Proceedings of the SIGCOMM '21 Poster and Demo Sessions最新文献

{"title":"Steal task scheduling from OS: enabling task-network co-schedule for time-critical traffic","authors":"Xuyan Jiang, Wenwen Fu, Xiangrui Yang, Yinhan Sun, Zhigang Sun","doi":"10.1145/3472716.3472868","DOIUrl":"https://doi.org/10.1145/3472716.3472868","url":null,"abstract":"Real-Time Ethernet (RT Ethernet) such as Time-Sensitive Networking and Time-Triggered Ethernet is widely deployed in the distributed real-time systems such as aerospace, automotive and industrial domainscite{craciunas2016combined}. Typically, there are many sensoring-computing-actuating Control Loops (CL) requiring real-time and deterministic end-to-end communication. From the network perspective, these CLs work as follow. The sensor generates sensoring traffic periodically to the computing node (i.e. the end system) via RT Ethernet. Then the computing node executes the corresponding task and generates actuating traffic to the actuator. Finally, the actuator receives the traffic and executes the actions accordingly. Sensoring and actuating traffic are both Time-critical (TC) traffic. In order to achieve end-to-end determinism in the loop, the deterministic and real-time process on the end system must be guaranteed. In another word, the task running on the Operating System (OS) of the end system must be scheduled with tight dependency towards the underlying network scheduling decision. Traditionally, the task scheduling strategy on the OS is asynchronous with the underlying network (i.e. frame scheduling). This may incur great uncertainty in the scenario where end-to-end determinism must be achieved. The main reason behind is that the local time of the OS is out of sync with the network whose time is synchronized by sophisticated time synchronization protocols like PTP cite{2006NTP}. In order to address this issue, a natural solution would be synchronizing the OS time with the whole network (e.g., integrating PTP into kernel). Moreover, a special scheduling mechanism should also be implemented which takes the packet arrival info into account. However, this requires massive modifications to the kernel and may incur compatibility issues when integrating with different types of the underlying network cite{yan2020tsn}. What's more, Commercial-off-the-shelf (COTS) end systems are more adopted in modern real-time system to reduce the cost and shorten the development period cite{pellizzoni2008coscheduling}. So it is hard to introduce fundamental changes to the key components of COTS systems. In this poster, we aim to tackle this problem from a different perspective. Specifically, we suggest that by carefully scheduling packets from the ingress engine of the NIC, a task-network co-scheduling can be achieved, which eliminates the necessity of the modification to the task scheduling mechanism on the OS completely. For this purpose, we introduce Frame-Task Co-Scheduling (FTCS), a mechanism that enables co-scheduling of task and network on COTS end systems. FTCS steals task scheduling from OS through time-triggered frame schedule and submission. The potential benefits are as follows. Firstly, FTCS achieves tight co-scheduling of critical task and traffic without any modification of software and hardware on COTS devices. Secondly, OS time on the end system doe","PeriodicalId":178725,"journal":{"name":"Proceedings of the SIGCOMM '21 Poster and Demo Sessions","volume":"180 2","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120984774","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}

{"title":"Cost-effective data analytics across multiple cloud regions","authors":"Junyi Shu, Xin Jin, Yun Ma, Xuanzhe Liu, Gang Huang","doi":"10.1145/3472716.3472842","DOIUrl":"https://doi.org/10.1145/3472716.3472842","url":null,"abstract":"We propose a cloud-native data analytics engine for processing data stored among geographically distributed cloud regions with reduced cost. A job is split into subtasks and placed across regions based on factors including prices of compute resources and data transmission. We present its architecture which leverages existing cloud infrastructures and discuss major challenges of its system design. Preliminary experiments show that the cost is reduced by 15.1% for a decision support query on a four-region public cloud setup.","PeriodicalId":178725,"journal":{"name":"Proceedings of the SIGCOMM '21 Poster and Demo Sessions","volume":"41 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121041209","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}

