金融分布式数据库异步全局索引研究

doi:10.3778/j.issn.1673-9418.2208104

摘要/Abstract

摘要： 伴随移动支付的快速发展，传统的金融核心业务所使用的集中式数据库面临性能、可用性等方面的挑战。为此有研究者提出了采用计算和存储分离架构的分布式数据库解决思路，通过网络连接构建在物理上分散、逻辑上集中的分布式数据库管理系统。全局索引是分布式数据库提升查询效率的重要手段，但当前的全局索引主要通过在传统网络中利用同步机制来实现。这种机制在金融数据管理的典型交易和批量导入数据场景下面临单表索引数量少、吞吐量降低、事务响应时间抖动等亟需解决的问题。为此，提出了一种基于分布式数据库的异步全局索引架构，通过新增MQ消息队列、分布式缓存，以及利用RDMA网络，实现一种满足金融典型场景需求的异步全局索引。为了验证有效性，与Oracle、CockroachDB在相同的场景下进行大量对比分析，实验结果表明，在金融核心业务数据的批量导入、交易等场景中，该方法的主要性能指标比现有方法提升60%以上，而对系统资源需求降低20%以上。

关键词: 分布式数据库, 异步, 全局索引, 低时延, RDMA网络

Abstract: With the rapid development of mobile payment, the traditional centralized database used by the financial core business is confronted with challenges in terms of performance and availability. For this reason, some researchers propose a distributed database solution based on the architecture of separation of computing and storage, and build a physically decentralized and logically centralized distributed database management system through network connection. Global index is an important method for distributed database to improve query efficiency, but current global indexing is mainly achieved through the use of synchronization mechanisms in traditional networks.Under the typical transaction and batch import data scenarios of financial data management, this mechanism faces the problems that need to be solved urgently, such as small number of indices in a single table, reduced throughput, and jittery transaction response time. Therefore, an asynchronous global index architecture of distributed database is proposed. This mechanism is applied to adding MQ message queue and distributed cache, and uses RDMA network to realize an asynchronous global index to meet the demands of the real financial scenarios. Through sufficient comparative experiments with Oracle and CockroachDB, experimental results show that the performance is improved by more than 60% compared with the existing methods, and the demand for system resources is reduced by more than 20% in the batch import and transaction of financial core business data.

Key words: distributed database, asynchronous, global index, low latency；RDMA network

金磐石, 李博涵, 秦小麟, 邢磊, 李晓栋, 王进. 金融分布式数据库异步全局索引研究[J]. 计算机科学与探索, 2023, 17(11): 2784-2794.

JIN Panshi, LI Bohan, QIN Xiaolin, XING Lei, LI Xiaodong, WANG Jin. Research on Asynchronous Global Index of Financial Distributed Database[J]. Journal of Frontiers of Computer Science and Technology, 2023, 17(11): 2784-2794.

