Comment
Author: Admin | 2025-04-27
A 0x11 value to indicate congestion in the data path. Figure 5. Host X informs hosts A and B about network congestion by sending them CNP packets After the destination (host X) receives the marked packet, it learns that congestion is happening in the data path and generates a CNP packet to the source host from whom it received the now-marked packet. As only some amounts of the data packets were marked with congestion experienced bits, the source reduces traffic throughput for that flow and continues to send packets. If congestion continues and the buffer usage rises over the WRED maximum threshold, the switch marks every packet with congestion experienced bits. This means that the sender receives many CNP packets, and based on its algorithm it should drastically reduce the data rate transmission toward the destination. This mitigates congestion and the buffer should start draining. As that happens, the traffic rate should rise until the next time congestion is signaled. How PFC Works Figure 6. Priority Flow Control no-drop queue, xOFF (Red) and xON (Green) thresholds, and buffer headroom When PFC is enabled on the Cisco Nexus 9000 switch, a class of service is dedicated to lossless transport. Traffic in this class is treated differently than traffic in other classes. Any port on the Cisco Nexus 9000 switch configured with PFC is allocated a dedicated no-drop queue and dedicated buffer for that queue. To provide lossless capabilities, the queue has two thresholds. The xOFF threshold is set higher in the
Add Comment