2018 14th International Conference on Network and Service Management (CNSM), 2018
The dynamically changing landscape of DDoS threats increases the demand for advanced security sol... more The dynamically changing landscape of DDoS threats increases the demand for advanced security solutions. The rise of massive IoT botnets enables attackers to mount high-intensity short-duration ”volatile ephemeral” attack waves in quick succession. Therefore the standard human-in-the-loop security center paradigm is becoming obsolete. To battle the new breed of volatile DDoS threats, the intrusion detection system (IDS) needs to improve markedly, at least in reaction times and in automated response (mitigation). Designing such an IDS is a daunting task as network operators are traditionally reluctant to act – at any speed – on potentially false alarms. The primary challenge of a low reaction time detection system is maintaining a consistently low false alarm rate. This paper aims to show how a practical FPGA-based DDoS detection and mitigation system can successfully address this. Besides verifying the model and algorithms with real traffic ”in the wild”, we validate the low false a...
HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific r... more HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés. ENDEAVOUR: Towards a flexible software-defined network ecosystemD4.8: Final Implementation of Selected Use Cases for IXP Members Daniel Kopp, Christoph Dietzel, Marco Chiesa, Rémy Lapeyrade, Philippe Owezarski
2017 13th International Conference on Network and Service Management (CNSM), 2017
While the scale, frequency and impact of the recent cyber-and DoS-attacks have all increased, the... more While the scale, frequency and impact of the recent cyber-and DoS-attacks have all increased, the traditional security management systems are still supervised by human operators in the decisional loop. To cope with the new breed of machine-driven attacks-particularly those designed to overload the humans in the loop-the next-generation anomaly detection and attack mitigation schema, i.e. the network security management, must improve greatly in speed and accuracy: become machine-driven, too. As infrastructure we propose an FPGA-accelerated Network Function Virtualization that potentially enhances the current multi-Tbps switching fabrics with SDN-based security capabilities of vastly higher performance and scalability. As key novelties, we contribute (i) sub-ms detection lag (ii) of the top 9 Akamai attacks [1] with (iii) a real-time SDN feedback loop between a distributed programmable data plane and a centralized SDN controller, (iv) coupled via a global N:1 mirror. We validate the concept in an actual datacenter network with a new security application that can detect and mitigate real-world dDoS attacks, with lags from 430 us up to 3 ms-several orders of magnitude faster than before.
2017 16th IEEE International Conference on Machine Learning and Applications (ICMLA), 2017
Unsupervised anomaly detection (AD) has shown promise against the frequently new cyberattacks. Bu... more Unsupervised anomaly detection (AD) has shown promise against the frequently new cyberattacks. But, as anomalies are not always malicious, such systems generate prodigious false alarm rates. The resulting manual validation workload often overwhelms the IT operators: it slows down the system reaction by orders of magnitude and ultimately thwarts its applicability. Therefore, we propose a real-time network AD system that reduces the manual workload by coupling 2 learning stages. The first stage performs adaptive unsupervised AD using a shallow autoencoder. The second stage uses a custom nearest-neighbor classifier to filter the false positives by modeling the manual classification. We implement a prototype for 10-50Gbps speeds and evaluate it with traffic from a national network operator: we achieve 98.5% true and 1.3% false positive rates, while reducing the human intervention rate by 5x.
While innovation in inter-domain routing has remained stagnant for over a decade, Internet Exchan... more While innovation in inter-domain routing has remained stagnant for over a decade, Internet Exchange Points (IXPs) are consolidating their role as economically advantageous interconnection points for reducing path latencies and exchanging ever increasing amounts of traffic. As such, IXPs appear as a natural place to foster network innovation and assess the benefits of Software-Defined Networking (SDN), a recent technological trend that has already boosted innovation within data-center networks. In this paper, we give a comprehensive overview of use cases for SDN at IXPs, which leverage the superior vantage point of an IXP to introduce advanced features like load-balancing and DDoS mitigation. We discuss the benefits of SDN solutions by analyzing real-world data from one of the largest IXPs. We also leverage insights into IXP operations to not only shape benefits for members but also for operators.
