8 December 2023 in Proceedings of the ACM on Management of Data
The exponential growth of spatial data poses new challenges to the performance of spatial databases. Spatial indexes like R-tree greatly accelerate the query performance and can be effectively constructed through packing, i.e., loading all data into the index at once. However, existing R-tree packing methods rely on a set of fixed heuristic rules, which may not be suitable for different data distributions and workload patterns. To address the limitations of existing R-tree packing methods, we propose PLATON, a top-down R-tree packing method with learned partition policy that explicitly optimizes the query performance with regard to the given data and workload instance. We develop a learned partition policy based on Monte Carlo Tree Search and carefully make design choices for the MCTS exploration strategy and simulation strategy to improve algorithm convergence. We propose a divide and conquer strategy and two optimization techniques, early termination and level-wise sampling, to drastically reduce the MCTS algorithm's time complexity and make it a linear-time algorithm. Experiments on both synthetic and real-world datasets demonstrate the superior performance of PLATON over existing R-tree variants and recently proposed learned/workload-aware spatial indexes.
17 June 2025 in Proceedings of the ACM on Management of Data
Key-value databases with Log Structured Merge tree are increasingly favored by modern applications. Apart from supporting fast lookup on primary key, efficient queries on non-key attributes are also highly demanded by many of these applications. To enhance query performance, many auxiliary structures like secondary indexing and filters have been developed. However, existing auxiliary structures suffer from three limitations. First, creating filter for every disk component has low lookup efficiency as all components need to be searched during query processing. Second, current secondary index design requires primary table access to fetch the data entries for each output primary key from the index. This indirect entries fetching process involves significant point lookup overhead in the primary table and hence hinders the query performance. Last, maintaining the consistency between the secondary index and the primary table is challenging due to the out-of-place update mechanism of the LSM-tree. To overcome the limitations in existing auxiliary structures for non-key attributes queries, this paper proposes a novel secondary index framework, NEXT, for LSM-based key-value storage system. NEXT utilizes a two-level structure which is integrated with the primary table. In particular, NEXT proposes to create secondary index blocks on each LSM disk component to map the secondary attributes to their corresponding data blocks. In addition, NEXT introduces a global index component which is created on top of all secondary index blocks to direct the secondary index operation to the target secondary index blocks. Finally, NEXT adopts two optimization strategies to further improve the query performance. We implement NEXT on RocksDB and experimentally evaluate its performance against existing methods. Experiments on both static and mixed workloads demonstrate that NEXT outperforms existing methods for different types of non-key attributes.
30 September 2024 in Proceedings of the ACM on Management of Data
With the increasing availability of trajectory data, it is important to have good indexes to facilitate query processing. In this work, we propose BT-Tree, which is built through a recursive bi-partitioning approach, for the processing of range and KNN queries for past trajectory data. We first propose a cost function based method (CFBM) to build the BT-Tree. Specifically, we design a novel cost function, which incorporates the characteristics of both the data and historical query workload, to decide how to partition a BT-Tree node. Then we propose a reinforcement learning (RL) based method to address CFBM's limitations, such as making locally optimal decisions that may lead to global suboptimality. Experiments on three real datasets with up to 800 million data points show that the CFBM generally outperforms the baselines in terms of query processing time and the RL based method consistently outperforms the baselines and has more significant advantages on larger datasets.