检索结果-内蒙古大学图书馆

19th European Conference on Computer systems, EuroSys 2024

作者： Zhao, Ziming Wu, Mingyu Chen, Haibo Zang, Binyu Institute of Parallel and Distributed Systems Seiee Shanghai Jiao Tong University Shanghai China Institute of Parallel and Distributed Systems Seiee Shanghai Jiao Tong University Engineering Research Center for Domain-specific Operating Systems Ministry of Education China Shanghai China

ISBN: (纸本)9798400704376

FaaS (function-as-a-service) is becoming a popular workload in cloud environments due to its virtues such as auto-scaling and pay-as-you-go. High-level languages like JavaScript and Java are commonly used in FaaS for programmability, but their managed runtimes complicate memory management in the cloud. This paper first observes the issue of frozen garbage, which is caused by freezing cached function instances where their threads have been paused but the unused memory (e.g., garbage) is not reclaimed due to the semantic gap between FaaS and the managed runtime. This paper presents the first characterization of the negative effects induced by frozen garbage with various functions, which uncovers that it can occupy more than half of FaaS instances' memory resources on average. To this end, this paper proposes Desiccant, a freeze-aware memory manager for managed workloads in FaaS, which reclaims idle memory resources consumed by frozen garbage from managed runtime instances and thus notably improves memory efficiency. The evaluation on various FaaS workloads shows that Desiccant can reduce FaaS functions' peak memory consumption by up to 6.72×. Such saved memory consumption allows caching more FaaS instances to reduce the frequency of cold boots (creating instances before function execution) and p99 latency by up to 4.49× and 37.5%, respectively. © 2024 ACM.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Controlling sensors and actuators collectively using the COCOS-framework

International Conference on Mobile Computing and Networking,...

引用

International Conference on Mobile Computing and Networking, MobiCom'07 Co-Located Workshops - Proceedings of the First ACM Workshop on Sensor and Actor Networks, SANET'07 2007年 53-54页

作者： Krüger, Maik Karnapke, Reinhardt Nolte, Jörg Distributed Systems/Operating Systems group BTU Cottbus Cottbus Germany

ISBN: (纸本)9781595937353

Cocos (COordinated COmmunicating Sensors)1 is a lean middleware platform for wireless sensor and actuator networks. The major programming abstractions of Cocos are distributed sensor spaces, which enable the collective controlling of sensors or actuators on remote nodes. Thus, Cocosbrings high-level data-parallel programming conceptssuch as global reductions into the world of small networked systems. In this paper we discuss the design rationale of Cocos as well as its in-network processing and aggregation capabilities.

关键词： Wireless sensor networks

来源：评论

学校读者我要写书评

暂无评论

Eliminating single points of failure in software-based redundancy

Eliminating single points of failure in software-based redun...

引用

9th European Dependable Computing Conference, EDCC 2012

作者： Ulbrich, Peter Hoffmann, Martin Kapitza, Rüdiger Lohmann, Daniel Schröder-Preikschat, Wolfgang Schmid, Reiner Department of Distributed Systems and Operating Systems Friedrich-Alexander University Erlangen-Nuremberg Erlangen Germany System Architecture and Platforms Corporate Technology Siemens AG Munich Germany

ISBN: (纸本)9780769546711

In the domain of safety-critical embedded and cyber-physical systems, software-based redundancy is generally understood as an effective and cheap approach to improve reliability. Especially redundant execution in terms of triple modular redundancy is a well-known solution. However, triple modular redundancy (TMR) leaves unprotected single points of failure (SPOFs), such as the voter, which have to be carefully considered in all safety considerations. We present Combined Redundancy (CoRed), a holistic approach that hardens safety-critical parts of a system against soft-errors, while effectively eliminating the vulnerability caused by SPOFs. CoRed leverages redundant execution in combination with encoded processing to tackle the unprotected voting and data distribution. Its implementation does not require specific knowledge about the application and can be easily integrated into existing projects. We evaluated CoRed in a realistic setting using a quad rotor helicopter and provide experimental evidence for soft-error resistance and comparable low resource demand. In our experimental comparison plain TMR left more than seven percent of failures undetected, whereas CoRed was able to eliminate all silent data corruptions while inducing an overhead of just seven percent. © 2012 IEEE.

关键词： Fault tolerance

来源：评论

学校读者我要写书评

暂无评论

Mary, Hugo, and Hugo*: Learning to schedule distributed data-parallel processing jobs on shared clusters

Mary, Hugo, and Hugo*: Learning to schedule distributed data...

