Pub Date : 2015-05-07DOI: 10.1109/ANCS.2015.7110115
J. larkins
Summary form only given. In this keynote we will present the latest trends in deploying Virtual Network Functions (VNF) for Carrier Networks. The discussion will address the challenges that operators face and their different approaches when virtualizing mobile and fixed networking applications, including EPG, BNG and many others. In addition we will highlight the latest developments in VNF acceleration, including potential acceleration targets and their locations in the Carrier Datacenter. And in conclusion explore the needs of a common acceleration abstraction and the standardization/industry efforts to achieve that goal, including Open Compute Project, Open Data Plane Project, OpenFlow and OpenFlow 2.0, ETSI and OPNFV.
{"title":"Recent trends in virtual network functions acceleration for carrier clouds","authors":"J. larkins","doi":"10.1109/ANCS.2015.7110115","DOIUrl":"https://doi.org/10.1109/ANCS.2015.7110115","url":null,"abstract":"Summary form only given. In this keynote we will present the latest trends in deploying Virtual Network Functions (VNF) for Carrier Networks. The discussion will address the challenges that operators face and their different approaches when virtualizing mobile and fixed networking applications, including EPG, BNG and many others. In addition we will highlight the latest developments in VNF acceleration, including potential acceleration targets and their locations in the Carrier Datacenter. And in conclusion explore the needs of a common acceleration abstraction and the standardization/industry efforts to achieve that goal, including Open Compute Project, Open Data Plane Project, OpenFlow and OpenFlow 2.0, ETSI and OPNFV.","PeriodicalId":186232,"journal":{"name":"2015 ACM/IEEE Symposium on Architectures for Networking and Communications Systems (ANCS)","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116897644","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}
Named Data Networking (NDN) is currently a hot research topic in the field of network architecture, and its transport control mechanism is one of the key technologies needed to be studied. Since the transport in NDN network has the characteristic of multi-source, the implicit congestion detection mechanism of the traditional TCP protocol is no longer suitable for the NDN network. In this paper, we propose a novel congestion control protocol for NDN network based on explicit feedback - ECP (Explicit Control Protocol), which detects the condition of network congestion proactively, and sends explicit feedback to the receiver. According to the feedback, the receiver can adjust the sending rate of Interests in order to control the sending rate of Datas from the sender, thus to realize the congestion control of the network. The simulation results based on NdnSIM show that the ECP protocol performs higher transfer efficiency and stability compared to the current NDN transport protocol using TCP implicit detection mechanism.
命名数据网络(NDN)是当前网络体系结构领域的研究热点，其传输控制机制是需要研究的关键技术之一。由于NDN网络传输具有多源特性，传统TCP协议的隐式拥塞检测机制已不适合NDN网络。本文提出了一种新的基于显式反馈的NDN网络拥塞控制协议——ECP (explicit control protocol)，该协议主动检测网络拥塞状况，并向接收方发送显式反馈。接收端可以根据反馈调整interest的发送速率，从而控制发送端数据的发送速率，从而实现网络的拥塞控制。基于NdnSIM的仿真结果表明，与目前采用TCP隐式检测机制的NDN传输协议相比，ECP协议具有更高的传输效率和稳定性。
{"title":"An interest control protocol for named data networking based on explicit feedback","authors":"Yongmao Ren, Jun Yu Li, Shanshan Shi, Lingling Li, Xiangqing Chang","doi":"10.1109/ANCS.2015.7110139","DOIUrl":"https://doi.org/10.1109/ANCS.2015.7110139","url":null,"abstract":"Named Data Networking (NDN) is currently a hot research topic in the field of network architecture, and its transport control mechanism is one of the key technologies needed to be studied. Since the transport in NDN network has the characteristic of multi-source, the implicit congestion detection mechanism of the traditional TCP protocol is no longer suitable for the NDN network. In this paper, we propose a novel congestion control protocol for NDN network based on explicit feedback - ECP (Explicit Control Protocol), which detects the condition of network congestion