https://DevOpsCloud.io -- Cloud Monk Losang Jinpa, Ph.D., MCSE/MCT, GitOps DevOps Engineer

Return to RFCs Related to Artificial Intelligence, Security-Related RFCs, Network Security, Container Security - Kubernetes Security, Cloud Security, Web Security, DevSecOps

See: 7567 on datatracker.ietf.org

The title of this RFC is “IETF Recommendations Regarding Active Queue Management (RFC 7567).”

RFC 7567 provides recommendations on the use of Active Queue Management (AQM) techniques to manage network congestion. The primary goal of AQM is to reduce network delays, improve throughput, and minimize packet loss during times of congestion by controlling how packets are queued and dropped in routers and network devices. This RFC revisits earlier practices and suggests updated methods to improve network performance in modern networks, emphasizing the importance of active queue management over passive, tail-drop queuing methods that often lead to increased latency and poor network performance. The related RFC is RFC 2309, which initially discussed the concept of AQM and its benefits. https://en.wikipedia.org/wiki/Active_queue_management https://tools.ietf.org/html/rfc2309

One of the central recommendations of RFC 7567 is the promotion of AQM algorithms such as Random Early Detection (RED) and its variants, which proactively drop packets before a queue becomes full. This prevents the build-up of large queues and reduces the likelihood of TCP congestion control mechanisms being triggered too aggressively. By dropping packets early, these algorithms signal to senders that they should reduce their transmission rate, avoiding network congestion before it worsens. The related RFC is RFC 3168, which discusses the use of Explicit Congestion Notification (ECN) as part of RED to inform endpoints about network congestion without packet loss. https://en.wikipedia.org/wiki/Random_early_detection https://tools.ietf.org/html/rfc3168

RFC 7567 also highlights the importance of reducing latency, particularly in real-time applications such as voice and video communication. Traditional queuing mechanisms, such as tail drop, often result in bufferbloat, where packets experience excessive delay due to large buffers being filled before any packets are dropped. This can lead to poor performance for latency-sensitive applications. AQM techniques that manage queue lengths dynamically can alleviate this issue by ensuring that queues are kept short, thus minimizing delays while maintaining good throughput. The related RFC is RFC 6817, which discusses the CoDel algorithm, an AQM technique designed to combat bufferbloat. https://en.wikipedia.org/wiki/Bufferbloat https://tools.ietf.org/html/rfc6817

RFC 7567 recommends the use of ECN as part of the AQM strategy to signal congestion to network endpoints without resorting to packet loss. By marking packets instead of dropping them, ECN allows endpoints to react to network congestion more gracefully, adjusting their transmission rates without the performance degradation associated with packet loss. This approach is particularly beneficial for interactive applications that are sensitive to packet loss but can tolerate minor adjustments in transmission rates. The related RFC is RFC 3168, which introduced the ECN mechanism into TCP/IP networks. https://en.wikipedia.org/wiki/Explicit_Congestion_Notification https://tools.ietf.org/html/rfc3168

Another key aspect of RFC 7567 is the recommendation that AQM techniques be enabled by default in network devices, particularly in routers and switches. Historically, AQM algorithms were not widely deployed because they were often seen as complex to configure and manage. However, with the advent of simpler and more effective algorithms like CoDel and PIE, it is now feasible to enable AQM by default in most network environments. This ensures that networks are better equipped to handle congestion without requiring extensive manual configuration. The related RFC is RFC 8033, which discusses the PIE algorithm and its advantages in managing network queues. https://en.wikipedia.org/wiki/Proportional_integral_controller https://tools.ietf.org/html/rfc8033

RFC 7567 also emphasizes the need for AQM techniques to be applied at network bottlenecks, where congestion is most likely to occur. These bottlenecks can be located at access links, peering points, or within data centers. By deploying AQM at these critical points, network operators can significantly improve the performance and reliability of their networks, particularly during periods of high demand. The RFC encourages network operators to identify these bottlenecks and ensure that appropriate AQM algorithms are in place. The related RFC is RFC 5129, which discusses congestion control principles in networking. https://en.wikipedia.org/wiki/Congestion_control https://tools.ietf.org/html/rfc5129

The title of this RFC is “IETF Recommendations Regarding Active Queue Management (RFC 7567).” RFC 7567 provides updated recommendations on the use of Active Queue Management techniques to reduce network congestion and improve overall performance. By promoting the use of algorithms like RED, CoDel, and PIE, and encouraging the use of ECN, the RFC ensures that network operators can effectively manage queues, reduce latency, and enhance the user experience. These recommendations represent an important step forward in addressing the challenges of modern network traffic management.