Open AccessResearch on Load Balancing of Cloud Storage Server Based on Fully Connected Group
Author(s) -
Xiaodan Chen,
Shuanglong Pang,
Haiyan Wang,
Desheng Zeng
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1982/1/012134
Subject(s) - server , bottleneck , computer science , round robin dns , cloud computing , upload , load balancing (electrical power) , scheduling (production processes) , distributed computing , volume (thermodynamics) , computer network , file server , operating system , response time , the internet , embedded system , engineering , quantum mechanics , grid , operations management , physics , geometry , mathematics , domain name system
The scheduling problem of online cloud storage traffic is a hot topic in current network research. Aiming at the scheduling problem of file upload requests in online cloud storage systems, the existing schemes rarely meet the requirement of providing users with different bandwidth guarantees. One of the purposes of using multiple volume servers is to let multiple volume servers share the system load evenly and avoid the system performance bottleneck. Once the load is aggregated and one of the multiple volume servers is overloaded, the volume server will become the system bottleneck again, which will greatly affect the overall performance of the system. On the basis of analyzing the existing scheduling algorithms of server clusters, this paper proposes a dormant scheduling algorithm of fully connected groups, which can automatically equip servers with I D, redirect tasks according to the actual processing capacity, and ensure load balance by unlimited increase. The experimental results show that the load balancing strategy among multi-volume servers ensures the overall reading and writing performance of the system, and effectively improves the I/O throughput and I/O times per second.