SyncSQL: A Language to Express Views over Data Streams

SyncSQL: A Language to Express Views over Data Streams Thanaa M. Ghanem1 1 Department

˚ Larson2 Per-Ake

Walid G. Aref1

Ahmed K. Elmagarmid1

of Computer Science, Purdue University, ghanemtm,aref,[email protected] 2 Microsoft

Research, [email protected]

Abstract

even when input streams represent append-only relations. For example, consider an application monitoring a parking lot where two sensors continuously monitor the lot’s entrance and exit. The sensors generate two streams of identifiers, say S1 and S2 , for cars entering and exiting the lot, respectively. A reasonable query in this environment is Q1 :“Continuously keep track of the identifiers of all cars inside the parking lot”. The answer of Q1 is a view that, at any time point T, contains the identifiers for cars that are inside the parking lot. S1 can be modeled as a stream that inserts tuples into an append-only relation, say