In SERVER-127745 we encountered a bug happening during resharding when the recipient was building an index and two step-up/step-down happened while the index build was still in progress.

During the PR review the question came up that we could have some generic concurrency/FSM test that would randomly exercise this code path, if we had this kind of test before it could potentially have caught the issue earlier.

So the idea is to have an FSM test that triggers resharding on collections that already have indexes, so that the recipient would need to rebuild the indexes. We have suites that trigger random stepdown while running FSM tests, so the above scenario should be covered.