There have been dramatic increases in industrial investment and growing research interests in Internet of Things over the past few years. Driven by the application demand from Internet of everything in the future, the collaboration between traditional cloud and edge cloud is already on the way to provide better service to diverse users. However, due to the high marginal operation and maintenance costs, there remains a need for collaboration between traditional clouds. JointCloud computing is a new computing paradigm that supports collaboration within heterogeneous clouds, which is a complex problem that is difficult to manage using the traditional methodology of “top-down, decomposition, and reduction” in the benign construction of the JointCloud ecosystem. In this article, we first describe the inherent mechanism behind the constructional evolution of the JointCloud ecosystem, which details the inner workings of its elements and the general principals of how these various elements influence each other. We then propose a series of forward-looking, feasible operational methods to support the constructional evolution of the JointCloud ecosystem that incorporate efficient development methods, data-driven operations, AI-assisted decision making and so on. Finally, a case study of a cross-cloud migration deployment is used to illustrate the scientific basis of the systems inherent mechanism, as well as the efficacy of the supporting methods.

中文翻译：

---

关节云生态系统构建演进的内在机制及案例分析


过去几年，工业投资急剧增加，物联网研究兴趣不断增长。在未来万物互联的应用需求驱动下，传统云与边缘云的协同已经开始，为多样化的用户提供更好的服务。但由于边际运维成本较高，传统云之间仍然需要协作。云众计算是一种支持异构云内协作的新型计算范式，在众云生态的良性构建中，这是一个用传统的“自上而下、分解、约简”的方法难以管理的复杂问题。在本文中，我们首先描述了云云生态系统构建演化背后的内在机制，详细介绍了其各个要素的内部运作方式以及各个要素之间如何相互影响的一般原理。在此基础上，我们提出了一系列具有前瞻性、切实可行的运营方法，包括高效的开发方法、数据驱动的运营、AI辅助决策等，以支撑云际生态的构建演进。最后通过跨云迁移部署的案例来说明系统内在机制的科学依据以及支撑方法的有效性。