If the balancer has been enabled during a test we might still have moveCollection/resharding operations ongoing after the test has finished even if the balancer was properly stopped. This means that the after-test checks may be running concurrently with moveCollection/resharding operations, which can lead to failures due to the expected inconsistencies caused by the ongoing DDL operations.

We have already tackled this on a few tickets (like SERVER-88191, SERVER-86564 and SERVER-89399) but we still see this kind of error on other types of checks.
The aim of this ticket is resolve this in a more generic way, like joining all the resharding ongoing operations once the ShardingTest stops and once the ShardedClusterFixture stops the balancer.