proactively, and sends explicit feedback to the receiver. According to the feedback, the receiver can adjust the sending rate of Interests in order to control the sending rate of Datas from the sender, thus to realize the congestion control of the network. The simulation results based on NdnSIM show that the ECP protocol performs higher transfer efficiency and stability compared to the current NDN transport protocol using TCP implicit detection mechanism.","PeriodicalId":186232,"journal":{"name":"2015 ACM/IEEE Symposium on Architectures for Networking and Communications Systems (ANCS)","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117264397","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 de facto implementation of Software Defined Networking (SDN), i.e., OpenFlow, only parses L2-L4 headers, which limits the use of SDN to employ control intelligence in application layer. In this paper, we advocate content parsing to empower SDN with finer grained control ability over traffic. Specifically, we propose a scalable content parser, called COPY, to identify and parse application layer protocols. COPY creates a distinguishable counting context free grammar (DCCFG) to specify the protocol's semantics in application layer, and translates multiple DCCFGs into one distinguishable counting automaton (DCA). DCA is generated without semantic loss from the single DCCFG, and thus provides accurate and scalable parsing ability. Our experiments show that COPY precisely identifies every packet in a labeled trace. When comparing with other six approaches on the real traces, COPY performs 4.2Gb/s and 24.7Gb/s with single- and eight-thread models, respectively, which improves 20%-860% than others, and consumes acceptable offline overhead in time and space.
{"title":"Parsing application layer protocol with commodity hardware for SDN","authors":"Hao Li, Chengchen Hu, Junkai Hong, Xiyu Chen, Yuming Jiang","doi":"10.1109/ANCS.2015.7110120","DOIUrl":"https://doi.org/10.1109/ANCS.2015.7110120","url":null,"abstract":"The de facto implementation of Software Defined Networking (SDN), i.e., OpenFlow, only parses L2-L4 headers, which limits the use of SDN to employ control intelligence in application layer. In this paper, we advocate content parsing to empower SDN with finer grained control ability over traffic. Specifically, we propose a scalable content parser, called COPY, to identify and parse application layer protocols. COPY creates a distinguishable counting context free grammar (DCCFG) to specify the protocol's semantics in application layer, and translates multiple DCCFGs into one distinguishable counting automaton (DCA). DCA is generated without semantic loss from the single DCCFG, and thus provides accurate and scalable parsing ability. Our experiments show that COPY precisely identifies every packet in a labeled trace. When comparing with other six approaches on the real traces, COPY performs 4.2Gb/s and 24.7Gb/s with single- and eight-thread models, respectively, which improves 20%-860% than others, and consumes acceptable offline overhead in time and space.","PeriodicalId":186232,"journal":{"name":"2015 ACM/IEEE Symposium on Architectures for Networking and Communications Systems (ANCS)","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116570623","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}
Pub Date : 2015-05-07DOI: 10.1109/ANCS.2015.7110132
S. Tripathi, R. Chickering, Jonathan Gainsley
In this paper, we describe a distributed network operating system that runs on merchant silicon based switches. Multiple switches can form a fabric for orchestration and state sharing purposes independent of the network topology. Each switch in a fabric shares its state with other switches over TCP/IP and keeps a global view of virtual ports along with topology to make individual switching, forwarding and encapsulation decisions. This allows the distributed switch OS to track Virtual Machines as they migrate and dynamically orchestrate VXLAN overlays without needing software overlays in servers. Since the Switch OS offloads encapsulation/decapsulation for VXLAN to the switch ASIC, there is no performance penalty.
