Cquirrel: Continuous Query Processing over Acyclic Relational Schemas.
Published in International Conference on Very Large Data Bases (VLDB), 2021
Abstract
We will demonstrate Cquirrel, a continuous query processing engine built on top of Flink. Cquirrel assumes a relational schema where the foreign-key constraints form a directed acyclic graph, and supports any selection-projection-join-aggregation query where all join conditions are between a primary key and a foreign key. It allows arbitrary updates to any of the relations, and outputs the deltas in the query answers in real-time. It provides much better support for multi-way joins than the native join operator in Flink. Meanwhile, it offers better performance, scalability, and fault tolerance than other continuous query processing engines.
Citation
Qichen Wang, Chaoqi Zhang, Danish Alsayed, Ke Yi, Bin Wu, Feifei Li, and Chaoqun Zhan. “Cquirrel: Continuous Query Processing over Acyclic Relational Schemas.” International Conference on Very Large Data Bases (VLDB), August 2021. System demonstration.