注册 登录  
 加关注
   显示下一条  |  关闭
温馨提示!由于新浪微博认证机制调整,您的新浪微博帐号绑定已过期,请重新绑定!立即重新绑定新浪微博》  |  关闭

Ooi Beng Chin 黄铭钧

Databases, Machine Learning and Systems

 
 
 

日志

 
 

LADS: Exploiting Single-Threaded Model in Multi-Core Systems  

2014-07-24 10:52:36|  分类: 默认分类 |  标签: |举报 |字号 订阅

  下载LOFTER 我的照片书  |

The widely adopted single-threaded OLTP model assigns a single thread to each static partition of the database for processing transactions in a partition. This simplifies concurrency control while retaining parallelism. However, it suffers performance loss arising from skewed workloads as well as transactions that span multiple partitions. In this paper, we present a dynamic single-threaded in-memory OLTP system, called LADS, that extends the simplicity of the single-threaded model. The key innovation in LADS is the separation of dependency resolution and execution into two nonoverlapping phases for batches of transactions. After the first phase of dependency resolution, the record actions of the transactions are partitioned and ordered. Each independent partition is then executed sequentially by a single thread, avoiding the need for locking. By careful mapping of the tasks to be performed to threads, LADS is able to achieve a high degree of balanced parallelism. We evaluate LADS against H-Store, a partition-based database; DORA, a data-oriented transaction processing system; and Silo, a multicore in-memory OLTP engine. The experimental study shows that LADS achieves up to 20x higher throughput than existing systems and exhibits better robustness with various workloads.

We started to work on this in 2014 as part of memepiC (distributed in-memory data flow system based on epiC programming model/api -- see previous post on memepiC), and submitted the first version to VLDB 15 early 2015, and deposited into CoRR in March 2015[1].  Due to fine tuning, the system has improved, but the idea of separating conflict resolution/concurrency control from execution remains the core concept [2].  However, more and more comparisons were done along the way, and the results show LADS achieves very good performance for high contention workloads.  Two versions of the paper could be found below.

 

References:

[1] C. Yao et al: DGCC: A New Dependency Graph based Concurrency Control Protocol for Multicore Database Systems, CoRR, March 2015.

[2] C. Yao, D. Agrawal, G. Chen, Q. Lin, B.C. Ooi, W.F. Wong, M. Zhang: Exploiting Single-Threaded Model in Multi-Core Systems, December 2015.  (the latest version is to appear in IEEE TKDE, 2016).

  评论这张
 
阅读(653)| 评论(0)
推荐 转载

历史上的今天

在LOFTER的更多文章

评论

<#--最新日志,群博日志--> <#--推荐日志--> <#--引用记录--> <#--博主推荐--> <#--随机阅读--> <#--首页推荐--> <#--历史上的今天--> <#--被推荐日志--> <#--上一篇,下一篇--> <#-- 热度 --> <#-- 网易新闻广告 --> <#--右边模块结构--> <#--评论模块结构--> <#--引用模块结构--> <#--博主发起的投票-->
 
 
 
 
 
 
 
 
 
 
 
 
 
 

页脚

网易公司版权所有 ©1997-2017