Increase slashing protection test SQL timeout to 500ms (#4574)

## Issue Addressed

NA

## Proposed Changes

We've been seeing a lot of [CI failures](https://github.com/sigp/lighthouse/actions/runs/5781296217/job/15666209142) with errors like this:

```
---- extra_interchange_tests::export_same_key_twice stdout ----
thread 'extra_interchange_tests::export_same_key_twice' panicked at 'called `Result::unwrap()` on an `Err` value: SQLError("Unable to open database: Error(None)")', validator_client/slashing_protection/src/extra_interchange_tests.rs:48:67
```

I'm assuming they're timeouts. I noticed that tests have a 0.1s timeout. Perhaps this just doesn't cut it when our new runners are overloaded.

## Additional Info

NA

validator_client/slashing_protection/src/slashing_database.rs

@@ -23,7 +23,7 @@ pub const POOL_SIZE: u32 = 1;
 #[cfg(not(test))]
 pub const CONNECTION_TIMEOUT: Duration = Duration::from_secs(5);
 #[cfg(test)]
-pub const CONNECTION_TIMEOUT: Duration = Duration::from_millis(100);
+pub const CONNECTION_TIMEOUT: Duration = Duration::from_millis(500);
 
 /// Supported version of the interchange format.
 pub const SUPPORTED_INTERCHANGE_FORMAT_VERSION: u64 = 5;