On runtime adaptive tile defragmentation for resource management in many-core systems

Before an application can be actually launched in a many-core system, the first thing that needs to be done is to get the application mapped to a number of tiles (cores). Such online application mapping process may unfortunately lead to a serious resource leak problem, referred as tile fragmentation, that free (uncommitted) tiles from any single contiguous region are just inadequate to accommodate the performance needs of an incoming application, although the total number of free tiles may still exceed what is required to service this application. When applications have to be mapped to noncontiguous tiles due to fragmentation, there will be obvious performance penalty due to increased communication distances. As a result, defragmentation that consolidates fragmented tiles needs to be routinely exercised, and this defragmentation process must not introduce high computation overhead that otherwise can adversely impact the system performance. In this paper, we propose a task migration-based adaptive tile defragmentation algorithm that helps consolidate running applications through online task migration. This algorithm relocates the applications’ tile regions so that a contiguous free tile region is formed and maintained. By doing so, future applications can be mapped to a region with low communication distance. Both the computation overhead and quality of defragmentation result of the proposed algorithm are adaptively set in response to the system workloads. Enabled by its low overhead, the proposed defragmentation algorithm is an effective resource management enhancement to the existing runtime task-to-tile mapping methods, with as much as 3× system throughput improvement observed in some experiments.     

软件研发中的精细化人力资源管理模型及系统

软件项目研发的成功与否,人力资源的管理起着决定性作用。从实践项目中提出了精细化人力资源管理的模型,将人力资源、软件项目、任务等抽象成具有状态、属性和操作的实体,通过定义任务效率、效率奇点等指标,综合SPI、CPI指标进行项目成本进度偏差分析,为人力资源的计划、调度、冲突处理等提供支撑。基于该模型进行平台研发,并通过在该平台中跟踪和分析真实的精细化人力资源管理案例,验证了模型的有效性。     

Toward a hybrid model for usability resource allocation in industrial software product development

The organizational aspects of user‐centered software development in a financial services company are presented. The financial services industry sector is one of the industrial sectors to embark on the development of computer software as a consumer product. The nature of business in the service sector predisposes it to encounter difficulties in developing software aimed at meeting customer demands. Lack of familiarity and experience with the product design and implementation processes, as well as reliance on usability for acceptance, are major obstacles encountered. Difficulties, insights, and lessons learned regarding organizational ergonomics issues faced by a user‐centered design group are provided, and a hybrid resource distribution model is proposed to guide other service sector companies in their future software development efforts. © 2007 Wiley Periodicals, Inc. Hum Factors Man 17: 245–262, 2007.     

CAP-W: Congestion-aware platform for wireless-based network-on-chip in many-core era

In order to fulfill the ever-increasing demand for high-speed and high-bandwidth, wireless-based MCSoC is presented based on a NoC communication infrastructure. Inspiring the separation between the communication and the computation demands as well as providing the flexible topology configurations, makes wireless-based NoC a promising future MCSoC architecture. However, congestion occurrence in wireless routers reduces the benefit of high-speed wireless links and significantly increases the network latency. Therefore, in this paper, a congestion-aware platform, named CAP-W, is introduced for wireless-based NoC in order to reduce congestion in the network and especially over wireless routers. The triple-layer platform of CAP-W is composed of mapping, migration, and routing layers. In order to minimize the congestion probability, the mapping layer is responsible for selecting the suitable free core as the first candidate, finding the suitable first task to be mapped onto the selected core, and allocating other tasks with respect to contiguity. Considering dynamic variation of application behaviors, the migration layer modifies the primary task mapping to improve congestion situation. Furthermore, the routing layer balances utilization of wired and wireless networks by separating short-distance and long-distance communications. Experimental results show meaningful gain in congestion control of wireless-based NoC compared to state-of-the-art works.     

Fuzzy resource space model and platform

The resource space model (RSM) is a model for organizing versatile resources in normal forms and providing uniform resource management operations. In applications, we find three important factors that influence the effectiveness of organizing and operating resources: natural semantics of resources, resource providers' beliefs, and resource users' beliefs. The relationship between the three factors influences the effectiveness of resource management operations. This paper proposes a fuzzy resource space model (FRSM), which consists of a fuzzy resource space and a fuzzy operation language expressing the resource providers' beliefs and the resource users' beliefs. By properly dealing with the two kinds of beliefs, the FRSM improves the RSM in effectively managing resources. The proposed model has been implemented in the Knowledge Grid platform VEGA-KG.     

