The rapid development of satellite communication technologies has significantly enhanced global connectivity, with Low Earth Orbit (LEO) satellite networks playing a crucial role. However, the high-speed movement of LEO satellites introduces dynamic and unstable network topologies, presenting challenges in routing and traffic engineering. Traditional routing algorithms, such as Dijkstra’s, often struggle to adapt to these dynamic conditions, leading to inefficiencies in load balancing and resource utilization. In response, this paper proposes an Iterative Region-Based Probabilistic Forwarding (IRPF) strategy within a Software-Defined Satellite Networking (SDSN) framework. Our strategy dynamically assigns forwarding probabilities to satellite ports based on link traffic weights, optimizing routing paths through a feedback iteration process. To prevent the algorithm from converging to local optima, we introduce a local adjustment mechanism that fine-tunes forwarding probabilities. The experimental results show that when the scale of the satellite network is relatively small, the IRPF strategy incurs lower total link traffic costs compared to the Dijkstra algorithm. Additionally, it improves the Gini coefficient and packet loss rate by approximately 20% and 30%, respectively. These results demonstrate the effectiveness and stability of this method in routing and traffic engineering.

With the continued migration of storage to cloud database systems, the impact of slow queries in such systems on services and user experience is increasing. Root-cause diagnosis plays an indispensable role in facilitating slow-query detection and revision. This paper proposes a method capable of both identifying possible root cause types for slow queries and ranking these according to their potential for accelerating slow queries. This enables prioritizing root causes with the highest impact, in turn improving slow-query revision effectiveness. To enable more accurate and detailed diagnoses, we propose the multimodal Ranking for the Root Causes of slow queries (RCRank) framework, which formulates root cause analysis as a multimodal machine learning problem and leverages multimodal information from query statements, execution plans, execution logs, and key performance indicators. To obtain expressive embeddings from its heterogeneous multimodal input, RCRank integrates self-supervised pre-training that enhances cross-modal alignment and task relevance. Next, the framework integrates root-cause-adaptive cross Transformers that enable adaptive fusion of multimodal features with varying characteristics. Finally, the framework offers a unified model that features an impact-aware training objective for identifying and ranking root causes. We report on experiments on real and synthetic datasets, finding that RCRank is capable of consistently outperforming the state-of-the-art methods at root cause identification and ranking according to a range of metrics.