{"title":"Distributed control plane for high performance switchbased VXLAN overlays","authors":"S. Tripathi, R. Chickering, Jonathan Gainsley","doi":"10.1109/ANCS.2015.7110132","DOIUrl":"https://doi.org/10.1109/ANCS.2015.7110132","url":null,"abstract":"In this paper, we describe a distributed network operating system that runs on merchant silicon based switches. Multiple switches can form a fabric for orchestration and state sharing purposes independent of the network topology. Each switch in a fabric shares its state with other switches over TCP/IP and keeps a global view of virtual ports along with topology to make individual switching, forwarding and encapsulation decisions. This allows the distributed switch OS to track Virtual Machines as they migrate and dynamically orchestrate VXLAN overlays without needing software overlays in servers. Since the Switch OS offloads encapsulation/decapsulation for VXLAN to the switch ASIC, there is no performance penalty.","PeriodicalId":186232,"journal":{"name":"2015 ACM/IEEE Symposium on Architectures for Networking and Communications Systems (ANCS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131376565","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}
Pub Date : 2015-05-07DOI: 10.1109/ANCS.2015.7110124
Stefano Garzarella, G. Lettieri, L. Rizzo
Supporting network I/O at high packet rates in virtual machines is fundamental for the deployment of Cloud data centers and Network Function Virtualization. Historically, SR-IOV and hardware passthrough were thought as the only viable solution to reduce the high cost of virtualization. In previous work [15] we showed how even plain device emulation can achieve VM-to-VM speeds of millions of packets per second (Mpps), though still at least 3 times slower than bare metal. In this paper, to fill this gap, we present ptnetmap, a virtual passthrough network device (based on the netmap framework). ptnetmap allows VMs to connect to any netmap port (physical devices, software switches, netmap pipes), conserving the speed and isolation of the native netmap system, and removing the constraints of hardware passthrough. Our work includes two key features not present in previous proposals: we provide a high speed path also to untrusted VMs, and do not require dedicated polling cores/threads, which is fundamental to achieve an efficient use of resources. Besides these features, our speed is also beyond previously published values. Running on top of ptnetmap, VMs can saturate a 10 Gbit link at 14.88 Mpps, talk at over 20 Mpps to untrusted VMs, and over 70 Mpps to trusted VMs. ptnetmap extends the netmap framework, and currently supports Linux and FreeBSD guests, and QEMU/KVM host. Support for bhyve/FreeBSD host is under development.
{"title":"Virtual device passthrough for high speed VM networking","authors":"Stefano Garzarella, G. Lettieri, L. Rizzo","doi":"10.1109/ANCS.2015.7110124","DOIUrl":"https://doi.org/10.1109/ANCS.2015.7110124","url":null,"abstract":"Supporting network I/O at high packet rates in virtual machines is fundamental for the deployment of Cloud data centers and Network Function Virtualization. Historically, SR-IOV and hardware passthrough were thought as the only viable solution to reduce the high cost of virtualization. In previous work [15] we showed how even plain device emulation can achieve VM-to-VM speeds of millions of packets per second (Mpps), though still at least 3 times slower than bare metal. In this paper, to fill this gap, we present ptnetmap, a virtual passthrough network device (based on the netmap framework). ptnetmap allows VMs to connect to any netmap port (physical devices, software switches, netmap pipes), conserving the speed and isolation of the native netmap system, and removing the constraints of hardware passthrough. Our work includes two key features not present in previous proposals: we provide a high speed path also to untrusted VMs, and do not require dedicated polling cores/threads, which is fundamental to achieve an efficient use of resources. Besides these features, our speed is also beyond previously published values. Running on top of ptnetmap, VMs can saturate a 10 Gbit link at 14.88 Mpps, talk at over 20 Mpps to untrusted VMs, and over 70 Mpps to trusted VMs. ptnetmap extends the netmap framework, and currently supports Linux and FreeBSD guests, and QEMU/KVM host. Support for bhyve/FreeBSD host is under development.","PeriodicalId":186232,"journal":{"name":"2015 ACM/IEEE Symposium on Architectures for Networking and Communications Systems (ANCS)","volume":"134 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134188606","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}
Theophilus Wellem, Yu-Kuen Lai, Danny Wen-Yaw Chung
Sketch-based data streaming algorithms are used in many network monitoring applications. The sketch data structure is capable of summarizing high-speed network traffic in real-time and producing accurate estimation result of a specific traffic measurement task. However, due to the fixed hardware implementation of the sketch counter array, the flexibility of adopting the same data structure for various traffic measurement applications is limited. In this work, we propose a scheme for the sketch-based monitoring applications to define the flow keys for updating the sketch data structure in more general settings. The system prototype is implemented on the NetFPGA-10G platform.
