Shaheen Sultana, Fahiem Altaf, Mayank K. Aditia, M. S. Burra, Chanchal Maurya, Soumyadev Maity
For many Intelligent Transportation System (ITS) applications, periodic transmission of beacon messages by a source vehicle is paramount for providing the unhindered service and to maintain its accuracy. A malicious source vehicle could deny beacon messages of a target ITS application and intelligently cover-up this denial with signal loss faced due to wireless channel error and surrounding environment obstacles. In this context, we propose a lightweight technique to detect a beacon denial attacker using Auto-Correlation Function (ACF). This technique possesses higher detection accuracy in comparison to fixed threshold scheme which depends upon the number of beacon messages denied by the source vehicle. Furthermore, we propose a random vehicle inspection scheme for inspecting randomly chosen set of vehicles for beacon denial attack. In addition to reducing computational overhead, the advantage of this scheme is that an attacker with knowledge about detection algorithm cannot decide the time and place of inspection. Simulation results confirm the accurate and lightweight nature of the proposed technique.
{"title":"Detection of beacon transmission denial attack in ITS using temporal auto-correlation and random inspections","authors":"Shaheen Sultana, Fahiem Altaf, Mayank K. Aditia, M. S. Burra, Chanchal Maurya, Soumyadev Maity","doi":"10.1145/3288599.3288616","DOIUrl":"https://doi.org/10.1145/3288599.3288616","url":null,"abstract":"For many Intelligent Transportation System (ITS) applications, periodic transmission of beacon messages by a source vehicle is paramount for providing the unhindered service and to maintain its accuracy. A malicious source vehicle could deny beacon messages of a target ITS application and intelligently cover-up this denial with signal loss faced due to wireless channel error and surrounding environment obstacles. In this context, we propose a lightweight technique to detect a beacon denial attacker using Auto-Correlation Function (ACF). This technique possesses higher detection accuracy in comparison to fixed threshold scheme which depends upon the number of beacon messages denied by the source vehicle. Furthermore, we propose a random vehicle inspection scheme for inspecting randomly chosen set of vehicles for beacon denial attack. In addition to reducing computational overhead, the advantage of this scheme is that an attacker with knowledge about detection algorithm cannot decide the time and place of inspection. Simulation results confirm the accurate and lightweight nature of the proposed technique.","PeriodicalId":346177,"journal":{"name":"Proceedings of the 20th International Conference on Distributed Computing and Networking","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129450553","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}
I. M. A. Jawarneh, P. Bellavista, L. Foschini, Giuseppe Martuscelli, R. Montanari, Amedeo Palopoli, F. Bosi
Current cloud deployments heavily depend on hypervisor-based virtualizations. The overarching characteristics of Docker and containerization have given them a momentum in their widespread adoption recently as alternatives for their counterparts. However, little research has been done for comparing the QoS of both technologies, thus leaving the domain without widely accepted performance metrics. Aiming at informing the decision of the best fit in a specific cloud deployment, we have designed performance metrics that compare the performance of both designs in an in-house cluster deployed by using OpenStack. We focus on well-established representatives as baselines, including KVM from the hypervisor-based side, LXD from the container-based side in addition to Docker. Our results show that containerization is not a predominant fit-all solution that can always replace hypervisors for all cluster deployment and application scenarios. It can instead be thought of as a complementary solution to use for specific application scenarios that are constrained with conditions that are solved by containerization merits.
