Hispanopedia Wiki
Articulo de referencia

Active queue management

In routers and switches , active queue management ( AQM ) is the policy of dropping packets inside a buffer associated with a network interface controller (NIC) before that buff...

Hispanopedia WikiContenido en espanolLectura gratuita

In routers and switches, active queue management (AQM) is the policy of dropping packets inside a buffer associated with a network interface controller (NIC) before that buffer becomes full, often with the goal of reducing network congestion or improving end-to-end latency. This task is performed by the network scheduler, which for this purpose uses various algorithms such as random early detection (RED), Explicit Congestion Notification (ECN), or controlled delay (CoDel). RFC 7567 recommends active queue management as a best practice.

Overview

An Internet router typically maintains a set of queues, one or more per interface, that hold packets scheduled to go out on that interface. Historically, such queues use a drop-tail discipline: a packet is put onto the queue if the queue is shorter than its maximum size (measured in packets or in bytes), and dropped otherwise.

Active queue disciplines drop or mark packets before the queue is full. Typically, they operate by maintaining one or more drop/mark probabilities, and occasionally dropping or marking packets according to the probabilities before the queue is full.

Benefits

Drop-tail queues have a tendency to penalise bursty flows, and to cause global synchronization between flows. By dropping packets probabilistically, AQM disciplines typically avoid both of these issues.[1]

By providing endpoints with congestion indication before the queue is full, AQM disciplines are able to maintain a shorter queue length than drop-tail queues, which combats bufferbloat and reduces network latency.

Drawbacks

Early AQM disciplines (notably RED and SRED) require careful tuning of their parameters in order to provide good performance. These systems are not optimally behaved from a control theory perspective.[2] Modern AQM disciplines (ARED, Blue, PI, CoDel, CAKE) are self-tuning, and can be run with their default parameters in most circumstances.

Network engineers have historically been trained to avoid packet loss, and have therefore sometimes been critical of AQM systems that drop packets: "Why should I drop perfectly good packets when I still have free buffer space?"[3]

Simulation

Se ha establecido una plataforma de simulación de gestión activa de colas y denegación de servicio (AQM&DoS) basada en el código de simulación NS-2 del algoritmo RRED . Esta plataforma puede simular diversos ataques DoS (DoS distribuido, DoS por suplantación de identidad, DoS de baja tasa, etc.) y algoritmos AQM (RED, RRED , SFB, etc.). Calcula y registra automáticamente el rendimiento promedio de los flujos TCP normales antes y después de los ataques DoS para facilitar el análisis del impacto de estos ataques en los flujos TCP normales y los algoritmos AQM. [ 4 ]

Algoritmos de gestión de colas activas

Referencias

  1. ↑ Floyd, Sally; Jacobson, Van (agosto de 1993). "Pasarelas de detección temprana aleatoria (RED) para evitar la congestión" . IEEE/ACM Transactions on Networking . 1 (4): 397– 413. CiteSeerX 10.1.1.147.3833 . doi : 10.1109/90.251892 . S2CID 221977646. Recuperado el 16 de marzo de 2008 .  
  2. ↑ CV Hollot, Vishal Misra, Don Towsley y Wei-Bo Gong, Análisis y diseño de controladores para enrutadores AQM que admiten flujos TCP
  3. ↑ "El genio criminal: ¡bufferbloat!" . Divagaciones de jg . Gettys.wordpress.com. 3 de diciembre de 2010 . Consultado el 23 de enero de 2014 .
  4. ^ "Plataforma de simulación AQM y DoS: Changwang Zhang" .
  5. ↑ Imer, Orhan; ​​Basar, Tamer (diciembre de 2005). "Un resultado de estabilidad global en el control de congestión basado en la utilidad" (PDF) . Actas de la 44.ª Conferencia IEEE sobre Decisión y Control y la Conferencia Europea de Control 2005 : 6740–6743 .
  6. ↑ Athuraliya, S. (junio de 2001). "REM: Gestión activa de colas". IEEE Transactions on Networking . 15 (3): 48– 53.
  7. ↑ Ratul Mahajan, Sally Floyd y David Wetherall, Control de flujos de alto ancho de banda en el enrutador congestionado , ICNP 2001
  8. ↑ Changwang Zhang, Jianping Yin, Zhiping Cai y Weifeng Chen, RRED: Algoritmo RED robusto para contrarrestar ataques de denegación de servicio de baja frecuencia , IEEE Communications Letters, vol. 14, págs. 489-491, 2010. Ref.
  9. ↑ Changwang Zhang, Jianping Yin y Zhiping Cai, RSFB: un algoritmo estocástico justo azul resiliente contra ataques DDoS de suplantación de identidad , en Simposio Internacional sobre Comunicación y Tecnología de la Información (ISCIT), 2009. Ref.
  10. ↑ "Gestión inteligente de colas" . bufferbloat.net . Consultado el 3 de octubre de 2020 .
  11. ↑ "OpenWrt·Traffic Shaping·SQM" . OpenWrt.org . Proyecto OpenWrt . 2 de enero de 2017. Consultado el 3 de octubre de 2020 .