引用

作者： Thamsen, Lauritz Beilharz, Jossekin Tran, Vinh Thuy Nedelkoski, Sasho Kao, Odej Complex and Distributed IT Systems Technische Universität Berlin Berlin Germany Operating Systems and Middleware Group Hasso Plattner Institute University of Potsdam Potsdam Germany Berlin Germany

distributed data-parallel processing systems like MapReduce, Spark, and Flink are popular for analyzing large datasets using cluster resources. Resource management systems like YARN or Mesos in turn allow multiple data-parallel processing jobs to share cluster resources in temporary containers. Often, the containers do not isolate resource usage to achieve high degrees of overall resource utilization despite overprovisioning and the often fluctuating utilization of specific jobs. However, some combinations of jobs utilize resources better and interfere less with each other when running on the same shared nodes than others. This article presents an approach for improving the resource utilization and job throughput when scheduling recurring distributed data-parallel processing jobs in shared clusters. The approach is based on reinforcement learning and a measure of co-location goodness to have cluster schedulers learn over time which jobs are best executed together on shared resources. We evaluated this approach over the last years with three prototype schedulers that build on each other: Mary, Hugo, and Hugo*. For the evaluation we used exemplary Flink and Spark jobs from different application domains and clusters of commodity nodes managed by YARN. The results of these experiments show that our approach can increase resource utilization and job throughput significantly. © 2020 The Authors. Concurrency and Computation: Practice and Experience published by John Wiley & Sons, Ltd.

关键词： Containers

来源：评论

学校读者我要写书评

暂无评论

VPRI: Efficient I/O Page Fault Handling via Software-Hardware Co-Design for IaaS Clouds 24

VPRI: Efficient I/O Page Fault Handling via Software-Hardwar...

引用

30th ACM Symposium on operating systems Principles, SOSP 2024

作者： Guo, Kaijie Li, Dingji Luo, Ben Shen, Yibin Peng, Kaihuan Luo, Ning Dai, Shengdong Liang, Chen Song, Jianming Yang, Hang Zhang, Xiantao Mi, Zeyu Alibaba Cloud Hangzhou China Institute of Parallel and Distributed Systems Shanghai Jiao Tong University China Engineering Research Center for Domain-specific Operating Systems Ministry of Education China

ISBN: (纸本)9798400712517

Device pass-through has been widely adopted by cloud service providers to achieve near bare-metal I/O performance in virtual machines (VMs). However, this approach requires static pinning of VM memory, making on-demand paging unavailable. The hardware device I/O page fault (IOPF) capability offers an optimal solution to this limitation. Current IOPF approaches, using either standard IOMMU capabilities (ATS+PRI) or devices with independent IOMMU implementations, have not gained widespread adoption in public Infrastructure-as-a-Service clouds. This is due to high costs, platform dependency, and significant impacts on performance and service level objectives (SLOs). We present the Virtualized Page Request Interface (VPRI), a novel IOPF system developed through software-hardware collaboration. VPRI is not only platform-independent, free from address translation complexities, but also cost-effective, and designed to minimize SLO impact. Our work enables large-scale deployment of IOPF capability in Alibaba Cloud with negligible impact on SLOs. When integrated with memory management software, it significantly enhances memory utilization in public IaaS clouds, effectively overcoming the static memory pinning restriction associated with pass-through devices. © 2024 Copyright is held by the owner/author(s). Publication rights licensed to ACM.

关键词： Infrastructure as a service (IaaS)

来源：评论

学校读者我要写书评

暂无评论

Approximate analysis of access contention for multiprocessor systems with common bus arbiters

引用

systems AND COMPUTERS IN JAPAN 1996年第9期27卷 12-22页

作者： Hoshiai, T Member NTT Network Service Systems Laboratories Musashino Japan 180 Takashige Hoshai:received his B.S. and M.S. degrees in 1984 and 1986 respectively from Toyohashi Univ. of Tech. where he received his Ph.D. degree. In 1986 he joined NIT Communication Switching Laboratories and is Senior Researcher at IIT Network Service Systems Laboratories. At present he is engaged in research on distributed systems at Bell Communication Lab. (USA). His main interests are real time operating systems computer architecture and distributed systems.

In multiprocessor systems with a common bus arbiter, contention for access to the common bus and shared memory is a major cause of performance degradation. The first-order moment match approximation is an effective technique of performance evaluation in such systems. Comparison with simulation results demonstrates that the first-order moment match technique gives good accuracy. The performance of completely coupled multiprocessor systems is compared with that of bus-arbitrated multiprocessor systems in order to determine the applications for which the latter is best suited.

关键词： Multiprocessor system access contention bus arbiter first-order moment matching

来源：评论

学校读者我要写书评

暂无评论

Jade: A High-throughput Concurrent Copying Garbage Collector 24

Jade: A High-throughput Concurrent Copying Garbage Collector

引用

19th European Conference on Computer systems, EuroSys 2024

作者： Wu, Mingyu Mao, Liang Lin, Yude Jin, Yifeng Li, Zhe Lyu, Hongtao Tang, Jiawei Lu, Xiaowei Tang, Hao Dong, Denghui Chen, Haibo Zang, Binyu Institute of Parallel and Distributed Systems Seiee Shanghai Jiao Tong University China Alibaba Group China Engineering Research Center for Domain-specific Operating Systems Ministry of Education China