{"title":"A software defined sketch system for traffic monitoring","authors":"Theophilus Wellem, Yu-Kuen Lai, Danny Wen-Yaw Chung","doi":"10.5555/2772722.2772755","DOIUrl":"https://doi.org/10.5555/2772722.2772755","url":null,"abstract":"Sketch-based data streaming algorithms are used in many network monitoring applications. The sketch data structure is capable of summarizing high-speed network traffic in real-time and producing accurate estimation result of a specific traffic measurement task. However, due to the fixed hardware implementation of the sketch counter array, the flexibility of adopting the same data structure for various traffic measurement applications is limited. In this work, we propose a scheme for the sketch-based monitoring applications to define the flow keys for updating the sketch data structure in more general settings. The system prototype is implemented on the NetFPGA-10G platform.","PeriodicalId":186232,"journal":{"name":"2015 ACM/IEEE Symposium on Architectures for Networking and Communications Systems (ANCS)","volume":"61 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127058794","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}
Pub Date : 2015-05-07DOI: 10.1109/ANCS.2015.7110119
Ori Rottenstreich, János Tapolcai
Packet classification is a building block in many network services such as routing, filtering, intrusion detection, accounting, monitoring, load-balancing and policy enforcement. Compression has gained attention recently as a way to deal with the expected increase of classifiers size. Typically, compression schemes try to reduce a classifier size while keeping it semantically-equivalent to its original form. Inspired by the advantages of popular compression schemes (e.g. JPEG and MPEG), we study in this paper the applicability of lossy compression to create packet classifiers requiring less memory than optimal semantically-equivalent representations. Our objective is to find a limited-size classifier that can correctly classify a high portion of the traffic so that it can be implemented in commodity switches with classification modules of a given size. We develop optimal dynamic programming based algorithms for several versions of the problem and describe how a small amount of traffic that cannot be classified can be easily treated, especially in software-defined networks. We generalize our solutions for a wide range of classifiers with different similarity metrics. We evaluate their performance on real classifiers and traffic traces and show that in some cases we can reduce a classifier size by orders of magnitude while still classifying almost all traffic correctly.
{"title":"Lossy compression of packet classifiers","authors":"Ori Rottenstreich, János Tapolcai","doi":"10.1109/ANCS.2015.7110119","DOIUrl":"https://doi.org/10.1109/ANCS.2015.7110119","url":null,"abstract":"Packet classification is a building block in many network services such as routing, filtering, intrusion detection, accounting, monitoring, load-balancing and policy enforcement. Compression has gained attention recently as a way to deal with the expected increase of classifiers size. Typically, compression schemes try to reduce a classifier size while keeping it semantically-equivalent to its original form. Inspired by the advantages of popular compression schemes (e.g. JPEG and MPEG), we study in this paper the applicability of lossy compression to create packet classifiers requiring less memory than optimal semantically-equivalent representations. Our objective is to find a limited-size classifier that can correctly classify a high portion of the traffic so that it can be implemented in commodity switches with classification modules of a given size. We develop optimal dynamic programming based algorithms for several versions of the problem and describe how a small amount of traffic that cannot be classified can be easily treated, especially in software-defined networks. We generalize our solutions for a wide range of classifiers with different similarity metrics. We evaluate their performance on real classifiers and traffic traces and show that in some cases we can reduce a classifier size by orders of magnitude while still classifying almost all traffic correctly.","PeriodicalId":186232,"journal":{"name":"2015 ACM/IEEE Symposium on Architectures for Networking and Communications Systems (ANCS)","volume":"41 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117258773","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}
Pub Date : 2015-05-07DOI: 10.1109/ANCS.2015.7110127
J. Garcia-Luna-Aceves, Maziar Mirzazad Barijough
We show that the mechanisms used in the name data networking (NDN) and the original content centric networking (CCN) architectures may not detect Interest loops, even if the network in which they operate is static and no faults occur. Furthermore, we show that no correct Interest forwarding strategy can be defined that allows Interest aggregation and attempts to detect Interest looping by identifying Interests uniquely. We introduce SIFAH (Strategy for Interest Forwarding and Aggregation with Hop-Counts), the first Interest forwarding strategy shown to be correct under any operational conditions of a content centric network. SIFAH operates by having forwarding information bases (FIBs) store the next hops and number of hops to named content, and by having each Interest state the name of the requested content and the hop count from the router forwarding an Interest to the content. We present the results of simulation experiments using the ndnSIM simulator comparing CCN and NDN with SIFAH. The results of these experiments illustrate the negative impact of undetected Interest looping when Interests are aggregated in CCN and NDN, and the performance advantages of using SIFAH.
