The FUSÉE Platform: Fast, Unified, and Scalable Event processing and Event messaging

在过去的十年里，大数据现象使主要参与者能够分析从各种来源收集的大量数据，以影响长期决策。下一个大数据挑战是高速提供大规模数据分析，以便对新收集的信息做出即时反应。 为了应对这一挑战，我建议使用基于事件的系统，旨在处理连续生成的数据流。基于事件的系统分为两类：从多个分布式来源收集数据的事件消息传递系统和能够实时分析信息的事件处理系统。这两种类型的系统都是独立构建的，然后连接在一起，形成一个完整的数据收集和分析解决方案。 在本提案中，我介绍了FUSéE平台：快速、统一、可扩展的事件处理和事件消息传递。FuséE通过共同考虑事件处理和事件消息传递，并利用针对各种目标大数据应用的整体优化机会，使自己有别于其他基于事件的系统，例如物联网、社交网络、在线游戏。 这项以事件为基础的运动在欧洲特别受欢迎，但在加拿大值得更多关注。FUSéE将在加拿大推广基于事件的系统的使用，并将成为全球首个研究基于事件的大数据应用的综合平台。 该项目的第一个贡献将是交付作为开放源码代码库的FUSéE平台，并在加拿大的一些工业和学术合作中实施概念验证。 其次，当事件处理和事件消息作为一个统一问题处理时，FUSéE的核心研究将识别优化机会，将这些优化形式化，提供创新算法，并使用我们的平台对其进行评估。 最后，FuséE统一平台将利用关键的数据管理和联网技术，在这些技术可用时提高性能。从这些集成工作中获得的见解将提高基于事件的系统在各种领域中的适用性。

In the past decade, the big data phenomenon has empowered major players with the ability to analyze large volumes of data collected from a wide variety of sources to influence long-term decision-making. The next big data challenge is to provide large-scale data analysis at high speed, in order to react instantaneously to newly gathered information. To address this challenge, I propose the use of event-based systems, designed to process streams of continuously generated data. Event-based systems fall into two categories: event messaging systems, which collect data from multiple distributed sources, and event processing systems, which are capable of analyzing information in real-time. Both types of systems are built independently and then connected together to form a complete solution for data collection and analysis. In this proposal, I present the FUSÉE platform: Fast, Unified, Scalable Event processing and Event messaging. FUSÉE distinguishes itself from other event-based systems by jointly considering both event processing and event messaging, and leveraging holistic optimization opportunities for a variety of targeted big data applications, such as the Internet of Things, social networks, online games. The event-based movement is particularly popular in Europe but deserves more attention in Canada. FUSÉE will promote the use of event-based systems in Canada and will be the first comprehensive platform worldwide for studying event-based big data applications. The first contribution of the project will be the delivery of the FUSÉE platform as an open-source codebase and the implementation of proofs of concept in a number of Canadian industrial and academic collaborations. Second, the core research of FUSÉE will identify optimization opportunities when event processing and event messaging are treated as a unified problem, formalize these optimizations, deliver innovative algorithms, and evaluate them using our platform. Finally, the FUSÉE unified platform will leverage key data management and networking technologies for improved performance when such technologies are available. The insights gained in these integration efforts will raise the applicability of event-based systems for a wide variety of domains.