A5111930131- Cloud Computing and Resource Management
- IoT and Edge/Fog Computing
- Software-Defined Networks and 5G
- Caching and Content Delivery
- Distributed and Parallel Computing Systems
- Peer-to-Peer Network Technologies
- Metallurgical Processes and Thermodynamics
- Anomaly Detection Techniques and Applications
- Materials Engineering and Processing
- Metal Extraction and Bioleaching
- Advanced Technologies in Various Fields
- Parallel Computing and Optimization Techniques
- Distributed systems and fault tolerance
- Green IT and Sustainability
- Advanced Malware Detection Techniques
- Graph Labeling and Dimension Problems
- Graph Theory and Algorithms
- Advanced Data Storage Technologies
- Interconnection Networks and Systems
- Optimization and Search Problems
- Digital Media and Visual Art
- Web Data Mining and Analysis
- 3D Shape Modeling and Analysis
- Computer Graphics and Visualization Techniques
- Advanced Optical Network Technologies
Communication University of China
2024
Northwest Institute of Mechanical and Electrical Engineering
2024
Xi'an Jiaotong University
2024
Science and Technology Department of Sichuan Province
2023
Xihua University
2023
Kunming University of Science and Technology
2022
Shanghai Key Laboratory of Trustworthy Computing
2022
East China Normal University
2022
University of Toronto
2016-2022
Alibaba Group (United States)
2022
Mobile cloud computing, with its promise to meet the urgent need for richer applications and services of resource-constrained mobile devices, is emerging as a new computing paradigm has recently attracted significant attention. However, there no clear definition well defined scope due commercial hype, diverse ways combining applications. This article makes first attempt present survey from perspective intended usages. Specifically, we introduce three common architectures classify...
With the increasing prevalence of large scale cloud computing environments, how to place requested applications into available servers regarding energy consumption has become an essential research problem, but existing application placement approaches are still not effective for live with dynamic characters. In this paper, we proposed a novel approach named EnaCloud, which enables dynamically consideration efficiency in platform. use Virtual Machine encapsulate application, supports...
Large-scale datacenters have been widely used to host cloud services, which are typically allocated different virtual machines (VMs) through resource multiplexing across shared physical servers. Although recent studies primarily focused on harnessing live migration of VMs achieve load balancing and power saving among servers, there has little attention the incurred performance interference cost both source destination servers during after such VM migration. To avoid potential violations...
In this paper, we present an analytical framework for characterizing and optimizing the power-performance tradeoff in Software-as-a-Service (SaaS) cloud platforms. Our objectives are two-folded: 1) We maximize operating revenue when serving heterogeneous SaaS applications with unpredictable user requests. 2) minimize power consumption processing To achieve these objectives, construct a unified profit-maximizing objective to jointly consider cost economic view. An offline solution supreme...
In this paper, we present an analytical framework for characterizing and optimizing the power-performance tradeoff in Software-as-a-Service (SaaS) cloud platforms. Our objectives are two-fold: (1) We maximize operating profit when serving heterogeneous SaaS applications with unpredictable user requests, (2) minimize power consumption processing requests. To achieve these objectives, take advantage of Lyapunov Optimization techniques to design analyze optimal control make online decisions on...
With rapid development in online shopping, e-commerce websites are facing intensive user requests from an increasing number of customers. Especially promotion seasons, these may encounter flash crowds which pull heavy pressure o private infrastructure and even make he website unavailable. Such severe can be addressed by leveraging hybrid cloud solution, relieves workloads the offloading excessive to IaaS public cloud. However, bursty fluctuation bring challenges distributing with targest...
It is typical for cloud providers to operate a number of geographically distributed data centers, where inter-data center traffic constitutes large portion provider's demand over the Internet. Though such incurs substantial operational costs that are charged by ISPs, it varies significantly across different overlay links, raising opportunity optimize both routing and scheduling minimize costs. As being transmitted multiple sources their respective destinations, an intermediate also able...
The collective Unmanned Weapon System-of-Systems (UWSOS) network represents a fundamental element in modern warfare, characterized by diverse array of unmanned combat platforms interconnected through heterogeneous architectures. Despite its strategic importance, the UWSOS is highly susceptible to hostile infiltrations, which significantly impede battlefield recovery capabilities. Existing methods enhance resilience predominantly focus on basic graph relationships, neglecting crucial...