{"title":"QoS and performance metrics for container-based virtualization in cloud environments","authors":"I. M. A. Jawarneh, P. Bellavista, L. Foschini, Giuseppe Martuscelli, R. Montanari, Amedeo Palopoli, F. Bosi","doi":"10.1145/3288599.3288631","DOIUrl":"https://doi.org/10.1145/3288599.3288631","url":null,"abstract":"Current cloud deployments heavily depend on hypervisor-based virtualizations. The overarching characteristics of Docker and containerization have given them a momentum in their widespread adoption recently as alternatives for their counterparts. However, little research has been done for comparing the QoS of both technologies, thus leaving the domain without widely accepted performance metrics. Aiming at informing the decision of the best fit in a specific cloud deployment, we have designed performance metrics that compare the performance of both designs in an in-house cluster deployed by using OpenStack. We focus on well-established representatives as baselines, including KVM from the hypervisor-based side, LXD from the container-based side in addition to Docker. Our results show that containerization is not a predominant fit-all solution that can always replace hypervisors for all cluster deployment and application scenarios. It can instead be thought of as a complementary solution to use for specific application scenarios that are constrained with conditions that are solved by containerization merits.","PeriodicalId":346177,"journal":{"name":"Proceedings of the 20th International Conference on Distributed Computing and Networking","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116170482","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}
In the aftermath of a large-scale disaster (such as earthquake), existing communication infrastructures are often critically impaired, preventing timely information exchange between the survivors, responders, and the coordination center. Typically, a temporary network, called Disaster Response Network (DRN), is set up using smart devices, movable base stations and easily deployable cellular antennas. However, such networks are challenged by rapid devices' energy depletion and component failures due to environmental adversities and hardware faults. State-of-the-art literature address energy challenges through intelligent routing, however robustness of DRN against component failures is largely unaddressed. In this paper, we investigate designing a novel network topology for DRNs, which is both energy-efficient and robust against component devices' failures. Specifically, the objective is to construct a sparse structure from the original DRN (termed, Sparse-DRN) while ensuring that there exists a connected tree backbone. Our performance evaluation shows that the Sparse-DRN offers a good trade-off between the energy efficiency and network robustness, while ensuring the QoS requirements i.e., packet delivery and network latency.
{"title":"Towards energy-efficient and robust disaster response networks","authors":"V. Shah, Satyaki Roy, S. Silvestri, Sajal K. Das","doi":"10.1145/3288599.3295589","DOIUrl":"https://doi.org/10.1145/3288599.3295589","url":null,"abstract":"In the aftermath of a large-scale disaster (such as earthquake), existing communication infrastructures are often critically impaired, preventing timely information exchange between the survivors, responders, and the coordination center. Typically, a temporary network, called Disaster Response Network (DRN), is set up using smart devices, movable base stations and easily deployable cellular antennas. However, such networks are challenged by rapid devices' energy depletion and component failures due to environmental adversities and hardware faults. State-of-the-art literature address energy challenges through intelligent routing, however robustness of DRN against component failures is largely unaddressed. In this paper, we investigate designing a novel network topology for DRNs, which is both energy-efficient and robust against component devices' failures. Specifically, the objective is to construct a sparse structure from the original DRN (termed, Sparse-DRN) while ensuring that there exists a connected tree backbone. Our performance evaluation shows that the Sparse-DRN offers a good trade-off between the energy efficiency and network robustness, while ensuring the QoS requirements i.e., packet delivery and network latency.","PeriodicalId":346177,"journal":{"name":"Proceedings of the 20th International Conference on Distributed Computing and Networking","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130571303","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}
In this work, we consider the problem of secure multi-party computation (MPC) with n parties where at most t are under the control of a computationally bounded adversary given the constraint t < n/2. We aim to design a round efficient protocol by minimizing the number of rounds in which the broadcast primitive is involved. In this setting, the previous best protocol can be attributed to Katz-Koo (EUROCRYPT 2007) which is set in the offline-online paradigm (where the parties generate preprocessing data during the offline phase to lighten the computation in the online phase). Their online phase is a constant round protocol with no invocations of broadcast, while the offline phase protocol needs total 29 rounds with a broadcast invocation in one round. Our work improves the round complexity of their offline phase protocol, by running in 4 rounds, with only a single broadcast round. Additionally, we also improve the communication complexity of the offline phase protocol by a factor of Ω(n3). As a technical contribution, we present the first two round computationally-secure verifiable secret-sharing (VSS) scheme that invokes broadcast in only one round.