ISBN: (纸本)9798400704376

Garbage collection (GC) pauses are a notorious issue threatening the latency of applications. To mitigate this problem, state-of-the-art concurrent copying collectors allow GC threads to run simultaneously with application threads (mutators) in nearly all GC phases. However, the design of concurrent copying collectors does not always lead to low application latency. To this end, this work studies the behaviors of mainstream concurrent copying collectors in OpenJDK and mainly focuses on long application pauses under heavy workloads. By analyzing the design of those collectors, this work uncovers that lengthy pre-reclamation cycles (including GC phases before actual memory release), high GC frequency, and large metadata maintenance overhead are major factors for long pauses. Therefore, this work proposes Jade, a concurrent copying collector aiming to achieve both short pauses and high GC efficiency. Compared with existing collectors, Jade provides a group-wise collection mechanism to shorten pre-reclamation cycles while controlling GC frequency. It also embraces a generational heap layout and a single-phase algorithm to maximize young GC's throughput. The evaluation results on representative latency-critical applications show that Jade can reach sub-millisecond-level pauses even under heavy workloads and significantly improve applications' peak throughput compared with state-of-the-art concurrent collectors. © 2024 ACM.

关键词： Refuse collection

来源：评论

学校读者我要写书评

暂无评论

Applying wireless sensor networks to renaturation monitoring: From requirements to deployment

Applying wireless sensor networks to renaturation monitoring...

引用

IFIP Wireless Days, WD

作者： Hannes Menzel Reinhardt Karnapke Steffen Büchner Jörg Nolte Operating Systems Distributed Systems Brandenburg University of Technology Cottbus

In Germany, renaturation of former coal mining regions is an important topic. One of these regions, called Chicken Creek, is used by ecologists to observe the development of an initial ecosystem without human interference - a scenario that seems perfect for a wireless sensor network. However, the sensor network applications are still mostly designed from scratch for every new scenario. The reason for this are strongly differing user requirements. This time-consuming problem can be overcome by providing a reusable module pool from which hard- and software components can be chosen. Based upon this module pool, a semi-automatic generation of sensor network applications becomes possible. In this paper we focus on the differences between the aforesaid theory and the practice during the initial deployment.

关键词： Wireless sensor networks Wireless communication Hardware Bridges Software Soil measurements Batteries

来源：评论

学校读者我要写书评

暂无评论

A new perspective on qality evaluation for control systems with stochastic timing 21

A new perspective on qality evaluation for control systems w...

引用

21st International Conference on Hybrid systems: Computation and Control, HSCC 2018

作者： Gaukler, Maximilian Michalka, Andreas Ulbrich, Peter Klaus, Tobias Department of Automatic Control Friedrich-Alexander-Universität Erlangen-Nürnberg Germany Department of Distributed Systems and Operating Systems Friedrich-Alexander-Universität Erlangen-Nürnberg Germany

ISBN: (纸本)9781450356428

As control applications are particularly sensitive to timing variations, the Quality of Control (QoC) is degraded by varying execution conditions of the underlying real-time system. In particular, transitions between different execution or environmental conditions pose a significant issue as they may impact the QoC unexpectedly. So far, the QoC is usually evaluated either in a stationary, timeinvariant way, which cannot analyze said transitions, or by simulation, which becomes ineficient when confronted with random influencing factors. In this paper, we propose a new perspective on QoC evaluation for modern, adaptive real-time systems with varying timing conditions. For this, we present a time-variant stochastic assessment approach that incorporates the effects mentioned before. Our results demonstrate that adaptive scheduling and runtime behavior considerably impacts the QoC. At the same time, the proposed scheme significantly outperforms a traditional simulation. © 2018 Copyright held by the owner/author(s).

关键词： Real time systems

来源：评论

学校读者我要写书评

暂无评论

Energy-aware Reconfiguration of Sensor Nodes

Energy-aware Reconfiguration of Sensor Nodes

引用

ITG FA 6.2 Workshop on Model-Based Testing, GI/ITG Workshop on Non-Functional Properties of Embedded systems, 13th GI/ITG Conference Measuring, Modelling, and Evaluation of Computer and Communications

作者： Andreas Weissel University of Erlangen-Nuremberg Distributed Systems and Operating Systems Germany

Energy consumption is one of the most challenbring constraints for the design and implementation of sensor networks. As sensor nodes are commonly battery-driven, the way the available energy is used determines the lifetime of the system. In recent research, a variety of approaches has been investigated to predict energy consumption off-line using simulation, emulation, code analysis and energy models of hardware components. However, for a wide range of applications the run-time behavior is dictated by sensor measurements or external events, allowing only a worst case analysis before deployment. In this paper, we investigate different approaches to derive energy characteristics during run-time and to adapt or reconfigure the applications running on individual sensor nodes. We discuss prototype implementations based on TinyOS for the BTnode architecture and their overhead regarding energy consumption and code size.

关键词：

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：