Hammerstein-Wiener nonlinear model based predictive control for relative QoS performance and resource management of software systems

Runtime management of Quality of Service (QoS) performance and resource provisioning is a vital issue in shared resource software environments. A useful performance management technique for such software systems is the relative guarantee feedback control scheme. The existing approaches for this class of control systems are mainly based on off-line linear or on-line model identification and control techniques, which tend to have performance issues in the presence of nonlinearities induced by this scheme. Instead of using such modeling techniques, this paper proposes a new approach for QoS performance management and resource provisioning by using an off-line identification of Hammerstein and Wiener nonlinear block structural model. Using the characteristic structure of the nonlinear model, a predictive feedback controller based on a gain schedule technique is incorporated in the design to achieve the performance objectives. The proposed approach is validated using experiments based on a prototype, demonstrating superior runtime QoS performance management and resource provisioning in a complex software system.     

开源平台下软件管理技术的研究

基于开源软件平台基本特点的分析,对目前具有代表性的apt/dpkg、yum/RPM、BSD ports和portage等软件管理技术进行分析,对比和总结了现有技术特征和一般结构,探索了开源平台下软件管理技术可能的发展方向.     

计算资源共享平台中非集中式的元数据管理

提出了计算资源共享平台中的一种非集中式元数据管理方法,它利用对等网络的方式把数据和元数据分散到网络上的其他节点,而不是在后台服务器端。该方法可以在平台运行数据密集型应用时支持大容量的并行工作机读写数据,同时具有随机的访问、灵活的访问粒度、支持高负载的并行读写的特点。运用了分布式哈希表技术,把大容量元数据划分成树型结构的分段树的形式。描述了数据与元数据的读写与追加新数据的过程。测试结果表明：该方法针对3D图像渲染类数据密集型应用,在不同的工作机并行访问与读写过程中,可以获得很高的聚集带宽与平均读写带宽。     

高能物理云平台中的弹性计算资源管理机制

虚拟化技术作为一种新的资源管理技术,正在高能物理领域得到越来越广泛的应用。静态虚拟机集群方式已经逐渐不能满足多作业队列对于计算资源动态的需求。为此,实现了一种云计算环境下面向多作业队列的弹性计算资源管理系统。系统通过高吞吐量计算系统HTCondor运行计算作业,使用开源的云计算平台Openstack管理虚拟计算节点,给出了一种结合虚拟资源配额服务,基于双阈值的弹性资源管理算法,实现资源池整体伸缩,同时设计了二级缓冲池以提高伸缩效率。目前系统已部署在高能所公共服务云IHEPCloud上,实际运行结果表明,当计算资源需求变化时系统能够动态调整各队列虚拟计算节点数量,同时计算资源的CPU利用率相比传统的资源管理方式有显著的提高。     

Scaling for agility: A reference model for hybrid traditional-agile software development methodologies

The adoption of agility at a large scale often requires the integration of agile and non-agile development elements for architecting a hybrid adaptive methodology. The challenge is ”which elements or components (agile or non-agile) are relevant to develop the context-aware hybrid adaptive methodology reference architecture?” This paper addresses this important challenge and develops a hybrid adaptive methodology reference architecture model using a qualitative constructive empirical research approach. In this way, we have uncovered the agility, abstraction, business value, business policy, rules, legal, context and facility elements or components that have not been explicitly modelled or discussed in International Standards (IS) such as the ISO/IEC 24744 metamodel. It is anticipated that a context-aware hybrid adaptive methodology can be architected by using the proposed context-aware hybrid adaptive methodology reference architecture elements for a particular situation when using a situational method engineering approach.     

Joust: a platform for liquid software

The authors describe a Java-based platform for liquid software, called Joust, that is specifically designed to support low-level, communication-oriented systems and to avoid the limitations of general-purpose OSs. The authors contrast the platform requirements for communication-oriented liquid software with those of computation-oriented software, identify the limitations of current platforms, and outline the benefits of Joust. They also offer an overview of Scout (the underlying OS upon which Joust is built), its runtime system, and its just-in-time (JIT) compiler     

基于Web的软件测试过程管理平台的分析与设计

阐述了一个中小型领域软件企业软件测试过程管理平台分析和设计过程.该平台基于Web技术,采用B/S结构构建,在Intranet/Internet环境下使用.以注重和改善测试过程管理中测试计划与风险管理、测试评审与缺陷跟踪、用户反馈和测试总结4个环节为目标,设计了一个的测试业务流程,实现了系统登录、系统配置管理、测试计划管理、测试用例管理、测试日志管理、测试报告管理、测试汇总管理和测试文档管理等功能.     