Concurrency and Computation: Practice and Experience, 2017
SummaryThe Graph500 benchmark attempts to steer the design of High‐Performance Computing systems ... more SummaryThe Graph500 benchmark attempts to steer the design of High‐Performance Computing systems to maximize the performance under memory‐constricted application workloads. A realistic simulation of such benchmarks for architectural research is challenging due to size and detail limitations. By contrast, synthetic traffic workloads constitute one of the least resource‐consuming methods to evaluate the performance. In this work, we provide a simulation tool for network architects that need to evaluate the suitability of their interconnect for BigData applications. Our development is a low computation‐ and memory‐demanding synthetic traffic model that emulates the behavior of the Graph500 communications and is publicly available in an open‐source network simulator. The characterization of network traffic is inferred from a profile of several executions of the benchmark with different input parameters. We verify the validity of the equations in our model against an execution of the ben...
Algorithms and Architectures for Parallel Processing, 2016
As BigData applications have gained momentum over the last years, the Graph500 benchmark has appe... more As BigData applications have gained momentum over the last years, the Graph500 benchmark has appeared in an attempt to steer the design of HPC systems to maximize the performance under memoryconstricted application workloads. A realistic simulation of such benchmarks for architectural research is challenging due to size and detail limitations, and synthetic traffic workloads constitute one of the least resource-consuming methods to evaluate the performance. In this work, we propose a synthetic traffic model that emulates the behavior of the Graph500 communications. Our model is empirically obtained through a characterization of several executions of the benchmark with different input parameters. We verify the validity of our model against a characterization of the execution of the benchmark with different parameters. Our model is well-suited for implementation in an architectural simulator.
Proceedings of the 2013 Interconnection Network Architecture: On-Chip, Multi-Chip, 2013
ABSTRACT Key to the economic viability of clouds and datacenters is their elastic scalability. Th... more ABSTRACT Key to the economic viability of clouds and datacenters is their elastic scalability. Therefore most active related research areas focus on the datacenter fabric scalability, efficiency, performance, virtualization, optimal virtual machine (VM) allocation and migration. Here we ask the questions: Given a set of tenant workloads running on generic servers interconnected by a 10--100G Ethernet fabric with modern network virtualization and transport protocols, how can the datacenter operator reach the optimal operation region? How is this optimum defined, traded between operator and tenants, and measured with what metrics? In this paper we propose an evaluation methodology and a set of simple, but descriptive, metrics as a first attempt to answer the questions raised above. As proof of concept, we investigate a multitenant virtualized datacenter network running a 3-tier workload. Our proposal enables a quantitative comparison between competing datacenter fabrics and virtualization architectures.
13th Symposium on High Performance Interconnects (HOTI'05), 2005
Driving computer interconnection networks closer to saturation minimizes cost/performance and pow... more Driving computer interconnection networks closer to saturation minimizes cost/performance and power consumption, but requires efficient congestion control to prevent catastrophic performance degradation during traffic peaks or "hot spot" traffic patterns. The In-finiBand™ Architecture provides such congestion control, but lacks guidance for setting its parameters. At its adoption, it was unproven that there were any settings that would work at all, avoid instability or oscillations. This paper reports on a simulation-driven exploration of that parameter space which verifies that the architected scheme can, in fact, work properly despite inherent delays in its feedback mechanism.
13th Symposium on High Performance Interconnects (HOTI'05)
A crucial part of any high-performance computing system is its interconnection network. In the OS... more A crucial part of any high-performance computing system is its interconnection network. In the OSMOSIS project, Corning and IBM are jointly developing a demonstrator interconnect based on optical cell switching with electronic control. Starting from the core set of requirements, we present the system design rationale and show how it impacts the practical implementation. Our focus is on solving the technical issues related to the electronic control path, and we show that it is feasible at the targeted design point.