{"title":"Mining social interactionsin connection traces of a campus wi-fi network","authors":"Eduardo Antonio Mañas Martínez, Elena Cabrera, K. Wasielewska, D. Kotz, J. Camacho","doi":"10.1145/3472716.3472844","DOIUrl":"https://doi.org/10.1145/3472716.3472844","url":null,"abstract":"Wi-Fi technologies have become one of the most popular means for Internet access. As a result, the use of mobile devices has become ubiquitous and instrumental for society. A device can be identified through its MAC address within an autonomous system. Although some devices attempt to anonymize MAC addresses via randomization, these techniques are not used once the device is associated to the network [7]. As a result, device identification poses a privacy problem in large-scale (e.g., campus-wide) Wi-Fi deployments [5]: if the mobile device can be located, the user who carries that device can also be located. In turn, location information leads to the possibility to extract private knowledge from Wi-Fi users, like social interactions, movement habits, and so forth. In this poster we report preliminary work in which we infer social interactions of individuals from Wi-Fi connection traces in the campus network at Dartmouth College [2]. We make the following contributions: (i) we propose several definitions of a pseudocorrelation matrix from Wi-Fi connection traces, which measure similarity between devices or users according to their temporal association profile to the Access Points (APs); (ii) we evaluate the accuracy of these pseudo-correlation variants in a simulation environment; and (iii) we contrast results with those found on a real trace.","PeriodicalId":178725,"journal":{"name":"Proceedings of the SIGCOMM '21 Poster and Demo Sessions","volume":"75 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127738231","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}

{"title":"SRv6-FEC: bringing forward erasure correction to IPv6 segment routing","authors":"Louis Navarre, F. Michel, O. Bonaventure","doi":"10.1145/3472716.3472863","DOIUrl":"https://doi.org/10.1145/3472716.3472863","url":null,"abstract":"IPv6 Segment Routing (SRv6) is a recent implementation of the source routing paradigm in IPv6 network. A programmability framework has recently been added to SRv6 and enables it to support diverse use-cases. We leverage this support to design, implement and assess a Forward Erasure Correction (FEC) technique that transparently protects IPv6 packets. We implement our encoders and decoders using eBPF on the Linux kernel and evaluate the benefits that they bring with IoT devices.","PeriodicalId":178725,"journal":{"name":"Proceedings of the SIGCOMM '21 Poster and Demo Sessions","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126814145","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}

{"title":"Quantifying the transient performance of congestion control algorithms","authors":"Yixin Shen, Zili Meng, Jing Chen, Mingwei Xu","doi":"10.1145/3472716.3472861","DOIUrl":"https://doi.org/10.1145/3472716.3472861","url":null,"abstract":"Congestion control algorithms (CCA) are devoted to improve their performance on long-term metrics (e.g., throughput, fairness), whereas the increasing low-latency application demand poses strict requirements on and motivates us to evaluate the transient performance of CCAs. However, due to the complexity of existing CCAs and changeable network conditions, it is challenging to quantify the transient performance. Our observation is that the transient performance could be modelled as damped oscillations. In this poster, we propose a framework to quantify the transient performance by calculating the extent of the damped oscillation. We use our framework to evaluate 3 widely-deployed CCAs and offer a comparison between their transient performance.","PeriodicalId":178725,"journal":{"name":"Proceedings of the SIGCOMM '21 Poster and Demo Sessions","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130753646","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}

{"title":"Constructing the face of network data","authors":"Ertza Warraich, M. Shahbaz","doi":"10.1145/3472716.3472852","DOIUrl":"https://doi.org/10.1145/3472716.3472852","url":null,"abstract":"Network datasets are an essential part of understanding, managing, and operating modern wide-area, data-center, and cellular networks. They are involved throughout the various stages of network development, from simulations, stress testing, to machine-learning training (for anomaly-based intrusion detection systems) and more. Despite the need, network datasets are rare due to concerns related to information privacy and sensitivity. In this paper, we aim to tackle this challenge and put forth a method, based on Generative Adversarial Networks (GANs), for generating new (and timely) datasets, automatically, that are provisioned as complete raw packets traces of a network and not just feature values.","PeriodicalId":178725,"journal":{"name":"Proceedings of the SIGCOMM '21 Poster and Demo Sessions","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130418713","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}

