Simulation framework for in-memory key-value store architecture exploration

用于内存中键值存储架构探索的模拟框架

• 批准号: 474466-2014
• 负责人: Chow, Paul
• 金额: $ 1.82万
• 依托单位: University of Toronto
• 依托单位国家: 加拿大
• 项目类别: Engage Grants Program
• 财政年份: 2014
• 资助国家: 加拿大
• 起止时间: 2014-01-01 至 2015-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=564972
• 关键词: Simulation; framework; memory; key; value

The emergence of large, Internet-scale applications, such as FaceBook, has required the development of systems for efficiently storing and retrieving information from data stores that are far larger than have been used in traditional data storage systems. It is especially important that these systems be able to respond rapidly to your clicks on a web page, such as when you might be looking to find who are the friends of your friends. The systems used today are generally based on the use of commodity compute-server systems that use the native memory of the compute servers to act as the storage media for the data that needs to be accessed quickly. Several open-source software applications exist for coalescing the individual compute-servers into large-scale data servers. In this project, we do not assume that standard compute servers are used as the underlying hardware platform. Instead, we build a simulation environment that reproduces the functionality required

大型互联网规模应用程序（例如 FaceBook）的出现，需要开发能够从比传统数据存储系统中使用的数据存储大得多的数据存储中高效存储和检索信息的系统。尤其重要的是，这些系统能够快速响应您对网页的点击，例如当您可能想要查找谁是您朋友的朋友时。当今使用的系统通常基于商品计算服务器系统的使用，该系统使用计算服务器的本机内存作为需要快速访问的数据的存储介质。存在多种开源软件应用程序，用于将单个计算服务器合并为大规模数据服务器。在这个项目中，我们不假设使用标准计算服务器作为底层硬件平台。相反，我们构建一个模拟环境来重现所需的功能