We introduce the new Wave model for exposing the temporal relationship among the queries in data-intensive distributed computing. The model defines the notion of query series to capture the recurrent nature of batched computation on periodically updated input streams. This seemingly simple concept captures a significant portion of the queries we observed in a production system. The recurring nature of the computation on the same steam opens up surprisingly significant opportunities for achieving better performance and higher resource utilization.