2014 IEEE 15th International Conference on High Performance Switching and Routing (HPSR), 2014
Performance-optimized datacenter networks aim to handle more efficiently the growing East-West in... more Performance-optimized datacenter networks aim to handle more efficiently the growing East-West intra-cluster traffic of BigData applications. The demanding latency constraints and traffic patterns of these applications expose the inherent bottlenecks of the often oversubscribed datacenter network topologies, favoring in stead the full-bisectional bandwidth fat-trees. And yet their topological benefits may remain unrealized in practical deployments, if such fabrics use single path or flow-level (ECMP hashing) multipath routing. Here we model in detail on Layer 2 the routing performance of modern fat-tree networks using stochastic permutations of bursty traffic. We first analytically simplify and then validate by accurate simulation models that the throughputs for 'static' d-mod-k and for ECMP-like multipath routing are 63% and 47%, respectively. We also find that ECMP routing results in a wide spread of link loads under random permutation traffic, which manifests as a 3x throughput reduction for 30% of the flows. Furthermore, ECMP can lead to collisions of mouse and elephant flows, often increasing the flow completion time (FCT) of delay-sensitive flows by a factor of 10. In contrast, packet-based multipath outperforms all the others in this study.. I. MOTIVATION Datacenter-based clouds employ copious amounts of distributed resources-endnodes, storage, networks-to perform BigData tasks impossible on monolithic systems or classic supercomputers. A notable peculiarity is the growing East-West traffic volume with latency constraints. Therefore the Performance-optimized Datacenter (PoD) interconnects are evolving from an established hierarchical model of access/aggregator/core, toward a 2-tiered, i.e., spine-leaf, topology. This shift circumvents the bandwidth partitioning inherent in hierarchical topologies, while promising performance robustness and delay distribution 'flatness': Uniform, quasiequidistant source-destination paths. However, this abundance of paths raises the issue of routing performance. The prevalent method today is (Equal Cost Multi-Path) ECMP-based: Each routing path is uniquely selected by hash-keying on a combination of L4, L3 and/or L2 addresses [1], [2], [3]-most often a 5-tuple hash covering Ethernet, IP and TCP header fields. While load-oblivious, the obvious merits of ECMP are its order preservation and relative simplicity. However, as any coarse-grained stochastic scheme, ECMP is exposed to the 'unlucky' permutations in the offered load traffic matrix, which can and will lead to a degree of flow interference and blocking effects in the network, with nefarious impacts on the workload performance. Here we contribute their quantitative analysis in fat-tree (spine-leaf) datacenter networks, as shown in Fig. 1.
Proceedings 1998 Design and Automation Conference. 35th DAC. (Cat. No.98CH36175)
This paper describes the design and implementation of the NUMAchine multiprocessor. As the market... more This paper describes the design and implementation of the NUMAchine multiprocessor. As the market for CC-NUMA multiprocessors expands, this research project provides a timely architectural design and cost-effective prototype. The key to the successful implementation of our 48-processor prototype is the use of off-the-shelf components and programmable logic devices. Since this machine will serve as a research vehicle for parallel software development, a number of hardware features to enhance experimentation have been included in the design.
Proceedings of the 8th International Workshop on Interconnection Network Architecture: On-Chip, Multi-Chip, 2014
ABSTRACT Performance optimized datacenters (PoDs) require efficient PoD interconnects to deal wit... more ABSTRACT Performance optimized datacenters (PoDs) require efficient PoD interconnects to deal with the increasing volumes of inter-server (east-west) traffic. To cope with these stringent traffic patterns, datacenter networks are abandoning the oversubscribed topologies of the past, and move towards full-bisection fat-tree fabrics. However, these fabrics typically employ either single-path or coarse-grained (flow-level) multi-path routing. In this paper, we use computer simulations and analysis to characterize the waste of bandwidth that is due to routing inefficiencies. Our analysis suggests that, under a randomly selected permutation, the expected throughputs of d-mod-k routing and of flow-level multi-path routing are close to 63% and 47%, respectively. Furthermore, nearly 30% of the flows are expected to undergo an unnecessary 3-fold slowdown. By contrast, packet-level multi-path routing consistently delivers full throughput to all flows, and proactively avoids internal hotspots, thus serving better the growing demands of inter-server (east-west) traffic.
2013 IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS), 2013
Current hypervisor software drops packets in the virtual network. This behavior is suboptimal, wa... more Current hypervisor software drops packets in the virtual network. This behavior is suboptimal, wastes network resources and harms performance. We propose to extend the flow control mechanisms that currently exist only in physical networks to the virtual networks. Using a simple setup we show the advantages of losslessness in a virtualized environment.
2011 IEEE 19th Annual Symposium on High Performance Interconnects, 2011
Page 1. Short and Fat: TCP Performance in CEE Datacenter Networks Daniel Crisan, Andreea S. Anghe... more Page 1. Short and Fat: TCP Performance in CEE Datacenter Networks Daniel Crisan, Andreea S. Anghel, Robert Birke, Cyriel Minkenberg and Mitch Gusat IBM Research, Zürich Research Laboratory Säumerstrasse 4, CH-8803 ...
