What steps should be taken if a database slows down under increased load?

When a database is running slowly under load, the following steps should be taken:

1. Monitoring and metrics collection: check CPU, memory, disk, network load, and the number of active connections.
2. Query analysis: identify slow or frequently executed queries using a profiler or logs.
3. Query optimization: add indexes, rewrite complex queries, use caching.
4. Database configuration check: ensure that memory settings, connection pools, and other configurations are optimal.
5. Scaling: if necessary, add replicas, sharding, or increase server resources.
6. Lock and contention check: identify and eliminate locks that slow down operation.

Example: if a PostgreSQL database is used in a Go application, enable slow query logging and analyze them, then optimize or add indexes.