Open AccessA dynamic resource chain task unloading method based on improved greedy algorithm
Author(s) -
Bo Hu,
Xiaolei Tian,
Chao Yang,
Wei Jiang,
Dong Shi Dan Jie,
MingHui Chen,
Chen Wu,
Shu Chen,
Weng Zhangjun
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 - 10.1088/1742-6596/1883/1/012021
Subject(s) - computer science , greedy algorithm , graph , task (project management) , cloud computing , key (lock) , distributed computing , algorithm , theoretical computer science , engineering , computer security , systems engineering , operating system
the computing network based on cloud computing is becoming more and more complex. How to effectively unload the computing tasks is the key problem to improve the service ability of the computing network. In this paper, a mathematical model of task unloading problem in computing network is established, and an optimal solution based on improved greedy algorithm is proposed, which can quickly find the way of resource chain task dynamic unloading and realize task fast allocation and calculation. In this paper, the search space for greedy algorithms is too large and there is a lack of support for HPC LAN. Therefore, the greedy algorithm search is improved from one layer to two layers, which are respectively searched by the strongly connected graph and strongly connected graph of the virtual graph, which can adapt to the computing network with complex structure, and has friendly characteristics for high-performance computing LAN.