摘要
针对使用时序数据库存储海量时序数据场景下,数据量增大导致数据写入和查询效率降低的问题,提出一种基于时标分层的时序数据存储引擎。通过将内存(RAM)、固态硬盘(SSD)、机械硬盘(HDD)三种存储介质存储特性与时序数据业务场景相结合,基于数据时标实现数据热度划分、数据分层与数据迁移操作,提高RAM和SSD使用率,提升数据写入和查询效率。结合该引擎对现有"海迅"实时库(HS)进行优化改造,改造后的数据库写入和查询性能分别提升70%和200%。
Aiming at the problems of data writing and query performance decreasing when using time series database to store massive time series data, this paper presentd a time series data storage engine based on time-scale stratification by combining the performance characteristics of Random Access Memory（ RAM）, Solid State Drive（ SSD） and Hard Disk Driver（ HDD） with the time-series data business scenes, implemented data partitioning, data layering and data migration operations based on data time scale, increases data writing and query efficiency by optimizing RAM and SSD usage. After the optimizing High Soon real-time database with this engine, the writing and query performance is improved by about 70% and 200%.
出处
《计算机应用》
CSCD
北大核心
2017年第A01期246-249,共4页
journal of Computer Applications
