2020 IEEE 9th International Conference on Cloud Networking (CloudNet)最新文献_第2页

Resource Inference for Task Migration in Challenged Edge Networks with RITMO 基于RITMO的挑战边缘网络任务迁移资源推断

2020 IEEE 9th International Conference on Cloud Networking (CloudNet) Pub Date : 2020-11-09 DOI: 10.1109/CloudNet51028.2020.9335807

Alessio Sacco, Flavio Esposito, G. Marchetto

{"title":"Resource Inference for Task Migration in Challenged Edge Networks with RITMO","authors":"Alessio Sacco, Flavio Esposito, G. Marchetto","doi":"10.1109/CloudNet51028.2020.9335807","DOIUrl":"https://doi.org/10.1109/CloudNet51028.2020.9335807","url":null,"abstract":"Edge computing, combined with the proliferation of IoT devices, is generating new business model opportunities and applications. Among those applications, Unmanned Aerial Vehicles (UAVs) have been deployed in several scenarios, from surveillance and monitoring to disaster response, to precision agriculture. To support such applications, however, edge network managers and application programmers need to overcome a few challenges, e.g., unstable network conditions, high loss rate, and node failures. Existing solutions designed to mitigate such inefficiencies by predicting future network conditions are often computationally intensive and hence less portable on constrained devices. In this paper, we propose RITMO, a distributed and adaptive task planning algorithm that aims at solving these challenges while running on a network of UAV devices. We model our system as a network of queues, and we exploit a simple yet effective ARIMA regressor, to dynamically predict the length of future UAV task queues. Such prediction is then used to proactively migrate the tasks in case of a failure or unbalanced loads. Our simulation results demonstrate how RITMO helps to reduce the overall latency perceived by the application and anticipates the node overloading by avoiding agents that are likely to exhaust their computational resources.","PeriodicalId":156419,"journal":{"name":"2020 IEEE 9th International Conference on Cloud Networking (CloudNet)","volume":"2011 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-11-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125634855","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 6

Scheduling Service Function Chains with Dependencies in the Cloud 在云中调度具有依赖关系的业务功能链

2020 IEEE 9th International Conference on Cloud Networking (CloudNet) Pub Date : 2020-11-09 DOI: 10.1109/CloudNet51028.2020.9335790

Tarek Menouer, Amina Khedimi, C. Cérin, Mohammed Chahbar

引用次数: 4

Cooperative Rule Caching for SDN Switches SDN交换机协同规则缓存

2020 IEEE 9th International Conference on Cloud Networking (CloudNet) Pub Date : 2020-11-09 DOI: 10.1109/CloudNet51028.2020.9335795

Ori Rottenstreich, A. Kulik, Ananya Joshi, J. Rexford, G. Rétvári, D. Menasché

{"title":"Cooperative Rule Caching for SDN Switches","authors":"Ori Rottenstreich, A. Kulik, Ananya Joshi, J. Rexford, G. Rétvári, D. Menasché","doi":"10.1109/CloudNet51028.2020.9335795","DOIUrl":"https://doi.org/10.1109/CloudNet51028.2020.9335795","url":null,"abstract":"Despite the tremendous success of SDNs in datacen-ters, their wide adoption still poses a key challenge: the packet-forwarding rules in switches require fast and power-hungry memories. Rule tables, which serve as caches, are of limited size in cheap and energy-constrained devices, motivating novel solutions to achieve high hit rates. In this paper, we leverage device connectivity in the fast data plane, where delays are in the order of few milliseconds, and propose multiple switches to work together to avoid accessing the control plane, where delays are orders of magnitude greater. As a low priority rule in a cache entails caching higher priority rules, we pose the problem of cooperative caching with dependencies. We provide models and algorithms for cooperative rule caching with dependencies, accounting for dependencies among rules implied by existing switch memory types. We develop caching algorithms for several typical use cases and study the difficulty to find an optimal cooperative rule placement as a function of the matching pattern, which lay the foundations of cooperative caching with dependencies.","PeriodicalId":156419,"journal":{"name":"2020 IEEE 9th International Conference on Cloud Networking (CloudNet)","volume":"70 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-11-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124570720","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 6

