期刊文献+

面向内存数据网格的分布式事务保障机制 被引量:1

Distributed Transaction Guarantee Mechanism for In-Memory Data Grid
下载PDF
导出
摘要 在线事务处理(online transaction processing,OLTP)应用面临并发量和数据量持续增长的问题,并且高并发读写操作使得后台数据库成为瓶颈。内存数据网格(in-memory data grid,IMDG)是基于内存的新型分布式数据访问平台,是解决系统数据库写操作瓶颈的有效技术途径之一。然而内存数据网格中数据访问操作涉及的数据分布是不可预知的,需要提供分布式事务保障。针对内存数据网格的系统特点,提出了一种分布式事务保障机制,设计实现了事务处理模型、请求处理和数据定位方法以及事务保障协议,并规范化地定义了客户端与服务器端以及服务器端之间的操作接口。在事务处理基准测试TPC-W上的实验结果表明,新机制可以提高在线事务应用的处理速度,并具备良好的扩展性。 Online transaction processing (OLTP) Web applications face the continuous growth of concurrent users and data size. The high concurrent read and write operations make databases become the bottleneck. In-memory data grid (IMDG) is a new distributed data access platform based on memory, and an effective technical approach to elimi-nate the bottleneck of write operation in the database. However data distribution involved in data access operation of IMDG is unpredictable. A distributed transaction guarantee mechanism is necessary. For the IMDG system characteristics, this paper gives a distributed transaction guarantee mechanism, designs and implements a distributed transaction processing model, the way to do request processing and data accessing and the transaction guarantee protocol, and normalizes the application programming interface (API) between clients and servers. The experimental results on TPC-W show that the mechanism can improve the OTLP speed and support good scalability.
出处 《计算机科学与探索》 CSCD 2014年第7期812-822,共11页 Journal of Frontiers of Computer Science and Technology
基金 国家自然科学基金(61173003) 国家高技术研究发展计划(863计划)(2011AA040504)~~
关键词 内存数据网格(IMDG) 分布式事务 两阶段提交协议(2PC) in-memory data grid (IMDG) distributed transaction two-phase commit protocol (2PC)
  • 相关文献

参考文献15

  • 1Oracle Coherence[EB/OL]. [2013-07-20]. http://www.oracle. com/technetwork/middleware/coherence/overview/.
  • 2JBoss Infinispan[EB/OL]. [2013-07-20]. http://www.jboss.org/ infinispan/.
  • 3Glendenning L, Beschastnikh I, Krishnamurthy A, et al. Scal- able consistency in scatter[C]//Proceedings of the 23rd ACM Symposium on Operating Systems Principles (SOSP '11), Cascais, Portugal, Oct 23-26, 2011. New York, NY, USA: ACM, 2011: 15-28.
  • 4Das S, Agrawal D, E1 Abbadi A. G-Store: a scalable data store for transactional multi key access in the cloud[C]//Pro- ceedings of the 1 st ACM Symposium on Cloud Computing (SoCC '10), Indianapolis, USA, Jun 10-11, 2010. New York, NY, USA: ACM, 2010: 163-174.
  • 5Carvalho M, Romano P, Rodrigues L. Asynchronous lease- based replication of software transactional memory[C]//Pro- ceedings of the ACM/IFIP/USENIX 11 th International Con- ference on Middleware (Middleware '10), Bangalore, India, Nov 29-Dec 3, 2010. Berlin, Heidelberg: Springer-Verlag, 2010: 376-396.
  • 6Cattell R. Scalable SQL and NoSQL data stores[EB/OL]. (2011-11) [2013-07-20]. http://www.cattell.net/datastores/ Datastores.pdf.
  • 7Hibernate[EB/OL]. (2013-03-20) [2013-07-20]. http://en. wikipedia.org/wiki/Hibernate_(Java).
  • 8Gilbert S, Lynch N. Brewer's conjecture and the feasibility of consistent, available, partition-tolerant Web services[J]. ACM SIGACT News, 2002, 33(2): 51-59.
  • 9Gray J, Helland P, O' Neil P, et al. The dangers of replication and a solution[C]//Proceedings of the 1996 ACM SIGMOD In- ternational Conference on Management of Data (SIGMOD '96), Montreal, Canada, Jun 4-6, 1996. New York, NY, USA: ACM, 1996: 173-182.
  • 10Dormando. Memcached--a distributed memory object caching system[EB/OL]. [2013-07-20]. http://memcached.org/.

同被引文献1

引证文献1

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部