CAREER: Scalable Self Managing Multimedia Storage

职业：可扩展的自管理多媒体存储

• 批准号: 0447671
• 负责人: Surendar Chandra
• 金额: --
• 依托单位: University of Notre Dame
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2005
• 资助国家: 美国
• 起止时间: 2005-02-01 至 2010-01-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0447671&HistoricalAwards=false
• 关键词: CAREER; Scalable; Self; Managing; Multimedia

This project develops an incrementally scalable and self-managing platform for profuse media capture and storage using a distributed architecture. The storage is tuned to exploratory environments where video sensors and storage bricks are quickly deployed to profusely capture the environment. The goal of this work is to balance the storage robustness requirements, brick and network capacity. To this end, this work addresses three important research areas: i) self-contained lifetime abstraction that lets the sensors and storage bricks to independently prune and manage the storage without coordination with other system components ii) self-organizing overlay mechanisms to locate candidate storage bricks for storing segments such that the system can be incrementally scalable, and iii) self-management mechanisms using storage replication, migration and rejuvenation to provide the distributed storage. The impact of this research will be to leverage the easy availability of multimedia sensors and extend the relevancy of high fidelity multimedia capture in remote locations without the associated management costs.

这个项目开发了一个增量可扩展和自我管理的平台，丰富的媒体捕获和存储使用分布式架构。存储针对探索性环境进行了调整，在这些环境中，视频传感器和存储砖可以快速部署，以大量捕获环境。这项工作的目标是平衡存储鲁棒性要求、砖和网络容量。为此，这项工作涉及三个重要的研究领域：i）自包含的寿命抽象，其允许传感器和存储块独立地修剪和管理存储而不与其它系统组件协调; ii）自组织覆盖机制，其定位用于存储段的候选存储块，使得系统可以是递增可扩展的;以及iii）使用存储复制的自管理机制，迁移和恢复以提供分布式存储。这项研究的影响将是利用多媒体传感器的容易获得性，并在远程位置扩展高保真多媒体捕获的相关性，而无需相关的管理成本。