The adaptive radix tree: ARTful indexing for main-memory databases

Abstract

Abstract-Main memory capacities have grown up to a point where most databases fit into RAM. For main-memory database systems, index structure performance is a critical bottleneck. Traditional in-memory data structures like balanced binary search trees are not efficient on modern hardware, because they do not optimally utilize on-CPU caches. Hash tables, also often used for main-memory indexes, are fast but only support point queries.To overcome these shortcomings, we present ART, an adaptive radix tree (trie) for efficient indexing in main memory. Its lookup performance surpasses highly tuned, read-only search trees, while supporting very efficient insertions and deletions as well. At the same time, ART is very space efficient and solves the problem of excessive worst-case space consumption, which plagues most radix trees, by adaptively choosing compact and efficient data structures for internal nodes. Even though ART's performance is comparable to hash tables, it maintains the data in sorted order, which enables additional operations like range scan and prefix lookup.

An experimental evaluation and analysis of database cracking

“…We show that full index creation with radix sort is twice as fast as with quick sort. We also show that ART [20] is up to 3.6 times faster than the AVL-tree in terms of lookup time. We also vary the query selectivity from very high selectivity to medium selectivity and compare the effects.…”

“…[25] Sideways cracking, Covered cracking [15], this paper resp. [25] Sorting: Quick(_insert) sort, Radix(_insert) [12,13] Full index: AVL-tree, B+-tree, ART [1,3,20] Parallel standard cracking P-SC [8,9] Parallel coarse-granular index P-CGI [2] Parallel-chunked standard cracking P-CSC [2] Parallel-chunked vectorized cracking P-CVC Variant of [23] Parallel-chunked coarse-granular index…”

“…Finally, we also extend the experimental parameters previously used in database cracking, e.g. by varying the query selectivities and by comparing against more recent, main memory optimized indexing techniques, including ART [20]. Our goal is to put database cracking in perspective by repeating several prior cracking works, giving new insights into cracking, and offering promising directions for future work.…”

“…The goal is to see whether or not it makes sense to use a sophisticated index structure as a baseline for cracking. We consider three index structures: (1) AVL-tree [1], (2) B+-tree [3], and (3) the very recently proposed ART [20]. We choose ART since it outperforms other main memory optimized search trees such as CSB+-tree [24] and FAST [19].…”

Practical rearrangement methods for dynamic double‐array dictionaries

“…Askitis and Sinha 33 proposed the HAT-trie that improves the burst trie using cache-conscious hash tables. Other dictionary structures include the adaptive radix tree 34 and the lexicographic tree. 35 In a recent study, Mavlyutov et al 36 concluded that the HAT-trie is an efficient data structure for managing Uniform Resource Identifier (URI) data from comparative experiments involving various data structures; however, they did not evaluate DA dictionaries.…”