{"title":"Round efficient computationally secure multi-party computation revisited","authors":"Laasya Bangalore, Ashish Choudhury, Gayathri Garimella","doi":"10.1145/3288599.3288600","DOIUrl":"https://doi.org/10.1145/3288599.3288600","url":null,"abstract":"In this work, we consider the problem of secure multi-party computation (MPC) with n parties where at most t are under the control of a computationally bounded adversary given the constraint t < n/2. We aim to design a round efficient protocol by minimizing the number of rounds in which the broadcast primitive is involved. In this setting, the previous best protocol can be attributed to Katz-Koo (EUROCRYPT 2007) which is set in the offline-online paradigm (where the parties generate preprocessing data during the offline phase to lighten the computation in the online phase). Their online phase is a constant round protocol with no invocations of broadcast, while the offline phase protocol needs total 29 rounds with a broadcast invocation in one round. Our work improves the round complexity of their offline phase protocol, by running in 4 rounds, with only a single broadcast round. Additionally, we also improve the communication complexity of the offline phase protocol by a factor of Ω(n3). As a technical contribution, we present the first two round computationally-secure verifiable secret-sharing (VSS) scheme that invokes broadcast in only one round.","PeriodicalId":346177,"journal":{"name":"Proceedings of the 20th International Conference on Distributed Computing and Networking","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123952707","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}
Microblogging sites are increasingly playing an important role in real-time disaster management. However, rumors and fake news often spread on such platforms, which if not detected, can derail the rescue operations. Therefore, it becomes imperative to verify some of the information posted on social media during disaster situations. To this end, it is necessary to correctly identify fact-checkable posts, so that their information content can be verified. In the present work, we address the problem of identifying fact-checkable posts on the Twitter microblogging site. We organized a shared task in the FIRE 2018 conference to study the problem of identification of fact-checkable tweets posted during a particular disaster event (the 2015 Nepal earthquake). This paper describes the dataset used in the shared task, and compares the performance of different methodologies for identifying fact-checkable tweets. We primarily experiment with two different types of approaches - classification-based and ranking-based. Our experiments show that a hybrid methodology involving both classification and ranking performs well and outperforms the methodologies that employ only classification or only ranking.
{"title":"Identifying fact-checkable microblogs during disasters: a classification-ranking approach","authors":"Divyank Barnwal, Siddharth Ghelani, Rohit Krishna, Moumita Basu, Saptarshi Ghosh","doi":"10.1145/3288599.3295587","DOIUrl":"https://doi.org/10.1145/3288599.3295587","url":null,"abstract":"Microblogging sites are increasingly playing an important role in real-time disaster management. However, rumors and fake news often spread on such platforms, which if not detected, can derail the rescue operations. Therefore, it becomes imperative to verify some of the information posted on social media during disaster situations. To this end, it is necessary to correctly identify fact-checkable posts, so that their information content can be verified. In the present work, we address the problem of identifying fact-checkable posts on the Twitter microblogging site. We organized a shared task in the FIRE 2018 conference to study the problem of identification of fact-checkable tweets posted during a particular disaster event (the 2015 Nepal earthquake). This paper describes the dataset used in the shared task, and compares the performance of different methodologies for identifying fact-checkable tweets. We primarily experiment with two different types of approaches - classification-based and ranking-based. Our experiments show that a hybrid methodology involving both classification and ranking performs well and outperforms the methodologies that employ only classification or only ranking.","PeriodicalId":346177,"journal":{"name":"Proceedings of the 20th International Conference on Distributed Computing and Networking","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123117654","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}
In this paper, the problem of minimizing the utilization of resources at regenerators under dynamic environment in translucent optical networks has been studied. The problem is an NP-complete one the solution of which remains intractable as the network size increases. A heuristic approach is proposed here to route each incoming request in a way to reduce the number of regenerations at regenerating nodes under dynamic scenario to the extent possible. Input to this algorithmic approach is the regenerator positions obtained from the output of execution of another heuristic CLR [14] which actually selects minimum possible number of regenerator nodes in the network. Performance comparisons show that the proposed heuristic provide a significant improvement over the existing one both in terms of maximum number of regenerations that occur amongst all the regenerator nodes and also connection blocking probability.
{"title":"An efficient heuristic to minimize number of regenerations in translucent optical network under dynamic scenario","authors":"Ira Nath, U. Bhattacharya, M. Chatterjee","doi":"10.1145/3288599.3295578","DOIUrl":"https://doi.org/10.1145/3288599.3295578","url":null,"abstract":"In this paper, the problem of minimizing the utilization of resources at regenerators under dynamic environment in translucent optical networks has been studied. The problem is an NP-complete one the solution of which remains intractable as the network size increases. A heuristic approach is proposed here to route each incoming request in a way to reduce the number of regenerations at regenerating nodes under dynamic scenario to the extent possible. Input to this algorithmic approach is the regenerator positions obtained from the output of execution of another heuristic CLR [14] which actually selects minimum possible number of regenerator nodes in the network. Performance comparisons show that the proposed heuristic provide a significant improvement over the existing one both in terms of maximum number of regenerations that occur amongst all the regenerator nodes and also connection blocking probability.","PeriodicalId":346177,"journal":{"name":"Proceedings of the 20th International Conference on Distributed Computing and Networking","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127611004","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}
The emerging Internet of Things (IoT) has several salient characteristics that differentiate it from existing wireless networking architectures. These include the deployment of very large numbers of (possibly) low-complexity terminals; the need for low-latency, short-packet communications (e.g., to support automation); light or no infrastructure; and primary applications of data gathering, inference and control. These characteristics have motivated the development of new fundamentals that can provide insights into the limits of communication in this regime. This paper discusses two issues in this context, namely security and low-latency, through the respective lenses of physical layer security and finite-blocklength information theory.
