Fault-Tolerant Communication Mechanisms in Data Center Network of HS-DCell

In the era of rapid technological advancements and widespread big data applications, data center networks serve as critical enablers for seamless communication and efficient data processing in large-scale computing ecosystems. Network reliability is fundamental to o ensure stable, high-performance operations. Traditional connectivity metrics assume simultaneous failures of all neighboring servers, which is unrealistic in complex network topologies and leads to inaccurate fault-tolerance assessments. To address this, [Formula: see text]-extra connectivity was proposed to better evaluate network reliability. [Formula: see text]-[Formula: see text], a novel server-centric network built upon DCell and BCCC, exhibits remarkable scalability with a linearly growing diameter as network levels increase. This paper investigates the [Formula: see text]-[Formula: see text] network, which leverages 3-port servers to reduce port counts, alleviate traffic congestion, and minimize latency, thereby offering robust scalability and fault tolerance. We derive the formula for the [Formula: see text]-extra connectivity of [Formula: see text]-[Formula: see text]. Unlike traditional metrics, which only consider disconnecting the network with the minimum number of nodes h-extra connectivity depends on the network’s hierarchical structure and server interconnections, ensuring each connected component contains at least [Formula: see text] nodes. This provides a more accurate measure of network robustness. The findings offer insights for designing resilient DCNs with enhanced fault-tolerance capabilities.