We can add an FSM workload where each worker thread acts on its own view namespace, and either creates, modifies, or drops the view definition. Since only one worker thread acts on any particular view namespace, we should be able to make strong assertions about what its view definition is at any point in time.

If we use the db[collName] collection that is prepared by the concurrency framework as the namespace to create the view on, then we can get additional coverage of a view on a sharded collection when the FSM workload runs as part of the concurrency_sharded suite.