By decoupling services and enabling elastic deployment, microservice architecture improves system scalability and evolutionary capability. At the same time, it substantially increases operational complexity. Failures often exhibit cross service propagation and a mismatch between observed symptoms and underlying root causes. To address the heterogeneity and fragmentation of multi source observability data such as logs, metrics, and distributed traces, this study proposes a unified modeling and intelligent root cause localization method for microservice systems. The approach treats each service as a basic modeling unit and maps heterogeneous observations into a shared representation space. Service dependency structure is explicitly incorporated to characterize system state at a global level. Through structure aware modeling on the dependency graph, anomaly information is propagated and constrained along real invocation relations. This design enables more accurate separation of local disturbances from structural anomalies. In addition, a consistency based measure derived from state deviation is constructed to score service anomalies. Dependency relations are then used for attribution and ranking, which unifies root cause localization and impact analysis within a single framework. Comparative results show that the proposed method achieves more stable and consistent advantages across multiple evaluation metrics. It captures anomaly propagation patterns in microservice systems more effectively and provides a unified and structure aware solution for intelligent diagnosis of complex distributed systems.

This paper addresses the challenges of evidence dispersion, attention shift, and inference inconsistency in long text and multi-document reasoning scenarios. It investigates the problems that large language models often encounter under conditions of extremely long contexts, multi-source information conflict, and structural complexity, and proposes a hierarchical, curriculum-based fine-tuning algorithm framework. This framework organizes the input into a hierarchical structure of questions and multi-document contexts. At the representation level, it constructs a three-level convergence path of tokens, fragments, and documents to form structured memory and mitigate semantic drift caused by context expansion. At the reasoning level, it introduces a question-guided evidence scoring and weight aggregation mechanism to achieve differentiable selection across document fragments and global evidence vector construction, thereby strengthening the alignment of key evidence and suppressing redundant interference. At the training organization level, it employs a curriculum strategy, progressively scheduling samples according to difficulty levels, enabling the model's capabilities to gradually transition from local consistency to cross-document evidence integration and overall induction. Comparative experimental results show that this method exhibits more stable performance in multi-document evidence localization and inference consistency evaluation, validating the effectiveness of hierarchical modeling and curriculum scheduling in shaping multi-document reasoning capabilities.

Large language model (LLM) inference in multi-tenant clouds is becoming an increasingly important contributor to data-center carbon emissions, yet existing carbon-aware scheduling techniques target long-running training jobs and are ill-suited for the short, bursty, SLO-sensitive nature of online serving. We propose CAPS (Carbon–Aware Prompt Scheduling), an online bi-objective scheduler that jointly optimizes goodput and per-request carbon cost for multi-tenant LLM inference. CAPS first employs a lightweight prompt complexity predictor to estimate token generation cost and latency risk for each incoming request. It then combines real-time grid carbon intensity, GPU energy profiles, and per-tenant SLO tiers to route each request to one of three execution pools: a low-latency pool, a low-carbon pool, or a delay-tolerant batch pool. A composite reward function balances goodput, carbon emissions, and SLO violation rate. In trace-driven simulations using public conversation traces and regional carbon intensity data, CAPS reduces average carbon emissions per 1K generated tokens by 26.8% compared to round-robin scheduling while achieving an SLO attainment rate that matches or exceeds a dedicated SLO-aware baseline.

Identifying abnormal behaviors plays a vital role in ensuring cloud infrastructure remains secure and operationally stable. Conventional methods that aggregate data at a single location create substantial privacy concerns, especially within shared cloud platforms hosting multiple organizations with confidential operational information. This paper proposes HierFedDP, a hierarchical federated learning framework integrated with a two-stage differential privacy mechanism for privacy-preserving anomaly detection in cloud services. Our approach employs a three-tier architecture consisting of local clients, edge servers, and a central cloud server, where clients apply local differential privacy to their updates before transmission. We introduce an edge aggregation frequency parameter that enables edge servers to perform multiple local aggregation rounds before communicating with the central cloud. Experiments on the CICIDS2017 dataset demonstrate that HierFedDP achieves detection performance comparable to standard local differential privacy approaches while reducing wide-area network (WAN) communication overhead by 49%. This significant communication reduction, achieved without sacrificing privacy guarantees or detection accuracy, makes HierFedDP particularly suitable for bandwidth-constrained, geo-distributed cloud deployments.