In environments where the database might be momentarily unreachable (network blips, pod restarts, or failovers), this leads to immediate application-level failures.

This feature should introduce a mechanism to retry the physical establishment of a connection within the existing acquisitionTimeout window.