We use cookies to improve your experience with our site.
Ni An, Meng-Jie Lyu, Wei-Bei Fan, Lei Han, Fu Xiao. Fault-tolerant Routing Under Conditional Faults Pattern in Data Center Network of RRect[J]. Journal of Computer Science and Technology. DOI: 10.1007/s11390-025-4126-8
Citation: Ni An, Meng-Jie Lyu, Wei-Bei Fan, Lei Han, Fu Xiao. Fault-tolerant Routing Under Conditional Faults Pattern in Data Center Network of RRect[J]. Journal of Computer Science and Technology. DOI: 10.1007/s11390-025-4126-8

Fault-tolerant Routing Under Conditional Faults Pattern in Data Center Network of RRect

  • With the expansion and increasing complexity of data center networks (DCNs), network fault-tolerance has become increasingly important. RRect was a server-centered DCN with a good interconnection structure. In this paper, we propose a fault-tolerant routing algorithm RRF P under a conditional fault pattern of RRect, which can find a fault-free path between any two fault-free vertices. Firstly, we study a fault pattern of RRect in the case of restricted faulty vertex sets, 1, 2-restricted connectivity. It is about 2, 3 times of RRect’s traditional connectivity, indicating that 1, 2-restricted connectivity better evaluate the fault-tolerant capability. Secondly, we design an effective fault-tolerant routing algorithm RRFP under the conditional fault pattern of RRect, and RRF P can accommodate more faulty vertices. Finally, we conduct experiments on RRF P to evaluate its performance. The experimental results show that in terms of running time, as parameters n and k change, RRFP performs better than Dijkstra with a 64% to 84% improvement and better than BFS with a 56 % to 71% improvement. The running time of all three algorithms is very short as parameter m changes. In terms of constructing path length, as parameters change, RRFP performs better than BFS, with a 12% to 19% improvement. Moreover, the path constructed by RRF P is approaching the shortest path of Dijkstra. Moreover, RRF P still maintains a low TFR and high reliability even with an increase in the number of fault elements.
  • loading

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return