The `ConnectionPool` interface has a notion of a `PoolId`, which is used to maintain separate pools for each `HostAndPort`. However, it does not seem to have a similar concept to distinguish between individual connections in each pool. As a result, many of the connection pool's logs (e.g., this one) address specific lifetime changes for individual connections (spawn, refresh, destruct) without providing a way to identify which connection this is happening to.

Incorporating a connection ID for each connection within a `SpecificPool` would help trace through individual connection lifetimes through logs so that it's easier to correlate operation success/failure with the connection that they used.