ECML: Improving Efficiency of Machine Learning in Edge Clouds ECML:提高边缘云中的机器学习效率

2020 IEEE 9th International Conference on Cloud Networking (CloudNet) Pub Date : 2020-11-09 DOI: 10.1109/CloudNet51028.2020.9335804

Aditya Dhakal, Sameer G. Kulkarni, K. Ramakrishnan

{"title":"ECML: Improving Efficiency of Machine Learning in Edge Clouds","authors":"Aditya Dhakal, Sameer G. Kulkarni, K. Ramakrishnan","doi":"10.1109/CloudNet51028.2020.9335804","DOIUrl":"https://doi.org/10.1109/CloudNet51028.2020.9335804","url":null,"abstract":"Edge cloud data centers (Edge) are deployed to provide responsive services to the end-users. Edge can host more powerful CPUs and DNN accelerators such as GPUs and may be used for offloading tasks from end-user devices that require more significant compute capabilities. But Edge resources may also be limited and must be shared across multiple applications that process requests concurrently from several clients. However, multiplexing GPUs across applications is challenging. With edge cloud servers needing to process a lot of streaming and the advent of multi-GPU systems, getting that data from the network to the GPU can be a bottleneck, limiting the amount of work the GPU cluster can do. The lack of prompt notification of job completion from the GPU can also result in poor GPU utilization. We build on our recent work on controlled spatial sharing of a single GPU to expand to support multi-GPU systems and propose a framework that addresses these challenges. Unlike the state-of-the-art uncontrolled spatial sharing currently available with systems such as CUDA-MPS, our controlled spatial sharing approach uses each of the GPU in the cluster efficiently by removing interference between applications, resulting in much better, predictable, inference latency We also use each of the cluster GPU's DMA engines to offload data transfers to the GPU complex, thereby preventing the CPU from being the bottleneck. Finally, our framework uses the CUDA event library to give timely, low overhead GPU notifications. Our evaluations show we can achieve low DNN inference latency and improve DNN inference throughput by at least a factor of 2.","PeriodicalId":156419,"journal":{"name":"2020 IEEE 9th International Conference on Cloud Networking (CloudNet)","volume":"81 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-11-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121888234","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 6

Virtualized Network Graph Design and Embedding Model 虚拟网络图设计与嵌入模型

2020 IEEE 9th International Conference on Cloud Networking (CloudNet) Pub Date : 2020-11-09 DOI: 10.1109/CloudNet51028.2020.9335799

Takehiro Sato, T. Kurimoto, S. Urushidani, E. Oki

引用次数: 1

A Novel Scalable On-chip Switch Architecture with Quality of Service Support for Hardware Accelerated Cloud Data Centers 一种支持硬件加速云数据中心服务质量的新型可扩展片上交换机架构

2020 IEEE 9th International Conference on Cloud Networking (CloudNet) Pub Date : 2020-11-09 DOI: 10.1109/CloudNet51028.2020.9335788

Fatih Yazıcı, Ayhan Sefa Yıldız, Alper Yazar, E. G. Schmidt

引用次数: 2

POX-PLUS: An SDN Controller with Dynamic Shortest Path Routing POX-PLUS:动态最短路径路由的SDN控制器

2020 IEEE 9th International Conference on Cloud Networking (CloudNet) Pub Date : 2020-11-09 DOI: 10.1109/CloudNet51028.2020.9335792

M. Alshammari, A. Rezgui

引用次数: 2

Robust Optimization for Probabilistic Protection with Multiple Types of Resources in Cloud 云环境下多类型资源概率保护的鲁棒优化

2020 IEEE 9th International Conference on Cloud Networking (CloudNet) Pub Date : 2020-11-09 DOI: 10.1109/CloudNet51028.2020.9335810

