If we get interrupted in AbstractReconnectionHandler#run, we abort the current attempt but keep scheduling the next reconnection:
We don't need to reschedule. There are two ways we can be interrupted:
the pool (i.e. the whole Cluster) is shutting down.
the client cancelled reconnection attempts with host.getReconnectionAttemptFuture().cancel(true). In that case the next scheduled attempt will immediately detect the cancellation and abort the reconnection process.
The current code does not cause any issue, but it schedules an unnecessary attempt in case #2 (we could abort right away instead of scheduling a task that will).