面向虚拟路由器平台的资源管理平面的设计与实现

通过研究与分析虚拟路由器平台在管理与维护等方面存在的问题,提出了三层的虚拟路由器体系框架：控制平面、转发平面和资源管理平面。控制平面和转发平面是虚拟路由器的两个基本功能平面,分别承载逻辑控制平面和逻辑转发平面。为了动态地管理虚拟路由器平台,引入了资源管理平面。它是管理物理资源和维护路由器实例的重要功能平面。通过阐述资源管理平面的框架以及实现机制,实现了支持路由器实例的静态创建和动态调整的原型系统。实验结果表明：资源管理平面的最大处理能力为3205条命令/秒,在1 min内完成路由器实例的创建任务,能够满足虚拟路由器平台的管理需求。     

基于云计算的网络平台共享资源模型的建设

随着计算机技术和网络技术的发展,人们的学习模式也逐渐发生着改变,传统的学习方式、方法正慢慢改变。越来越多的人获得有价值信息的途径转向于媒体,特别是网络。本文首先介绍目前＂云计算＂的相关定义、特点及应用情况,其次介绍＂云计算＂模型在网络资源共享中的地位,最后构建基于云计算的网络平台共享资源模型。     

网格计算资源管理中的信誉度模型

网格资源管理作为网格研究的核心内容,受到研究人员的高度关注。目前已经提出的诸多管理模型中,基于计算经济的模型被人们普遍看好。文中针对计算经济模型,提出了资源信誉度、资源提供者信誉度、资源消费者信誉度的概念。在此基础上,深入研究了信誉度评价的评价模型和算法以及信誉度评价实现中的一些处理策略。对信誉度评价所需的资源使用记录进行了定义,根据评价需要对现有RUR进行了扩展。最后通过仿真实验对信誉度评价算法进行验证,得到了相应的实验结果和结论。     

A quantitative model for the support of software management

The software production and maintenance management support project (SPMMS) will define and produce a complete, consistent and efficient information system supporting all management activities in the life cycle of software.Adequate management of the production and maintenance of software can be achieved only with the support of quantitative analysis. This analysis must examine all attributes of the process by which software is developed and maintained. Initial results are available of an investigation into the use of quantitative models within an SPMMS.There are several issues and problems of project management. Existing models for cost and some approaches to the measurement of quality have disadvantages. An experience database helps estimating and planning and control. Criteria for model selection can be given (as a basis for assessing external and internal models). A CQA model is incorporated within SPMMS. This is part of the SPMMS approach to integrated and automatic management support.     

基于模型驱动的软件集成平台研究与实现

针对当前管理软件业务需求快速变化的特点,提出了"以用户为中心"的面向业务的软件集成平台.分析了当前管理软件开发遇到的问题,提出面向业务的软件集成平台(BOSIP)作为求解的基本思想;提出采用模型驱动作为BOSIP的实现技术体系,研究了模型驱动原理,给出了运行期模型驱动观点;根据运行期模型驱动思想,给出BOSIP实现框架;最后针对油气勘探业务领域实现了数据管理平台,对基于运行期模型驱动的BOSIP进行了验证.     

Knowledge boundaries in enterprise software platform development: Antecedents and consequences for platform governance

The widespread uptake of platform strategies turns many vendors of enterprise software into curators of an ecosystem of firms that collaboratively develop and commercialize a shared technology. As a platform owner's effectiveness in integrating knowledge across ecosystem participants will distinguish it from its competitors, we investigate the management of development‐related knowledge across firm boundaries. Our exploratory, multiple‐case study of 4 platforms illustrates how “knowledge boundaries” emerge between platform owners and complementors. We observe that knowledge boundaries are influenced by a platform's functional extent, interface design, and evolutionary dynamics, which create differences, dependencies, and novelty of development knowledge, resulting in qualitatively distinct types of knowledge boundaries. To overcome knowledge boundaries, platform owners provide various resources at the boundary, including information portals, documentation, helpdesks, and alignment workshops. We observe that in shaping these resources, platform owners face a trade‐off between providing knowledge at the right scope, while allowing for the scalability of knowledge resources for the entire ecosystem. Depending on their scope and scale, we classify knowledge boundary resources as broadcasting, brokering, and bridging, each representing qualitatively distinct patterns in managing knowledge in platform ecosystems. We conclude with implications for researchers and managers.