参考文献

[1] 罗毅, 林承军. 加强分布式数据库建设赋能银行数字化转型[J]. 中国金融电脑, 2021(11): 30-31.
LUO Y, LIN C J. Strengthen the construction of distributed databases to enable the digital transformation of banks[J].China Financial Computer, 2021(11): 30-31.
[2] 杨頲, 韩路, 许勤, 等. 大型商业银行基于Hadoop分布式数据仓库建设初探[J]. 计算机应用与软件, 2017, 34(8): 72-75.
YANG T, HAN L, XU Q, et al. A preliminary study on the construction of Hadoop based distributed data warehouse for large commercial banks[J]. Computer Application and Software, 2017, 34(8): 72-75.
[3] 万维威, 谷鹏. 基于分布式网络设备的分布式数据库设计[J]. 通信技术, 2022, 55(3): 318-329.
WAN W W, GU P. Design of distributed database based on distributed network equipment[J]. Communication Techno-logy, 2022, 55(3): 318-329.
[4] 马友忠, 孟小峰. 云数据管理索引技术研究[J]. 软件学报, 2015, 26(1): 145-166.
MA Y Z, MENG X F. Research on cloud data management index technology[J]. Journal of Software, 2015, 26 (1): 145-166.
[5] 朱涛, 郭进伟, 周欢, 等. 分布式数据库中一致性与可用性的关系[J]. 软件学报, 2018, 29(1): 131-149.
ZHU T, GUO J W, ZHOU H, et al. Relationship between consistency and availability in distributed database[J]. Jour-nal of Software, 2018, 29(1): 131-149.
[6] 朱晓真, 邓昌胜, 陈章立. 用于将大数据量的文本文件导入分布式数据库的方法及装置: ZL202011481442.4[P]. 2021-04-02.
ZHU X Z, DENG C S, CHEN Z L. Method and device for importing large amount of text files into distributed database: ZL202011481442.4[P]. 2021-04-02.
[7] 崔斌, 高军, 童咏昕, 等. 新型数据管理系统研究进展与趋势[J]. 软件学报, 2019, 30(1): 164-193.
CUI B, GAO J, TONG Y X, et al. Research progress and trend of new data management system[J]. Journal of Soft-ware, 2019, 30(1): 164-193.
[8] KALIA A, KAMINSKY M, ANDERSEN D G. Using RDMA efficiently for key-value services[J]. ACM SIGCOMM Com-puter Communication Review, 2014, 44(4): 295-306.
[9] SRINIVASAN V, CAREY M J. Performance of B-Tree concur-rency control algorithms[J]. VLDB Journal, 1991, 20(2): 416425.
[10] 张晓祎. 基于非易失内存的高效索引方法及应用研究[D]. 武汉: 华中科技大学, 2019.
ZHANG X Y. Research on efficient indexing methods and applications based on nonvolatile memory[D]. Wuhan: Hua-zhong University of Science and Technology, 2019.
[11] 吴植民. 一种基于B+树异步更新索引的读优化方法: ZL201410697537.8[P]. 2015-03-11.
WU Z M. A read optimization method based on asynchro-nous update index of B+tree: ZL201410697537.8[P]. 2015-03-11.
[12] LIU J, WU J, KINI S P, et al. High performance RDMA-based MPI implementation over InfiniBand[C]//Proceedings of the 17th Annual International Conference on Supercom-puting, San Francisco, Jun 23-26, 2003. New York: ACM, 2003: 295-304.
[13] WU S, JIANG D, OOI B C, et al. Efficient B-tree based indexing for cloud data processing[J]. Proceedings of the VLDB Endowment, 2010, 3(1/2): 1207-1218.
[14] 林朝晖, 于俊清, 何云峰, 等. 高维分布式局部敏感哈希索引方法[J]. 计算机科学与探索, 2013, 7(9): 811-818.
LIN C H, YU J Q, HE Y F, et al. High-dimensional distri-buted indexing based on locality-sensitive hashing[J]. Jour-nal of Frontiers of Computer Science and Technology, 2013, 7(9): 811-818.
[15] 王习特, 朱宗梅, 于雪苹, 等. 异构分布式环境中的并行离群点检测算法[J]. 湖南大学学报(自然科学版), 2020, 47(10): 100-110.
WANG X T, ZHU Z M, YU X P, et al. Parallel outlier detec-tion algorithm in heterogeneous distributed environment[J]. Journal of Hunan University (Natural Science Edition), 2020, 47(10): 100-110.
[16] 周维, 路劲, 周可人, 等. 基于并发跳表的云数据处理双层索引架构研究[J]. 计算机研究与发展, 2015, 52(7): 1531-1545.
ZHOU W, LU J, ZHOU K R, et al. Research on cloud data processing double-layer index architecture based on concur-rent hop table[J]. Journal of Computer Research and Deve-lopment, 2015, 52(7): 1531-1545.
[17] 胡耀艺, 胡卉芪, 周烜, 等. 面向非易失性内存缓冲区的SQLite-CC[J]. 华东师范大学学报(自然科学版), 2021, (6): 124-134.
HU Y Y, HU H Q, ZHOU X, et al. SQLite-CC for non-volatile memory buffer[J]. Journal of East China Normal University (Natural Science Edition), 2021(6): 124-134.
[18] HUANG D, LIU Q, CUI Q, et al. TiDB: a raft-based HTAP database[J]. Proceedings of the VLDB Endowment, 2020, 13(12): 3072-3084.
[19] HUANG J, OUYANG X, JOSE J, et al. High-performance design of HBase with RDMA over InfiniBand[C]//Procee-dings of the 26th IEEE International Parallel and Distri-buted Processing Symposium, Shanghai, May 21-25, 2012. Washington: IEEE Computer Society, 2012: 774-785.
[20] WU S, WU K L. An indexing framework for efficient retrie-val on the Cloud[J]. IEEE Data Engineering Bulletin, 2009, 32(1): 75-82.
[21] HAZEL T, JEFFORDS J P, BUTEAU G L. Method and system for append-only storage and retrieval of information: US20130226931A1[P]. 2013-08-29.
[22] JOSE J, SUBRAMONI H, MIAO L, et al. Memcached des-ign on high performance RDMA capable interconnects[C]//Proceedings of the 2011 International Conference on Paral-lel Processing, Taipei, China, Sep 13-16, 2011. Washington: IEEE Computer Society, 2011: 743-752.
[23] ARONOVICH L, SPIEGLER I. CM-tree: a dynamic clus-tered index for similarity search in metric databases[J]. Data & Knowledge Engineering, 2007, 63(3): 919-946.
[24] BARON J, BAZ H, BIXLER T, et al. AWS certified solu-tions architect official study guide: associate exam[M]. New York: John Wiley & Sons, 2016.