Data from RACE: Resource Aware Cost-Efficient Scheduler for Cloud Fog Environment

The quality of human life increases with increased usage of the Internet of Things (IoT). However, this massive use of IoT produces a large amount of data which creates a problem for data storage and analysis. Cloud computing is used to solve such issues like storage and processing requirements for data generated by IoT. However, several applications like traffic control, health observation, and games, etc., are time-sensitive and need a quick response. The delay created by sending data to the Cloud and then returning the Server response to the user of such programs has an unintended impact. To overwhelmed these limitations, the Fog computing concept was launched in 2012. Fog computing is an elongation of Cloud computing in which user services are extended to networking devices. In recent times the Fog computing is considered as a favorable approach to cope with the delay-sensitive applications. Moreover, Fog computing in combination with the Cloud computing model can also provide promising solutions to deal with compute-intensive applications. The scheduling challenges faced by the Fog service provider are to fulfill user requests with the maximum resource utilization of Fog devices while reducing the monetary cost of Cloud resource usage. In this research, a resource-aware scheduler has been proposed to distribute the incoming application modules to Fog devices that maximize resource utilization at the Fog layer, reduces the monetary cost of using Cloud resources with minimum execution time of applications, and minimum bandwidth usage. The proposed scheduler Resource Aware Cost-Efficient Scheduler (RACE) consists of two algorithms: ModuleScheduler: which categorizes the incoming application modules according to their computation and bandwidth requirements. The ModuleScheduler also creates a prioritized list considering the categories of application modules. Compare Module: Finds the suitable Fog device for modules execution in the prioritized list. The proposed scheduler (RACE) schedule the modules at the Fog layer based on their computation requirement and bandwidth requirement until the devices at the Fog layer have enough CPU capacity to accommodate the modules.人类生活质量随着物联网（Internet of Things, IoT）应用规模的扩大而提升。然而，物联网的大规模部署会产生海量数据，为数据存储与分析带来严峻挑战。云计算常被用于解决物联网所产生数据的存储与处理需求相关问题。但诸如交通管控、健康监测、游戏等诸多应用均具有时间敏感性，需要快速响应。将数据传输至云端再将服务器响应回传给此类应用的用户所产生的延迟，会带来意想不到的负面影响。 为克服上述局限，雾计算（Fog Computing）理念于2012年被提出。雾计算是云计算的延伸，其将用户服务拓展至网络设备当中。近年来，雾计算被视为应对时延敏感型应用的优选方案。此外，雾计算与云计算模型相结合，还可为计算密集型应用提供颇具前景的解决方案。 雾计算服务提供商所面临的调度挑战，是在最大化雾设备资源利用率的前提下满足用户请求，同时降低云端资源使用的经济成本。本研究提出一种资源感知调度器，用于将待处理的应用模块分配至雾设备，以实现雾层资源利用率最大化、应用执行时长最小化、带宽使用量最小化，并降低云端资源使用的经济成本。 所提出的资源感知成本高效调度器（Resource Aware Cost-Efficient Scheduler, RACE）包含两种算法： 1. 模块调度器（ModuleScheduler）：根据应用模块的计算与带宽需求对其进行分类，并结合模块类别生成优先级列表； 2. 匹配模块（CompareModule）：为优先级列表中的待执行应用模块匹配合适的雾设备。 RACE调度器将依据应用模块的计算与带宽需求，将其调度至雾层设备，直至雾层设备的CPU容量不足以容纳更多待调度模块为止。