{"title":"Enabling correct interest forwarding and retransmissions in a content centric network","authors":"J. Garcia-Luna-Aceves, Maziar Mirzazad Barijough","doi":"10.1109/ANCS.2015.7110127","DOIUrl":"https://doi.org/10.1109/ANCS.2015.7110127","url":null,"abstract":"We show that the mechanisms used in the name data networking (NDN) and the original content centric networking (CCN) architectures may not detect Interest loops, even if the network in which they operate is static and no faults occur. Furthermore, we show that no correct Interest forwarding strategy can be defined that allows Interest aggregation and attempts to detect Interest looping by identifying Interests uniquely. We introduce SIFAH (Strategy for Interest Forwarding and Aggregation with Hop-Counts), the first Interest forwarding strategy shown to be correct under any operational conditions of a content centric network. SIFAH operates by having forwarding information bases (FIBs) store the next hops and number of hops to named content, and by having each Interest state the name of the requested content and the hop count from the router forwarding an Interest to the content. We present the results of simulation experiments using the ndnSIM simulator comparing CCN and NDN with SIFAH. The results of these experiments illustrate the negative impact of undetected Interest looping when Interests are aggregated in CCN and NDN, and the performance advantages of using SIFAH.","PeriodicalId":186232,"journal":{"name":"2015 ACM/IEEE Symposium on Architectures for Networking and Communications Systems (ANCS)","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123508503","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}
Pub Date : 2015-05-07DOI: 10.1109/ANCS.2015.7110122
Henrique Rodrigues, Richard D. Strong, A. S. Akyurek, T. Simunic
Dynamic optical interconnects using fast Optical Circuit Switches (OCS) are emerging as a scalable and energy efficient alternative to increasing network demands. Initial concerns regarding slow switching speeds of OCSs were recently overcome, with prototypes enabling circuit setup in a few microseconds. This can potentially broaden the classes of traffic patterns that can be carried efficiently by an alloptical interconnect. However, application performance on such newer interconnects has not been fully understood yet. In this paper, we explore the the gap between advances in faster OCS hardware and the potential success of such newer technologies in terms of application performance and cluster energy efficiency. We evaluate the performance of latencysensitive distributed applications running on a fast OCS environment, analyzing its impact to overall server and network energy efficiency. We also discuss scheduling inefficiencies of current fast OCSs and evaluate ideas to solve them. We find that while some distributed applications suffer minimal performance penalty when running on fast OCSs, more flexible schedulers, like the ones outlined in this paper, improve application performance and OCS efficiency by up to 2.44 times compared to the strategies in the literature. Dynamic optical interconnects using fast Optical Circuit Switches (OCS) are emerging as a scalable and energy efficient alternative to increasing network demands. Initial concerns regarding slow switching speeds of OCSs were recently overcome, with prototypes enabling circuit setup in a few microseconds. This can potentially broaden the classes of traffic patterns that can be carried efficiently by an all-optical interconnect. However, application performance on such newer interconnects has not been fully understood yet. In this paper, we explore the the gap between advances in faster OCS hardware and the potential success of such newer technologies in terms of application performance and cluster energy efficiency. We evaluate the performance of latency-sensitive distributed applications running on a fast OCS environment, analyzing its impact to overall server and network energy efficiency. We also discuss scheduling inefficiencies of current fast OCSs and evaluate ideas to solve them. We find that while some distributed applications suffer minimal performance penalty when running on fast OCSs, more exible schedulers, like the ones outlined in this paper, improve application performance and OCS efficiency by up to 2.44 times compared to the strategies in the literature.
