A Chunk-Based Hash Table Caching Method for In-Memory Hash Joins

In-memory query processing can be accelerated by caching intermediate query results. Among various types of intermediate results, hash tables used by hash join are ideal objects for caching, as they can benefit a wide range of queries. In this paper, we introduce a fine-grained hash table caching method to benefit the hash-join operator. Our insight is that the fine-grained management of cached hash tables at the granularity of chunks can achieve optimal caching efficiency. As hash chunks can be reused more effectively, we further propose a cache-enabled hash join operator to adapt the cache under chunk granularity and accelerate in-memory hash join execution. Furthermore, we also consider concurrent accesses to cached hash chunks and present the scheduling strategy to improve throughput and consider response time simultaneously. To the best of our knowledge, our work is the first one that studies the delicate management of intermediate result caching for the concurrent workload. We integrated our caching method into a prototype to evaluate its performance. Experiments show that it can achieve significant performance improvement over brute force caching methods.