Although we call findCmdFailPoint.wait(), this doesn't really help because wait() will only wait for the failpoint block to be entered once more from any command. The way the failpoint is defined, we will increment the timesEntered field for the failpoint on every find command while the failpoint is enabled. However, we will only hang the operation if the read is applied on the namespace that we pass in. Therefore, calling wait() will not guarantee that we have actually created the hanging operation that we care about on the primary.