Users Guide

Fabric OS Administrator’s Guide 433
53-1002920-02
Setting bottleneck detection alerts
15
Port Alerts? LatencyThresh CongestionThresh Time (s) QTime(s)
================================================================================
1 Y 0.100 0.800 300 300
2 C -- 0.800 600 600
3 L 0.100 -- 300 300
4N-- -- -- --
NOTE
If there are no per-port overrides, “Per-port overrides for alert parameters” section is not displayed.
Setting bottleneck detection alerts
You can configure Fabric OS to log per-port alerts based on the latency and congestion history of
the port. Alerts are generated based on the number of affected seconds over a specified period of
time. If the number of affected seconds is higher than the threshold, an alert is generated.
This evaluation is done independently for latency and congestion.
NOTE
A congestion bottleneck detection alert is generated whenever a frame timeout occurs irrespective
of the number of affected seconds in the observation window.
The following bottleneckmon -alert parameters determine whether an alert is generated and the
reason for the alert.
The -time parameter specifies the time window. For this example, -time equals 12 seconds.
The -cthresh and -lthresh parameters specify the thresholds on number of affected seconds that
trigger alerts for congestion and latency bottlenecks, respectively.
For example, Figure 56 shows an interval of 12 seconds, in which 6 seconds are affected by a
congestion bottleneck and 3 seconds are affected by a latency bottleneck. This example uses the
default values for these parameters, where -cthresh = 0.8 (80%) and -lthresh = 0.1 (10%).
The following command results in the example illustrated in Figure 56:
bottleneckmon -alert -time 12 -cthresh 0.8 -lthresh 0.1
FIGURE 56 Affected seconds for bottleneck detection