Publication
ElasticBF constructs multiple small filters for each SSTable and dynamically loads them into memory as needed based on access frequency, realizing fine-grained and elastic adjustment of Bloom filter false positive rates to significantly improve read performance in large key-value stores.
Yongkun Li, Chengjin Tian, Fan Guo, Cheng Li, Yinlong Xu
ElasticBF dynamically adjusts Bloom filter sizes per SSTable based on hotness, reducing false positive rates for frequently accessed data and significantly improving read performance in LSM-tree KV stores.