{"title":"Fundamentals for IoT networks: secure and low-latency communications","authors":"H. Poor, Mario Goldenbaum, Wei Yang","doi":"10.1145/3288599.3288643","DOIUrl":"https://doi.org/10.1145/3288599.3288643","url":null,"abstract":"The emerging Internet of Things (IoT) has several salient characteristics that differentiate it from existing wireless networking architectures. These include the deployment of very large numbers of (possibly) low-complexity terminals; the need for low-latency, short-packet communications (e.g., to support automation); light or no infrastructure; and primary applications of data gathering, inference and control. These characteristics have motivated the development of new fundamentals that can provide insights into the limits of communication in this regime. This paper discusses two issues in this context, namely security and low-latency, through the respective lenses of physical layer security and finite-blocklength information theory.","PeriodicalId":346177,"journal":{"name":"Proceedings of the 20th International Conference on Distributed Computing and Networking","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122513398","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}
Load balanced transaction scheduling in on-demand computing system is known to be NP-hard problem. In order to solve this problem, this paper introduces a hybrid approach named cuckoo search-ant colony optimization. The approach dynamically generates an optimal schedule by clustering the on-demand computing resources considering their load and completes the transaction execution within their deadlines. The approach also balances the load of the system before scheduling the transactions. For clustering the resources we use cuckoo search method. We use ant colony optimization for selecting the appropriate and optimal resources. We evaluate the performance of the proposed algorithm with six existing algorithms.
{"title":"Load balanced transaction scheduling in on-demand computing using cuckoo search-ant colony optimization","authors":"D. P. Mahato","doi":"10.1145/3288599.3298791","DOIUrl":"https://doi.org/10.1145/3288599.3298791","url":null,"abstract":"Load balanced transaction scheduling in on-demand computing system is known to be NP-hard problem. In order to solve this problem, this paper introduces a hybrid approach named cuckoo search-ant colony optimization. The approach dynamically generates an optimal schedule by clustering the on-demand computing resources considering their load and completes the transaction execution within their deadlines. The approach also balances the load of the system before scheduling the transactions. For clustering the resources we use cuckoo search method. We use ant colony optimization for selecting the appropriate and optimal resources. We evaluate the performance of the proposed algorithm with six existing algorithms.","PeriodicalId":346177,"journal":{"name":"Proceedings of the 20th International Conference on Distributed Computing and Networking","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130354499","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}
A Delay Tolerant Network being a facilitator of communication establishment in heterogeneous networking environments may present an efficient solution of communication in a Wireless Sensor Network that may exist within the ambit of the DTN. WSNs may be found easily in a DTN like those temporarily created using sensor-deployment over drowned places or villages, etc. Collecting information from those cut-off areas is of utmost importance to provide efficient rescue and relief operation. Current work suggests distributing Delsar Life Detection Sensors for probing such cut-off areas for tracing of struck or trapped life. Localization techniques have been used to locate the exact positions of the victims so that appropriate help could be administered to them with efficiency. At the time of distributing the sensors, one or two tag sensors with high power are positioned strategically in the cut-off areas that collect data from the sensors and from where data is further collected using some data mules. The results have been obtained both over simulation in the NS2 simulator and over a practical field implementation. Results suggest that the proposed architecture is able to predict presence of life with more than 70 percent accuracy. This shows that the proposed novel solution to WSN assisted DTN is practicable and efficient for implementation during disasters to communicate with cut-off and far-off remote areas.