Mitsuki Ito, Fujun He, E. Oki

{"title":"Robust Optimization for Probabilistic Protection with Multiple Types of Resources in Cloud","authors":"Mitsuki Ito, Fujun He, E. Oki","doi":"10.1109/CloudNet51028.2020.9335810","DOIUrl":"https://doi.org/10.1109/CloudNet51028.2020.9335810","url":null,"abstract":"This paper proposes a robust optimization model for probabilistic protection with multiple types of resources to minimize the required backup capacity for each type of resource against multiple random failures of physical machines in a cloud provider. If random failures occur, the required capacities for virtual machines are allocated to the dedicated backup physical machines, which are determined in advance. Probabilistic protection restricts the probability that the workload caused by failures exceeds the backup capacity by a given survivability parameter. We introduce three survivability parameters for central processing unit (CPU), memory, and the entire cloud provider considering both CPU and memory. By using the relationship between the three survivability parameters, the proposed model guarantees probabilistic protection for each resource, CPU and memory, and the entire cloud provider. By adopting the robust optimization technique, we formulate the proposed model as a multi-objective mixed integer linear programming problem. To deal with the multi-objective optimization problem, we apply the lexicographic weighted Tchebycheff method with which a Pareto optimal solution is obtained. Our proposed model reduces the average value between the backup capacity ratios of CPU and memory compared with the conventional model.","PeriodicalId":156419,"journal":{"name":"2020 IEEE 9th International Conference on Cloud Networking (CloudNet)","volume":"478 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-11-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115216600","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 2

Roadblocks of I/O Parallelization: Removing H/W Contentions by Static Role Assignment in VNFs I/O并行化的障碍:通过VNFs中的静态角色分配消除H/W争用

2020 IEEE 9th International Conference on Cloud Networking (CloudNet) Pub Date : 2020-11-09 DOI: 10.1109/CloudNet51028.2020.9335803

Masahiro Asada, Ryota Kawashima, Hiroki Nakayama, Tsunemasa Hayashi, H. Matsuo

{"title":"Roadblocks of I/O Parallelization: Removing H/W Contentions by Static Role Assignment in VNFs","authors":"Masahiro Asada, Ryota Kawashima, Hiroki Nakayama, Tsunemasa Hayashi, H. Matsuo","doi":"10.1109/CloudNet51028.2020.9335803","DOIUrl":"https://doi.org/10.1109/CloudNet51028.2020.9335803","url":null,"abstract":"Achieving 100 Gbps+ throughput with commodity servers is a challenging goal, even with state-of-the-art Data Plane Development Kit (DPDK). Fundamental performance of CPU/Memory is now the bottleneck and simple code optimization of Network Functions (NFs) cannot be the solution. Hardware accelerators including FPGA are getting attentions for performance boost; however, relying on specific features degrades manageability of NFV-nodes. Common Receive Side Scaling (RSS) provides a means of H/W-level parallelization, but per-flow throughput is not accelerated. Existing software-based approaches distribute processing load of NFs, but I/O is still serialized for each datapath. We tackled I/O parallelization and uncovered encounterd certainly misty contentions in our previous study. Specifically, per-thread CPU cycle consumptions proportionally grew as increasing parallelization level, although the overhead of conceivable mutual executions (e.g. CAS operations) was trivial. In this paper, we pursue the cause of the issue and upgrade our I/O parallelization scheme. Our careful investigation of NFV-node internals ranging from application to device driver layers indicates that hidden H/W-level contentions involving DMA heavily consume CPU cycles. We propose a contention avoidance design of thread role assignment and prove our design can flatten per-thread CPU cycle consumptions.","PeriodicalId":156419,"journal":{"name":"2020 IEEE 9th International Conference on Cloud Networking (CloudNet)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-11-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128913351","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Supporting IoT Applications with Serverless Edge Clouds 通过无服务器边缘云支持物联网应用

2020 IEEE 9th International Conference on Cloud Networking (CloudNet) Pub Date : 2020-11-09 DOI: 10.1109/CloudNet51028.2020.9335805

I. Wang, Elizabeth Liri, K. Ramakrishnan

引用次数: 8