Author(s)

Ying Li, Mingchen Shao, Chaoying Gao

Corresponding Author

Ying Li

Abstract

The distributed task parallel scheduling problem of flash memory database is studied to improve the ability of optimal access to flash memory database and parallel scheduling of distributed tasks. The traditional method uses particle swarm optimization control algorithm for parallel scheduling of distributed tasks in flash database. The coupling characteristics of association rules of mass distributed task data stream in flash database cannot be effectively utilized, which leads to poor scheduling performance. In this paper, a distributed task parallel scheduling algorithm for flash database based on chaotic evolution feature clustering is proposed. The dual state stationary chaotic evolution feature clustering flash database scheduling model is designed, and the complex activation function is set. The layer excavates the frequent pattern set of distributed task parallel scheduling in flash database, extracts the characteristic of data information flow, and designs the transfer balance operator of distributed task parallel scheduling in flash database. Distributed task parallel scheduling of distributed task flash memory database is carried out according to real part and virtual part path, so as to improve database access ability. The simulation results show that the algorithm can effectively improve the throughput and scheduling success rate of the distributed task parallel scheduling of flash database, and the distributed task parallel scheduling of flash database is universal.

Keywords

Flash database, distributed tasks, parallel scheduling, throughput