Realization of Flexible Scaling Technology for Computing Tasks in Power Automation System Based on Container

With the popularization and construction of the new power system, the amount of data to be processed by the power automation system will increase explosively, and the original mode of main and standby machines will no longer meet the demand. At present, some dynamic scheduling algorithms studied in the industry do not consider the security, reliability and real-time requirements of power system. By studying the characteristics of some existing scheduling algorithms and the implementation principle of container technology, this paper proposes a container based elastic scaling technology for computing tasks, which can dynamically adjust the resources of each application instance without affecting the resource use of other application instances. At the same time, it supports highly reliable mutual redundancy and real-time switching, and supports the real-time elastic expansion of resources after adding new nodes. Finally, through the test and verification, the technology can support the dynamic allocation and distributed operation of power application resources, switch the active and standby computers in real time in case of failure, support the elastic expansion of resources, and improve the reliability and processing capacity of the upper application.