We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
A quantitative evaluation of persistent memory hash indexes.
- Authors
Chen, Zhiwen; Hu, Daokun; Che, Wenkui; Sun, Jianhua; Chen, Hao
- Abstract
Persistent memory (PMem) is increasingly being leveraged to build hash-based indexing structures featuring cheap persistence, high performance, and instant recovery. Especially with the release of Intel Optane DC Persistent Memory Modules, we have witnessed a flourish in (re)designing persistent hash indexes. However, most of them are focus on the evaluation of specific metrics with important properties sidestepped. Thus, it is essential to understand how the proposed hash indexes perform under a unified testing framework and how they differentiate from each other if a wider range of performance metrics are considered. To this end, this paper provides a comprehensive evaluation of persistent hash tables. In particular, we focus on the evaluation of several state-of-the-art hash tables including CCEH, Dash, PCLHT, Clevel, Viper, Halo, SOFT, and Plush, with the second-generation PMem hardware. Our evaluation was conducted using a unified benchmarking framework and representative workloads. Besides characterizing common performance properties, we also explore how hardware configurations (such as PMem bandwidth, CPU instructions, and NUMA) affect the performance of PMem-based hash tables. With our in-depth analysis, we identify design trade-offs and good paradigms in prior arts and suggest desirable optimizations and directions for the future development of PMem-based hash tables.
- Subjects
COMMONS; INDEXING
- Publication
VLDB Journal International Journal on Very Large Data Bases, 2024, Vol 33, Issue 2, p375
- ISSN
1066-8888
- Publication type
Article
- DOI
10.1007/s00778-023-00812-1