Online cloud services are widely deployed as Long-Running Applications (LRAs) hosted in containers. Placing LRA containers turns out to be particularly challenging due the complex interference between co-located and operation constraints production clusters such fault tolerance, disaster avoidance incremental deployment. Existing schedulers typically provide APIs for operators manually specify container scheduling requirements offer only qualitative guidelines placement. Such schedulers, do...
Fair bandwidth allocation in datacenter networks has been a focus of research recently, yet this not received adequate attention the context private cloud, where link is often shared among applications running data parallel frameworks, such as MapReduce. In paper, we introduce rigorous definition performance-centric fairness, with guiding principle that performance should be proportional to their weights. We investigate problem maximizing application while maintaining strict fairness and...
In multi-tenant data centers, each tenant desires reassuring predictability from the virtual network fabric - bandwidth guarantee, work conservation, and bounded tail latency. Achieving these goals simultaneously relies on rapid precise traffic admission. However, slow convergence (tens of milliseconds) prior works can hardly satisfy increasingly rigorous performance demand under dynamic patterns. Further, state-of-the-art load balance schemes are all guarantee-agnostic bring great risks...
The cleaning of copper slag is an important way to realize the sustainable development metallurgy industry. MgO a commonly used additive in process and has impact on physicochemical properties slag. However, there are few studies mechanism phase transformation ISASMELT with different amounts MgO. In this study, effect mass action concentration component, loss were elucidated. results showed that when was added slag, high melting point hortonolite formed When amount more than 4%, pleonaste...
With the ability to simplify code deployment with one-click upload and lightweight execution, serverless computing has emerged as a promising paradigm increasing popularity. However, there remain open challenges when adapting data-intensive analytics applications context, in which users of encounter difficulty coordinating computation across different stages provisioning resources large configuration space. This paper presents our design implementation Astra, configures orchestrates jobs an...
Tasks in a data-parallel job communicate with each other through number of concurrent flows, which is described as coflow. These flows are correlated the sense that performance coflow dictated by flow takes longest time to complete. Minimizing completion times, however, turns out be challenge, given correlation across and how they routed collectively datacenter network. In this paper, we propose Tailor, simple yet effective mechanism objective trimming times To achieve our objective, Tailor...
In production datacenters operated by Web service providers such as Google, multiple data parallel applications, MapReduce, are employed to facilitate processing at a large scale, with strong demand for intra-datacenter bandwidth in their communication stages. A noteworthy phenomenon these applications is the presence of barriers, which implies that job will not finish until last task completes. Existing flow-level sharing datacenter networks designed and optimized meet application-level...
Link utilization has received extensive attention since data centers become the most pervasive platform for data-parallel applications. A specific job of such applications involves communication among multiple machines. The recently proposed coflow abstraction depicts through a group parallel flows, and captures application performance corresponding requirements. Existing techniques to improve link utilization, however, either restrict themselves achieving work conservation, or merely focus...
We present Lynn, a multi-dimensional dynamic resource management system for distributed applications in clouds. Lynn is the first to uniformly manage full kinds of resources, including CPU, memory and particularly, network I/O bandwidth as whole schedule tasks. This comprehensive policy allows more fine-grained control over hardware thus efficiency. To support today's sophisticated clouds ease burden software developers, defines concept virtual cluster this case provides transparent...
Highly-parallel, high-performance scientific applications must maximize performance inside of a power envelope while maintaining scalability. Emergent parallel and distributed systems offer growing number operating modes that provide unprecedented control processor speed, memory latency, bandwidth. Optimizing these for requires an understanding the combined effects thread concurrency on execution time. In this paper, we describe how analytical model separates pure computation time (C) stall...
In the context of private datacenters that are operated by Web service providers such as Google, multiple applications using data parallel frameworks, MapReduce, coexist and share a limited supply link bandwidth capacities. It has been shown failures norm, rather than exception, in datacenters, will negatively affect performance failed tasks need to be relaunched placed on newly selected servers. this paper, we argue even with presence failures, should allocated competing performance-centric...