Switches Know the Exact Amount of Congestion

Arslan, Serhat; McKeown, Nick

doi:10.1145/3375235.3375245

Cited by 12 publications

(5 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…[ Arslan and McKeown 2019]. Nesse caso, os pacotes contêm campos de cabec ¸alho que são instruc ¸ões de telemetria, os quais podem ser usados para coletar e gravar dados de rede com maior granularidade.…”

Section: In-band Network Telemetryunclassified

“…Além disso, o nível de ocupac ¸ão do buffer pode afetar outras métricas usadas pelos sistemas de monitoramento, como RTT, atraso e jitter. Coletar dados de buffers em tempo real, no entanto, sempre foi uma tarefa muito cara, devido à necessidade de bits adicionais nos cabec ¸alhos e o aumento do consumo de energia dos dispositivos [Arslan and McKeown 2019].…”

Section: Introduc ¸ãOunclassified

“…Esse cenário mudou nos dispositivos de rede modernos, que com um aumento insignificante de potência ou perda de capacidade, é possível obter informac ¸ões sobre os buffers em tempo real [Arslan and McKeown 2019]. Além disso, com um plano de dados programável e telemetria de rede, é possível observar dados mais precisos da rede.…”

Section: Introduc ¸ãOunclassified

See 2 more Smart Citations

Estimando métricas de serviço através de In-band Network Telemetry

Almeida

Verdi

Pasquini

2021

Anais Do XXXIX Simpósio Brasileiro De Redes De Computadores E Sistemas Distribuídos (SBRC 2021)

View full text Add to dashboard Cite

Recentemente, novas abordagens de telemetria ﬁna na rede, através de In-band network telemetry em equipamentos programáveis, têm entregue novas e precisas informações sobre o estado da rede. Neste contexto, este trabalho apresenta indícios de que é possível utilizar métricas ﬁnas de telemetria de rede em plano de dados programável, em conjunto com métodos de aprendizado de máquina, para estimar métricas de qualidade de serviço. Um prova de conceito minimalista foi realizada e resultados preliminares indicam que, com o auxílio de algoritmos de aprendizado de máquina, é possível estimar métricas de um serviço de vídeo a partir de métricas ﬁnas relacionadas aos buffers dos switches.

show abstract

Section: In-band Network Telemetryunclassified

Section: Introduc ¸ãOunclassified

See 1 more Smart Citation

Estimando métricas de serviço através de In-band Network Telemetry

Almeida

Verdi

Pasquini

2021

Anais Do XXXIX Simpósio Brasileiro De Redes De Computadores E Sistemas Distribuídos (SBRC 2021)

View full text Add to dashboard Cite

show abstract

“…Recent progress in programmable hardware and the utilization of the P4 language [13] have enabled network devices to autonomously report the network's state, eliminating the need for direct control plane intervention [14]. In this scenario, packets incorporate telemetry instructions within their header fields, facilitating the fine-grained collection and recording of network data.…”

Section: In-band Network Telemetrymentioning

confidence: 99%

Using Machine Learning and In-band Network Telemetry for Service Metrics Estimation

Almeida

Pasquini

Verdi

2021

2021 IEEE 10th International Conference on Cloud Networking (CloudNet)

View full text Add to dashboard Cite

Active Queue Management (AQM) is a mechanism employed to alleviate transient congestion in network device buffers, such as routers and switches. Traditional AQM algorithms use fixed thresholds, like target delay or queue occupancy, to compute random packet drop probabilities. A very small target delay can increase packet losses and reduce link utilization, while a large target delay may increase queueing delays while lowering drop probability. Due to dynamic network traffic characteristics, where traffic fluctuations can lead to significant queue variations, maintaining a fixed threshold AQM may not suit all applications. Consequently, we explore the question: What is the ideal threshold (target delay) for AQMs? In this work, we introduce DESiRED (Dynamic, Enhanced, and Smart iRED), a P4-based AQM that leverages precise network feedback from In-band Network Telemetry (INT) to feed a Deep Reinforcement Learning (DRL) model. This model dynamically adjusts the target delay based on rewards that maximize application Quality of Service (QoS). We evaluate DESiRED in a realistic P4based test environment running an MPEG-DASH service. Our findings demonstrate up to a 90x reduction in video stall and a 42x increase in high-resolution video playback quality when the target delay is adjusted dynamically by DE-SiRED.

show abstract

“…New specialized intensive I/Ocapable programmable data planes [24] are suitable for data-heavy operations. However, most approaches on distributed executions of data analytics to date ignore the network altogether and at best use observable noisy signals [2] such as latency or packet loss without insight on the causes and the underlying infrastructure.…”

Section: Introductionmentioning

confidence: 99%

Accelerating the performance of data analytics using network-centric processing

Boughzala

Koldehofe

2021

Proceedings of the 15th ACM International Conference on Distributed and Event-Based Systems

View full text Add to dashboard Cite

Distributed execution of real-time data analytics such as event stream processing is the key to scalability, performance and reliable detection of situation changes. Although real-time analytics is highly I/O centric, existing methods supporting the efficient execution of data analytics functions mostly rely on traditional compute models that are available in data centers, e.g., CPU or GPU based processing models, but treat the network mainly as a blackbox. However, with recent advance in software-defined networking (SDN) and the standardization of packet processing pipeline, data analytics functions can be offloaded to programmable switches and benefit from hardware acceleration in an easier and more flexible way than a decade ago. In this paper we focus on the potential of in-network processing to enhance the performance of the overall real-time data analytics application. We aim to contribute to an (i) understanding on how in-network processing can accelerate real-time data analytics and (ii) assess what models of in-network computing can accelerate which event processing functions considering the limitations of network models compared to traditional compute models. We motivate the potential and illustrate the research problems in the context of load balancing which is an important concept in the data-parallel execution of event processing systems. CCS CONCEPTS• Networks → Programmable networks, In-network processing; • Software and its engineering → Publish-subscribe eventbased architectures, Message oriented middleware.

show abstract

Switches Know the Exact Amount of Congestion

Cited by 12 publications

References 18 publications

Estimando métricas de serviço através de In-band Network Telemetry

Estimando métricas de serviço através de In-band Network Telemetry

Using Machine Learning and In-band Network Telemetry for Service Metrics Estimation

Accelerating the performance of data analytics using network-centric processing

Contact Info

Product

Resources

About