Add Prometheus metrics for `using_database_replica`

As we're adding a fairly critical database utility in #95, we'd like to be able to effectively monitor its operation with Prometheus.

### Implementation

Add the following counter:

- `flagsmith_db_selected_replica_count_total`, labeled with `result` (`success`/`failure`) and `replica_name`.

Inside `using_database_replica`:

- On each successful `ensure_connection()` call, increment the counter: `flagsmith_db_selected_replica_count_total.labels(result="success", replica_name= chosen_replica).inc()`
- In each `OperationalError` exception handling branch, increment the counter: `flagsmith_db_selected_replica_count_total.labels(result="failure", replica_name=attempted_replica).inc()`

Check how existing metrics are used and defined by searching for an existing metric, e.g. `flagsmith_task_processor_enqueued_tasks_total`.

Refer to [developer documentation](https://github.com/Flagsmith/flagsmith-common?tab=readme-ov-file#guidelines) to learn how to develop and test Prometheus metrics.