2018 14th International Conference on Network and Service Management (CNSM), 2018
The dynamically changing landscape of DDoS threats increases the demand for advanced security sol... more The dynamically changing landscape of DDoS threats increases the demand for advanced security solutions. The rise of massive IoT botnets enables attackers to mount high-intensity short-duration ”volatile ephemeral” attack waves in quick succession. Therefore the standard human-in-the-loop security center paradigm is becoming obsolete. To battle the new breed of volatile DDoS threats, the intrusion detection system (IDS) needs to improve markedly, at least in reaction times and in automated response (mitigation). Designing such an IDS is a daunting task as network operators are traditionally reluctant to act – at any speed – on potentially false alarms. The primary challenge of a low reaction time detection system is maintaining a consistently low false alarm rate. This paper aims to show how a practical FPGA-based DDoS detection and mitigation system can successfully address this. Besides verifying the model and algorithms with real traffic ”in the wild”, we validate the low false a...
HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific r... more HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés. ENDEAVOUR: Towards a flexible software-defined network ecosystemD4.8: Final Implementation of Selected Use Cases for IXP Members Daniel Kopp, Christoph Dietzel, Marco Chiesa, Rémy Lapeyrade, Philippe Owezarski
2017 13th International Conference on Network and Service Management (CNSM), 2017
While the scale, frequency and impact of the recent cyber-and DoS-attacks have all increased, the... more While the scale, frequency and impact of the recent cyber-and DoS-attacks have all increased, the traditional security management systems are still supervised by human operators in the decisional loop. To cope with the new breed of machine-driven attacks-particularly those designed to overload the humans in the loop-the next-generation anomaly detection and attack mitigation schema, i.e. the network security management, must improve greatly in speed and accuracy: become machine-driven, too. As infrastructure we propose an FPGA-accelerated Network Function Virtualization that potentially enhances the current multi-Tbps switching fabrics with SDN-based security capabilities of vastly higher performance and scalability. As key novelties, we contribute (i) sub-ms detection lag (ii) of the top 9 Akamai attacks [1] with (iii) a real-time SDN feedback loop between a distributed programmable data plane and a centralized SDN controller, (iv) coupled via a global N:1 mirror. We validate the concept in an actual datacenter network with a new security application that can detect and mitigate real-world dDoS attacks, with lags from 430 us up to 3 ms-several orders of magnitude faster than before.
2017 16th IEEE International Conference on Machine Learning and Applications (ICMLA), 2017
Unsupervised anomaly detection (AD) has shown promise against the frequently new cyberattacks. Bu... more Unsupervised anomaly detection (AD) has shown promise against the frequently new cyberattacks. But, as anomalies are not always malicious, such systems generate prodigious false alarm rates. The resulting manual validation workload often overwhelms the IT operators: it slows down the system reaction by orders of magnitude and ultimately thwarts its applicability. Therefore, we propose a real-time network AD system that reduces the manual workload by coupling 2 learning stages. The first stage performs adaptive unsupervised AD using a shallow autoencoder. The second stage uses a custom nearest-neighbor classifier to filter the false positives by modeling the manual classification. We implement a prototype for 10-50Gbps speeds and evaluate it with traffic from a national network operator: we achieve 98.5% true and 1.3% false positive rates, while reducing the human intervention rate by 5x.
While innovation in inter-domain routing has remained stagnant for over a decade, Internet Exchan... more While innovation in inter-domain routing has remained stagnant for over a decade, Internet Exchange Points (IXPs) are consolidating their role as economically advantageous interconnection points for reducing path latencies and exchanging ever increasing amounts of traffic. As such, IXPs appear as a natural place to foster network innovation and assess the benefits of Software-Defined Networking (SDN), a recent technological trend that has already boosted innovation within data-center networks. In this paper, we give a comprehensive overview of use cases for SDN at IXPs, which leverage the superior vantage point of an IXP to introduce advanced features like load-balancing and DDoS mitigation. We discuss the benefits of SDN solutions by analyzing real-world data from one of the largest IXPs. We also leverage insights into IXP operations to not only shape benefits for members but also for operators.
