Research on Data Producing Services across Compute Engines

Abstract

Currently, data production services are facing the following challenges: achieving cross-computing engine scheduling, automated accurate labeling, and on-demand task scheduling. In this paper, cross-computing engine scheduling technology for business data production, domain-oriented business data automatic labeling technology and business data production visualization process scheduling technology are investigated, while high concurrency-related technical solutions are studied to improve the service processing performance. At the end of the paper, technical risk assessment analysis is also given in four aspects, including computing resource scheduling, labeling reliability, process scheduling optimization and thread safety.