With the rapid advancement of big data processing and artificial intelligence (AI), data center networks (DCNs) must deliver more efficient resource management and data transmission mechanisms. Unfortunately, due to the significant differences in bandwidth requirements, transmission patterns, and temporal characteristics across various traffic types in DCNs (such as short flows, long flows, and bursty flows), traditional buffer allocation strategies fail to adapt flexibly to these disparities. In this paper, we propose Enhanced Buffer Management (EBM), a novel buffer-sharing scheme designed for scenarios that require higher performance from DCNs. Unlike prior approaches, EBM employs a multi-level flow identification and adaptive threshold adjustment mechanism to enhance the flexibility and efficiency of buffer management under varying traffic conditions. Specifically, EBM first performs coarse-grained and fine-grained classification of traffic based on packet size, inter-arrival interval, and other flow characteristics. It then applies an improved threshold computation function to allocate buffer space differentially across traffic classes while maintaining allocation smoothness. Our evaluation results demonstrate that EBM significantly improves performance under realistic workloads. For instance, it reduces the 99th percentile Flow Completion Time (FCT) slowdown by 32.7% for short flows in the web-search workload and by 45.1% for incast flows in the hadoop workload, all without sacrificing overall throughput.

Summary With the continuous expansion scale of data center networks (DCNs), the probability of network failures becomes high. Trustworthy fault‐tolerant routing is extremely significant for reliable communication in data centers. In this article, we tackle the challenge by proposing a novel fault‐tolerant routing scheme for a torus‐based DCN. First, we present a multipath information transmission model based on the trust degree of reachable paths and propose a novel Hamiltonian odd–even turning model without deadlock. Second, we design an efficient deadlock‐free fault‐routing algorithm by constructing the longest fault‐free path between any two fault‐free nodes in DCN. Extensive simulation results show that the proposed fault‐tolerant routing outperforms the previous algorithms. Compared with the most advanced fault‐tolerant routing algorithms, the proposed algorithm has a 21.5% to 25.3% increase in throughput and packet arrival rate. Moreover, it can reduce the average delay of 18.6% and the maximum delay of 23.7% in the network respectively.