Concurrency and Computation: Practice and Experience, 2017
SummaryThe Graph500 benchmark attempts to steer the design of High‐Performance Computing systems ... more SummaryThe Graph500 benchmark attempts to steer the design of High‐Performance Computing systems to maximize the performance under memory‐constricted application workloads. A realistic simulation of such benchmarks for architectural research is challenging due to size and detail limitations. By contrast, synthetic traffic workloads constitute one of the least resource‐consuming methods to evaluate the performance. In this work, we provide a simulation tool for network architects that need to evaluate the suitability of their interconnect for BigData applications. Our development is a low computation‐ and memory‐demanding synthetic traffic model that emulates the behavior of the Graph500 communications and is publicly available in an open‐source network simulator. The characterization of network traffic is inferred from a profile of several executions of the benchmark with different input parameters. We verify the validity of the equations in our model against an execution of the ben...
Algorithms and Architectures for Parallel Processing, 2016
As BigData applications have gained momentum over the last years, the Graph500 benchmark has appe... more As BigData applications have gained momentum over the last years, the Graph500 benchmark has appeared in an attempt to steer the design of HPC systems to maximize the performance under memoryconstricted application workloads. A realistic simulation of such benchmarks for architectural research is challenging due to size and detail limitations, and synthetic traffic workloads constitute one of the least resource-consuming methods to evaluate the performance. In this work, we propose a synthetic traffic model that emulates the behavior of the Graph500 communications. Our model is empirically obtained through a characterization of several executions of the benchmark with different input parameters. We verify the validity of our model against a characterization of the execution of the benchmark with different parameters. Our model is well-suited for implementation in an architectural simulator.
Proceedings of the 2013 Interconnection Network Architecture: On-Chip, Multi-Chip, 2013
ABSTRACT Key to the economic viability of clouds and datacenters is their elastic scalability. Th... more ABSTRACT Key to the economic viability of clouds and datacenters is their elastic scalability. Therefore most active related research areas focus on the datacenter fabric scalability, efficiency, performance, virtualization, optimal virtual machine (VM) allocation and migration. Here we ask the questions: Given a set of tenant workloads running on generic servers interconnected by a 10--100G Ethernet fabric with modern network virtualization and transport protocols, how can the datacenter operator reach the optimal operation region? How is this optimum defined, traded between operator and tenants, and measured with what metrics? In this paper we propose an evaluation methodology and a set of simple, but descriptive, metrics as a first attempt to answer the questions raised above. As proof of concept, we investigate a multitenant virtualized datacenter network running a 3-tier workload. Our proposal enables a quantitative comparison between competing datacenter fabrics and virtualization architectures.
13th Symposium on High Performance Interconnects (HOTI'05), 2005
Driving computer interconnection networks closer to saturation minimizes cost/performance and pow... more Driving computer interconnection networks closer to saturation minimizes cost/performance and power consumption, but requires efficient congestion control to prevent catastrophic performance degradation during traffic peaks or "hot spot" traffic patterns. The In-finiBand™ Architecture provides such congestion control, but lacks guidance for setting its parameters. At its adoption, it was unproven that there were any settings that would work at all, avoid instability or oscillations. This paper reports on a simulation-driven exploration of that parameter space which verifies that the architected scheme can, in fact, work properly despite inherent delays in its feedback mechanism.
13th Symposium on High Performance Interconnects (HOTI'05)
A crucial part of any high-performance computing system is its interconnection network. In the OS... more A crucial part of any high-performance computing system is its interconnection network. In the OSMOSIS project, Corning and IBM are jointly developing a demonstrator interconnect based on optical cell switching with electronic control. Starting from the core set of requirements, we present the system design rationale and show how it impacts the practical implementation. Our focus is on solving the technical issues related to the electronic control path, and we show that it is feasible at the targeted design point.
