Are Updatable Learned Indexes Ready?
arXiv:2207.02900 · doi:10.14778/3551793.3551848
Abstract
Recently, numerous promising results have shown that updatable learned indexes can perform better than traditional indexes with much lower memory space consumption. But it is unknown how these learned indexes compare against each other and against the traditional ones under realistic workloads with changing data distributions and concurrency levels. This makes practitioners still wary about how these new indexes would actually behave in practice. To fill this gap, this paper conducts the first comprehensive evaluation on updatable learned indexes. Our evaluation uses ten real datasets and various workloads to challenge learned indexes in three aspects: performance, memory space efficiency and robustness. Based on the results, we give a series of takeaways that can guide the future development and deployment of learned indexes.
References in corpus (13)
- The Wide-field Infrared Survey Explorer (WISE): Mission Description and Initial On-orbit Performance
- Neo: A Learned Query Optimizer
- The PGM-index: a multicriteria, compressed and learned approach to data indexing
- Bao: Learning to Steer Query Optimizers
- FITing-Tree: A Data-aware Index Structure
- Learning Multi-dimensional Indexes
- Are Updatable Learned Indexes Ready?
- APEX: A High-Performance Learned Index on Persistent Memory
- From WiscKey to Bourbon: A Learned Index for Log-Structured Merge Trees
- RadixSpline: A Single-Pass Learned Index
- Tsunami: A Learned Multi-dimensional Index for Correlated Data and Skewed Workloads
- Bounding the Last Mile: Efficient Learned String Indexing
- Shift-Table: A Low-latency Learned Index for Range Queries using Model Correction