and your application's logs. Connect over SSH (how to connect) to read them. For anything important, ship these somewhere off the server so you still have them if the server fails.
Platform metrics
The console shows health and basic metrics for your resources, so you can spot a server that is constantly maxed out (a sign to resize) or a disk filling up.
Account activity
Actions taken in your account are recorded, which matters for security and for working out what changed before a problem. If you run a team, pair this with least-privilege roles and review access from time to time.
For deeper, programmatic metrics, the platform exposes monitoring through its API; see the developer documentation.