We shut off this failpoint which was hanging initial sync before stopping the replica set. If the replication coordinator realized that the node was in shut down, the node will shut down fine. However, if the node is not in shut down when the initial sync attempt fails, it will fassert.

I think we should let shut down interrupt the failpoint instead of turning the failpoint off to make sure that the replication coordinator is aware of shutdown when the initial syncer finishes the attempt.