Chapter 3-13. Detecting Congestion in Fibre Channel Fabrics

Tx-datarate-burst

Tx-datarate-burst detects multiple occurrences of tx-datarate at a lower granularity of one second, essentially helping in detecting traffic bursts. The OBFL on Cisco MDS switches logs the rising and falling threshold events with time and date stamps. Tx-datarate-burst 以一秒的较低粒度检测 tx-datarate 的多次出现,基本上有助于检测流量突发。Cisco MDS 交换机上的 OBFL 会记录阈值上升和下降事件,并标注时间和日期。

Refer to Example 3-24, which is based on Port-Monitor polling interval of 10 seconds, a rising-threshold of 5, a falling threshold of 1, and a datarate percentage of 90%. There are two time periods where interface fc1/13 was running at a high Tx utilization. The first interval of high utilization was Mar 3 11:18:12 through 11:18:35. These times are 10 seconds earlier than the displayed timestamps because the event started 10 seconds earlier but detected only after the polling interval. This is a 23-second duration. The second was on Mar 3 from 11:26:07 to 11:27:28. This is an 81-second duration. Since this was the tx-datarate-burst counter, both rising thresholds had five 1-second bursts of over 90% in the 10-second polling interval. When the falling-threshold hits, there were zero 1-second intervals where the tx-datarate was 90% or more. 请参阅例 3-24,该例基于端口监控器 10 秒的轮询间隔、5 的上升阈值、1 的下降阈值和 90% 的数据速率百分比。有两个时间段,接口 fc1/13 Tx 利用率很高。第一个高利用率时段是 3 3 11:18:12 11:18:35。这些时间比显示的时间戳早 10 秒,因为事件开始时间早 10 秒,但在轮询间隔后才检测到。持续时间为 23 秒。第二个事件发生在 3 3 日,从 11:26:07 11:27:28。持续时间为 81 秒。由于这是

你可能感兴趣的:(Linux,kernel,网络,linux,服务器,kernel,tcp/ip)