2014 IEEE 15th International Conference on High Performance Switching and Routing (HPSR), 2014
Performance-optimized datacenter networks aim to handle more efficiently the growing East-West in... more Performance-optimized datacenter networks aim to handle more efficiently the growing East-West intra-cluster traffic of BigData applications. The demanding latency constraints and traffic patterns of these applications expose the inherent bottlenecks of the often oversubscribed datacenter network topologies, favoring in stead the full-bisectional bandwidth fat-trees. And yet their topological benefits may remain unrealized in practical deployments, if such fabrics use single path or flow-level (ECMP hashing) multipath routing. Here we model in detail on Layer 2 the routing performance of modern fat-tree networks using stochastic permutations of bursty traffic. We first analytically simplify and then validate by accurate simulation models that the throughputs for 'static' d-mod-k and for ECMP-like multipath routing are 63% and 47%, respectively. We also find that ECMP routing results in a wide spread of link loads under random permutation traffic, which manifests as a 3x throughput reduction for 30% of the flows. Furthermore, ECMP can lead to collisions of mouse and elephant flows, often increasing the flow completion time (FCT) of delay-sensitive flows by a factor of 10. In contrast, packet-based multipath outperforms all the others in this study.. I. MOTIVATION Datacenter-based clouds employ copious amounts of distributed resources-endnodes, storage, networks-to perform BigData tasks impossible on monolithic systems or classic supercomputers. A notable peculiarity is the growing East-West traffic volume with latency constraints. Therefore the Performance-optimized Datacenter (PoD) interconnects are evolving from an established hierarchical model of access/aggregator/core, toward a 2-tiered, i.e., spine-leaf, topology. This shift circumvents the bandwidth partitioning inherent in hierarchical topologies, while promising performance robustness and delay distribution 'flatness': Uniform, quasiequidistant source-destination paths. However, this abundance of paths raises the issue of routing performance. The prevalent method today is (Equal Cost Multi-Path) ECMP-based: Each routing path is uniquely selected by hash-keying on a combination of L4, L3 and/or L2 addresses [1], [2], [3]-most often a 5-tuple hash covering Ethernet, IP and TCP header fields. While load-oblivious, the obvious merits of ECMP are its order preservation and relative simplicity. However, as any coarse-grained stochastic scheme, ECMP is exposed to the 'unlucky' permutations in the offered load traffic matrix, which can and will lead to a degree of flow interference and blocking effects in the network, with nefarious impacts on the workload performance. Here we contribute their quantitative analysis in fat-tree (spine-leaf) datacenter networks, as shown in Fig. 1.
Proceedings 1998 Design and Automation Conference. 35th DAC. (Cat. No.98CH36175)
This paper describes the design and implementation of the NUMAchine multiprocessor. As the market... more This paper describes the design and implementation of the NUMAchine multiprocessor. As the market for CC-NUMA multiprocessors expands, this research project provides a timely architectural design and cost-effective prototype. The key to the successful implementation of our 48-processor prototype is the use of off-the-shelf components and programmable logic devices. Since this machine will serve as a research vehicle for parallel software development, a number of hardware features to enhance experimentation have been included in the design.
Proceedings of the 8th International Workshop on Interconnection Network Architecture: On-Chip, Multi-Chip, 2014
ABSTRACT Performance optimized datacenters (PoDs) require efficient PoD interconnects to deal wit... more ABSTRACT Performance optimized datacenters (PoDs) require efficient PoD interconnects to deal with the increasing volumes of inter-server (east-west) traffic. To cope with these stringent traffic patterns, datacenter networks are abandoning the oversubscribed topologies of the past, and move towards full-bisection fat-tree fabrics. However, these fabrics typically employ either single-path or coarse-grained (flow-level) multi-path routing. In this paper, we use computer simulations and analysis to characterize the waste of bandwidth that is due to routing inefficiencies. Our analysis suggests that, under a randomly selected permutation, the expected throughputs of d-mod-k routing and of flow-level multi-path routing are close to 63% and 47%, respectively. Furthermore, nearly 30% of the flows are expected to undergo an unnecessary 3-fold slowdown. By contrast, packet-level multi-path routing consistently delivers full throughput to all flows, and proactively avoids internal hotspots, thus serving better the growing demands of inter-server (east-west) traffic.
2013 IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS), 2013
Current hypervisor software drops packets in the virtual network. This behavior is suboptimal, wa... more Current hypervisor software drops packets in the virtual network. This behavior is suboptimal, wastes network resources and harms performance. We propose to extend the flow control mechanisms that currently exist only in physical networks to the virtual networks. Using a simple setup we show the advantages of losslessness in a virtualized environment.
2011 IEEE 19th Annual Symposium on High Performance Interconnects, 2011
Page 1. Short and Fat: TCP Performance in CEE Datacenter Networks Daniel Crisan, Andreea S. Anghe... more Page 1. Short and Fat: TCP Performance in CEE Datacenter Networks Daniel Crisan, Andreea S. Anghel, Robert Birke, Cyriel Minkenberg and Mitch Gusat IBM Research, Zürich Research Laboratory Säumerstrasse 4, CH-8803 ...
Uploads
Papers by Mitch Gusat