No abstract

In cloud service delivery, the cloud user is concerned about “what” function and performance the cloud service could provide, while the cloud provider is concerned about “how” to design proper underlying cloud resources to meet the cloud user’s requirements. We take the cloud user’s requirement as intent and aim to translate the intent autonomously into cloud resource decisions. In recent years, intent-driven management has been a widely spread management concept which aims to close the gap between the operator’s high-level requirements and the underlying infrastructure configuration complexity. Intent-driven management has drawn attention from telecommunication industries, standards organizations, the open source software community and academic research. There are various application of intent-driven management which are being studied and implemented, including intent-driven Software Defined Network (SDN), intent-driven wireless network configuration, etc. However, application of intent-driven management into the cloud domain, especially the translation of cloud performance-related intent into the amount of cloud resource, has not been addressed by existing studies. In this work, we have proposed an Intent-based Cloud Service Management (ICSM) framework, and focused on realizing the RDF (Resource Design Function) to translate cloud performance-related intent into concrete cloud computation resource amount settings that are able  to meet the intended performance requirement. Furthermore, we have also proposed an intent breach prevention mechanism, P-mode, which is essential for commercial cloud service delivery. We have validated the proposals in a sensor-cloud system, designed to meet the user’s intent to process a large quantity of images collected by the sensors in a restricted time interval. The validation results show that the framework achieved 88.93 ~ 90.63% precision for performance inference, and exceeds the conventional resource approach in the aspects of human cost, time cost and design results. Furthermore, the intent breach prevention mechanism P-mode significantly reduced the breach risk, at the same time keeping a high level of precision.

No abstract

No abstract

With the increasing maturity of artificial intelligence (AI) technology, business automation technology has also become a trend. Particularly, network operation and maintenance (O&M) is expected to soon become automated and more efficient. However, the automation of O&M is hindered by the lack of network failure data and the cost of collecting data. We thus propose an approach to build a low-cost environment that can produce the same data as the actual production environment and use tools such as chaos engineering to generate training models for fault data. This paper attempts to build the underlying physical network layer using a low-cost single-board computer Raspberry Pi instead of an expensive PC server, while keeping the virtual network layer the same and performing fault simulation, data collection, and AI model training on the constructed virtual network layer. A comparison of the accuracy of the trained AI models verifies the feasibility of replacing the traditional PC server with an inexpensive Raspberry Pi device while keeping the structure and services of the virtual network layer unchanged. Also, a brief comparison with existing techniques is discussed. Our proposed approach solves the problem of insufficient data for AI training while reducing cost and risk.