{"title":"The last RTT matters and we can preact","authors":"Chengjun Jia, Yifan Li, Xiaohe Hu, Jun Li","doi":"10.1145/3472716.3472847","DOIUrl":"https://doi.org/10.1145/3472716.3472847","url":null,"abstract":"In this poster, we find that the congestion control algorithms which try to acquire close-to-zero queue would encounter bandwidth waste because there would be a RTT gap for other flows to take over after a flow finishes. By distinguishing the packets of the last round in a flow's lifespan, we can improve the bandwidth utilization of datacenter network. We modify the INT process in HPCC and the FCTs of long flows can be reduced by about 20%-30% from our ns3 simulation.","PeriodicalId":178725,"journal":{"name":"Proceedings of the SIGCOMM '21 Poster and Demo Sessions","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130129249","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}

{"title":"A cloud-scale per-flow backpressure system via FPGA-based heavy hitter detection","authors":"Enge Song, Nianbing Yu, Tian Pan, Liang Xu, Yisong Qiao, Jianyuan Lu, Yilong Lv, Xiaoyu Zhang, Mingxu Xie, Jian Guo, Jun He, Jinkui Mao, Chenhao Jia, Shunmin Zhu","doi":"10.1145/3472716.3472855","DOIUrl":"https://doi.org/10.1145/3472716.3472855","url":null,"abstract":"Virtual private clouds provide sharing resources to a massive number of tenants for economics of scale. In such clouds, off-the-shelf x86 boxes are widely deployed as network intermediate nodes. However, due to rapid growth of cloud traffic and significant slowdown of CPU improvement in recent years, although horizontal scaling is still leveraged, CPU overload and packet losses caused by heavy hitters are occasionally observed in production environment, which seriously damage tenant's SLAs. To address this, we propose a cloud-scale per-flow backpressure system designed in Alibaba Cloud. The basic idea is to (1) trigger the heavy-hitter flow acquisition at the intermediate node in an on-demand manner only when the CPU utilization exceeds a predefined threshold and (2) backpressure the identified heavy-hitter flow to the traffic source via rate limiting at sender's NIC or hypervisor. To handle the extremely large traffic rate of cloud traffic, we leverage a high-speed FPGA for heavy hitter detection acceleration. To accommodate highly concurrent flows in the cloud, we design a hierarchical memory system for accurate heavy hitter counting during a large time window. Under the per-flow backpressure mechanism, the rate of the heavy-hitter flow is accurately throttled while the rate of mice flows is completely unaffected during the backpressure.","PeriodicalId":178725,"journal":{"name":"Proceedings of the SIGCOMM '21 Poster and Demo Sessions","volume":"28 1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122591405","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}

{"title":"Implementing the plugin distribution system","authors":"Nicolas Rybowski, Q. D. Coninck, Tom Rousseaux, Axel Legay, O. Bonaventure","doi":"10.1145/3472716.3472860","DOIUrl":"https://doi.org/10.1145/3472716.3472860","url":null,"abstract":"Recent works proposed to dynamically extend protocol implementations through protocol plugins. While addressing deployment issues, they raise safety concerns (do they terminate, do they act maliciously,. . . ). To fill this gap, a system distributing trust in plugin's verification properties was proposed in the literature. However, it was not implemented. This poster demonstrates the feasibility of this approach by providing an open-source implementation of this system. We also extend the state-of-the-art verification works about protocol plugins by considering a new property called side-effects.","PeriodicalId":178725,"journal":{"name":"Proceedings of the SIGCOMM '21 Poster and Demo Sessions","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124033175","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}

{"title":"Sarasate","authors":"Bin Gui, Fangping Lan, Anduo Wang","doi":"10.1145/3472716.3472848","DOIUrl":"https://doi.org/10.1145/3472716.3472848","url":null,"abstract":"Policy information in computer networking today is hard to manage. This is in sharp contrast to relational data structured in a database that allows easy access. In this demonstration, we ask why cannot (or how can) turn network policies into relational data. Our key observation is that oftentimes a policy does not prescribe a single ``definite'' network state, but rather is an ``incomplete'' description of all the legitimate network states. Based on this idea, we adopt conditional tables and the usual SQL interface (a relational structure developed for incomplete database) as a means to represent and query sets of network states in exactly the same way as a single definite network snapshot. More importantly, like relational tables that improve data productivity and innovation, relational policies allow us to extend a rich set of data mediating methods to address the networking problem of coordinating policies in a distributed environment.","PeriodicalId":178725,"journal":{"name":"Proceedings of the SIGCOMM '21 Poster and Demo Sessions","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122882906","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}