I-Corps: Networked Federated Infrastructure-as-a-Service

这个I-Corps项目的更广泛的影响/商业潜力在于向IT云服务市场引入一种新颖的颠覆性思想——云可以自下而上而不是自上而下地创建。这种方法允许许多中小型提供商创建云联盟，从而使云基础设施服务市场民主化，这些云联盟的规模可以与该领域中大型公司提供的单一垂直集成公共云产品相媲美。考虑到现代生活对云服务的依赖程度，这个想法可能会对现有的云市场、相关支持技术的发展以及我们与周围日常技术的互动方式产生深远的破坏性影响。如果成功，该项目将有助于创造新颖的解决方案，使依赖于动态的、深度互联的云基础设施的新型企业和服务能够无缝地将人、环境、健康和其他监测设备、科学仪器、工业设备、数据存储和处理设施连接在一起。这个I-Corps项目探索了实现云资源管理算法的软件控制代理的分布式架构的商业潜力。这是一种可扩展的资源描述语言，用于描述属于不同组织的云基础设施和对这些描述进行推理的管理算法，以便根据个人用户的需求实例化动态软件定义的基础设施。本研究的最终结果是一个功能良好的原型系统，该系统目前管理着支持分布式系统和网络研究的测试平台。这种云资源管理方法本质上是分布式和联合的，并且具有高度可扩展性，能够轻松地合并和支持新的云提供商和新的云资源。它与现有的商业和开源云解决方案明显不同，这些解决方案是集中的，孤立的，只能使用类似的技术。

The broader impact/commercial potential of this I-Corps project lies in introducing a novel disruptive idea into the market of IT cloud services - the idea that clouds can be created bottom up instead of top-down. The approach democratizes the market for cloud infrastructure services by allowing many small and medium providers to create federations of clouds that can rival the size of current monolithic vertically integrated public cloud offerings available from large corporations in this space. Considering the extent to which modern life depends on cloud services today, this idea can have a profoundly disruptive influence on the existing cloud marketplace, the evolution of the associated enabling technologies and how we interact with the day-to-day technologies surrounding us. If successful, this project will help create novel solutions that enable new types of businesses and services dependent on the dynamic, deeply interconnected cloud infrastructure seamlessly linking together people, environmental, health and other monitoring devices, scientific instruments, industrial equipment, data storage and processing facilities.This I-Corps project explores the commercial potential of a distributed architecture of software control agents implementing the cloud resource management algorithms. This is the extensible resource description language used to describe cloud infrastructures belonging to different organizations and the management algorithms reasoning over those descriptions in order to instantiate dynamic software-defined infrastructures for individual users based on their needs. The end result of this research is a functioning prototype system which today manages the testbed in support of distributed systems and networking research. This approach to cloud resource management is distributed and federated in nature and highly extensible by being able to easily incorporate and support new cloud providers and new cloud resources. It is markedly different from the existing commercial and open-source cloud solutions which are centralized, siloed and only able to work with like technologies.