Study on the Key Technology for Establishing a Cloud Platform-Oriented Digital Oilfield Based on High-performance Computing

This study aims to overcome the principal problems in the current information-support–platformoriented digital oilfield. These problems include a low utilization efficiency of the software and hardware resources, the high maintenance cost of the system, and limited automatic control. Starting with the existing software and hardware resources and with the current pattern of use of cloud computing, this study designs a cloud platform-oriented digital oilfield, based on highperformance computing (HPC). It also comprehensively describes the architecture of the cloud platform and then solves three key problems, namely, united resource management (computing and license resources ), policybased resource scheduling, and self-service remote cluster visualization. Finally, this paper describes the achievement of establishing and operating a new type of cloud platform-oriented digital oilfield, based on HPC with self-service intelligent cluster development, mainstream application system integration, unified job management, comprehensive real-time monitoring of resources, and statistical statement analysis. The results show that integrating cloud computing into HPC can comprehensively improve the efficiency of use of the hardware and software resources, and reduce system operation and maintenance costs, as well as providing administrators with a friendly and fast way to use and monitor those resources. Subject Categories and Descriptors I.2.10 [Vision and Scene Understanding]: High-performance cloud platform research; I.4.10 [Image Representation] General Terms : High-performance Computing, Cloud Platform