{"title":"Dynamic optical switching for latency sensitive applications","authors":"Henrique Rodrigues, Richard D. Strong, A. S. Akyurek, T. Simunic","doi":"10.1109/ANCS.2015.7110122","DOIUrl":"https://doi.org/10.1109/ANCS.2015.7110122","url":null,"abstract":"Dynamic optical interconnects using fast Optical Circuit Switches (OCS) are emerging as a scalable and energy efficient alternative to increasing network demands. Initial concerns regarding slow switching speeds of OCSs were recently overcome, with prototypes enabling circuit setup in a few microseconds. This can potentially broaden the classes of traffic patterns that can be carried efficiently by an alloptical interconnect. However, application performance on such newer interconnects has not been fully understood yet. In this paper, we explore the the gap between advances in faster OCS hardware and the potential success of such newer technologies in terms of application performance and cluster energy efficiency. We evaluate the performance of latencysensitive distributed applications running on a fast OCS environment, analyzing its impact to overall server and network energy efficiency. We also discuss scheduling inefficiencies of current fast OCSs and evaluate ideas to solve them. We find that while some distributed applications suffer minimal performance penalty when running on fast OCSs, more flexible schedulers, like the ones outlined in this paper, improve application performance and OCS efficiency by up to 2.44 times compared to the strategies in the literature. Dynamic optical interconnects using fast Optical Circuit Switches (OCS) are emerging as a scalable and energy efficient alternative to increasing network demands. Initial concerns regarding slow switching speeds of OCSs were recently overcome, with prototypes enabling circuit setup in a few microseconds. This can potentially broaden the classes of traffic patterns that can be carried efficiently by an all-optical interconnect. However, application performance on such newer interconnects has not been fully understood yet. In this paper, we explore the the gap between advances in faster OCS hardware and the potential success of such newer technologies in terms of application performance and cluster energy efficiency. We evaluate the performance of latency-sensitive distributed applications running on a fast OCS environment, analyzing its impact to overall server and network energy efficiency. We also discuss scheduling inefficiencies of current fast OCSs and evaluate ideas to solve them. We find that while some distributed applications suffer minimal performance penalty when running on fast OCSs, more exible schedulers, like the ones outlined in this paper, improve application performance and OCS efficiency by up to 2.44 times compared to the strategies in the literature.","PeriodicalId":186232,"journal":{"name":"2015 ACM/IEEE Symposium on Architectures for Networking and Communications Systems (ANCS)","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123058035","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}
Pub Date : 2015-05-07DOI: 10.1109/ANCS.2015.7110134
Abhishek Dwaraki, S. Seetharaman, S. Natarajan, T. Wolf
SDN-enabled networks offer a great degree of flexibility, programmability and support for multiple applications. Applications running on top of a SDN controller could further benefit if network state information were made available to them as part of the SDN framework. Our work investigates the requirements for abstracting network state on the control plane. We intend to show that, by meeting these pre-requisites, network state can be made available with the guarantee of state safety and integrity. We envision that our state management abstraction will provide safety in the data plane and aid better programmability in the control plane.
{"title":"State abstraction and management in software-defined networks","authors":"Abhishek Dwaraki, S. Seetharaman, S. Natarajan, T. Wolf","doi":"10.1109/ANCS.2015.7110134","DOIUrl":"https://doi.org/10.1109/ANCS.2015.7110134","url":null,"abstract":"SDN-enabled networks offer a great degree of flexibility, programmability and support for multiple applications. Applications running on top of a SDN controller could further benefit if network state information were made available to them as part of the SDN framework. Our work investigates the requirements for abstracting network state on the control plane. We intend to show that, by meeting these pre-requisites, network state can be made available with the guarantee of state safety and integrity. We envision that our state management abstraction will provide safety in the data plane and aid better programmability in the control plane.","PeriodicalId":186232,"journal":{"name":"2015 ACM/IEEE Symposium on Architectures for Networking and Communications Systems (ANCS)","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122395055","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：