{"title":"Sensor node assisted DTN for a post-disaster scenario","authors":"A. Gupta, J. K. Mandal, Indrajit Bhattacharya","doi":"10.1145/3288599.3295590","DOIUrl":"https://doi.org/10.1145/3288599.3295590","url":null,"abstract":"A Delay Tolerant Network being a facilitator of communication establishment in heterogeneous networking environments may present an efficient solution of communication in a Wireless Sensor Network that may exist within the ambit of the DTN. WSNs may be found easily in a DTN like those temporarily created using sensor-deployment over drowned places or villages, etc. Collecting information from those cut-off areas is of utmost importance to provide efficient rescue and relief operation. Current work suggests distributing Delsar Life Detection Sensors for probing such cut-off areas for tracing of struck or trapped life. Localization techniques have been used to locate the exact positions of the victims so that appropriate help could be administered to them with efficiency. At the time of distributing the sensors, one or two tag sensors with high power are positioned strategically in the cut-off areas that collect data from the sensors and from where data is further collected using some data mules. The results have been obtained both over simulation in the NS2 simulator and over a practical field implementation. Results suggest that the proposed architecture is able to predict presence of life with more than 70 percent accuracy. This shows that the proposed novel solution to WSN assisted DTN is practicable and efficient for implementation during disasters to communicate with cut-off and far-off remote areas.","PeriodicalId":346177,"journal":{"name":"Proceedings of the 20th International Conference on Distributed Computing and Networking","volume":"86 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130363051","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}
N. Rao, Qiang Liu, S. Sen, Zhengchun Liu, R. Kettimuthu, Ian T Foster
Distributed scientific and big-data computations are becoming increasingly dependent on access to remote files. Wide-area file transfers are supported by two basic schemes: (i) application-level tools, such as GridFTP, that provide transport services between file systems housed at geographically separated sites, and (ii) file systems mounted over wide-area networks, using mechanisms such as LNet routers that make them transparently available. In both cases, the file transfer performance critically depends on the configuration consisting of host, file, IO, and disk subsystems, which are complex by themselves, as well as on their complex compositions implemented using buffers and IO-network data transitions. We present extensive file transfer rate measurements collected over dedicated 10 Gbps connections with 0-366 ms round-trip times, using GridFTP and XDD file transfer tools, and Lustre file system extended over wide-area networks using LNet routers. Our test configurations are composed of: three types of host systems; XFS, Lustre, and ext3 file systems; and Ethernet and SONET wide-area connections. We present analytics based on the convexity-concavity of throughput profiles which provide insights into throughput and its superior or inferior trend compared to linear interpolations. We propose the utilization-concavity coefficient, a scalar metric that characterizes the overall performance of any file transfer method consisting of specific configuration and scheme. Our results enable performance optimizations by highlighting the significant roles of (i) buffer sizes and parallelism in GridFTP and XDD, and (ii) buffer utilization and credit mechanism in LNet routers.
{"title":"Measurements and analytics of wide-area file transfers over dedicated connections","authors":"N. Rao, Qiang Liu, S. Sen, Zhengchun Liu, R. Kettimuthu, Ian T Foster","doi":"10.1145/3288599.3288641","DOIUrl":"https://doi.org/10.1145/3288599.3288641","url":null,"abstract":"Distributed scientific and big-data computations are becoming increasingly dependent on access to remote files. Wide-area file transfers are supported by two basic schemes: (i) application-level tools, such as GridFTP, that provide transport services between file systems housed at geographically separated sites, and (ii) file systems mounted over wide-area networks, using mechanisms such as LNet routers that make them transparently available. In both cases, the file transfer performance critically depends on the configuration consisting of host, file, IO, and disk subsystems, which are complex by themselves, as well as on their complex compositions implemented using buffers and IO-network data transitions. We present extensive file transfer rate measurements collected over dedicated 10 Gbps connections with 0-366 ms round-trip times, using GridFTP and XDD file transfer tools, and Lustre file system extended over wide-area networks using LNet routers. Our test configurations are composed of: three types of host systems; XFS, Lustre, and ext3 file systems; and Ethernet and SONET wide-area connections. We present analytics based on the convexity-concavity of throughput profiles which provide insights into throughput and its superior or inferior trend compared to linear interpolations. We propose the utilization-concavity coefficient, a scalar metric that characterizes the overall performance of any file transfer method consisting of specific configuration and scheme. Our results enable performance optimizations by highlighting the significant roles of (i) buffer sizes and parallelism in GridFTP and XDD, and (ii) buffer utilization and credit mechanism in LNet routers.","PeriodicalId":346177,"journal":{"name":"Proceedings of the 20th International Conference on Distributed Computing and Networking","volume":"41 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128652812","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}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：