A5082859077- Data Stream Mining Techniques
- Cloud Computing and Resource Management
- Software System Performance and Reliability
- Software-Defined Networks and 5G
- High Entropy Alloys Studies
- High-Temperature Coating Behaviors
- Additive Manufacturing Materials and Processes
- Particle physics theoretical and experimental studies
- Dark Matter and Cosmic Phenomena
- Anomaly Detection Techniques and Applications
- Big Data and Business Intelligence
- Particle Detector Development and Performance
- Advanced materials and composites
- Tuberculosis Research and Epidemiology
- Caching and Content Delivery
- Infectious Diseases and Tuberculosis
- Time Series Analysis and Forecasting
- Diagnosis and treatment of tuberculosis
- Network Packet Processing and Optimization
Zhejiang Center for Disease Control and Prevention
2025
Shenyang Ligong University
2023
Xi’an University
2023
Xi'an University of Technology
2023
Objective To identify the composition of comorbidities among patients with newly diagnosed pulmonary tuberculosis and assess impact on clinical characteristics patients. Methods This study was conducted in 13 hospitals across counties Zhejiang province, China. Patient data collected this included demographic characteristics, chest radiography results, etiological comorbidities. Descriptive statistics were to describe all participants. Univariate multivariate logistic regression analyzes...
Open vSwitch (OvS) has been widely used in cloud networks view of its programmability and flexibility. However, we observe a huge performance drop when it loads practical networking services (e.g., tunneling firewalling). Our further analysis reveals that the root cause lies gap between needs supporting various selections packet header fields one-size-fits-all data path vanilla OvS. Motivated by this, design Hoda, high-performance OvS dataplane with multiple specialized paths. Specifically,...
Service meshes play a central role in the modern application ecosystem by providing an easy and flexible way to connect microservices of distributed application. However, because how they interpose on traffic, can substantially increase latency its resource consumption. We develop tool called MeshInsight help developers quantify overhead service deployment scenarios interest make informed trade-offs about their functionality vs. overhead. Using MeshInsight, we confirm that have high...
This study demonstrates the salient facts and challenges of host failure operations in hyperscale data centers. A incident can involve hundreds distinct host-level metrics. The faulting mechanism inside connects these heterogeneous metrics through direct indirect correlation, making it extremely difficult to sort out propagation procedures root cause from intertwined indicators. To deeply understand host, we develop HEAL -- a novel analysis toolkit. discovers dynamic causality sparse by...
This study demonstrates the salient facts and challenges of host failure operations in hyperscale data centers. A incident can involve hundreds distinct host-level metrics. The faulting mechanism inside connects these heterogeneous metrics through direct indirect correlation, making it extremely difficult to sort out propagation procedures root cause from intertwined indicators. To deeply understand host, we develop HEAL -- a novel analysis toolkit. discovers dynamic causality sparse by...
Operating and maintaining hyperscale data centers involving millions of service hosts has been an extremely intricate task to tackle for top Internet companies. Incessant system failures cost operators countless hours browsing through performance metrics diagnose the underlying root cause prevent recurrence. Although many state-of-the-art (SOTA) methods have used time-series causal discovery construct relationships among anomalous metrics, they only focus on homogeneous service-level fail...
Abstract High entropy alloy is considered a kind of structural and functional materials with wide application prospects. In the present work, FeCrCoNiMo x was prepared used as research object. The influence Mo large atomic radius on crystal structure, microstructure, mechanical properties discussed variation microstructure composition clarified. changes gradually from single phase FCC centered cubic solid solution structure to + σ μ increasing content. also uniaxial dendrite hardness change...
This study demonstrates the salient facts and challenges of host failure operations in hyperscale data centers. A incident can involve hundreds distinct host-level metrics, covering broad aspects. The faulting mechanism inside connects these heterogeneous metrics through direct indirect correlation, making it extremely difficult to sort out propagation procedures root cause from intertwined indicators. To deeply understand host, we develop HEAL -- a novel analysis toolkit. synergistically...