信息安全研究 ›› 2018, Vol. 4 ›› Issue (1): 63-72.

• 自主可控专题 • 上一篇    下一篇

一种安全可靠大数据存储平台的设计

蒋旭1,孙磊1,谭炜波2   

  1. 1. 天津市海量数据处理技术实验室
    2. 天津神舟通用数据技术有限公司
  • 收稿日期:2018-01-14 出版日期:2018-01-15 发布日期:2018-01-13
  • 通讯作者: 蒋旭
  • 作者简介:蒋旭,出生于1984年,工程硕士,工程师,主要研究领域分布式并行数据库、大数据存储、数据安全。 孙磊,出生于1982年,硕士,高级工程师,主要研究领域为分布式并行数据库、大数据分析。 谭炜波,出生于1984年,工程硕士在读,工程师,主要研究领域分布式并行数据库、数据安全。

The Design of One kind of secure reliable bigdata storage platform

  • Received:2018-01-14 Online:2018-01-15 Published:2018-01-13

摘要: 基于目前国产通用关系型数据库软件,提出“大数据”适应性改造方案。该方案提出基于行列混合的压缩存储引擎(HCC),解决“大数据”的磁盘I/O读取性能问题并降低了存储采购成本;利用智能索引、Hash索引、子串索引和自定义分词索引技术,解决“大数据”精确查询的性能问题;采用多机并行计算技术(MPP)和多CPU核心并行计算技术(SMP),解决“大数据”统计分析性能问题;通过数据全生命周期管理,解决“大数据”硬件资源优化分配问题;构建在线平滑扩展的完全无共享平台架构,解决“大数据”膨胀带来的系统扩展性问题;通过云化的大数据安全框架,设计了分布式的大数据安全解决方案;在应用改造方案中设计并实现了“大数据”存储平台,通过测试与应用效果分析,验证了技术方案的合理性,平台的技术指标接近国外同类产品的目标。

关键词: 大数据, 并行计算, 水平扩展, 大数据安全, 大数据存储

Abstract: Based on the current domestic DBMS, the "BIGDATA" adaptation solution is put forward to improve the traditional DBMS. Designed and implemented the storage engine using the hybrid column compression technology to solve disk I/O problem of the "BIGDATA". Designed and implemented the intelligent indexing, hash index, substring indexes and custom word indexing technology to solve the performance problem of the key-search queries on the "BIGDATA" environment. Designed and implemented the multi-machine parallel (MPP) computing technology, multi-CPU core parallel (SMP) computing technology to solve the performance problem of the statistical analysis on the "BIGDATA" environment. Designed and implemented the data lifecycle management solutions to make purpose of the most valuable data using the most advantage of hardware resources. Designed and implemented the Share-Nothing architecture using auto extends online technology to solve the problem of the data become larger and larger over time. Through the cloud-based big data security framework, a distributed big data security solution is designed. To sum up in a word. Achieved a " big Data " storage platform in application transformation solution. Through testing and application effect analysis, verified the rationality of technical solutions. Specification of the Platform close to the target of similar foreign products.

Key words: big data, parallel computing, horizontal expansion, secure reliable bigdata, bigdata storage