Brocade Fabric OS Documentation Updates - Supporting Fabric OS v6.4.x (53-1002063-09, August 2012)

36 Fabric OS Documentation Updates

53-1002063-09

Documentation updates for Fabric OS v6.4.3

bottleneckMon

This command has been enhanced in two ways:

1. Support for more granular alerts: New options have been provided to enable RASlog alerts and

SNMP traps for only one kind of bottleneck, latency or congestion, while preserving the existing

ability to enable alerts for both kinds of bottleneck.

2. Credit recovery enhancements: The credit recovery configuration command has been enhanced to

support on-demand checking of certain types of backend ports for credit loss.

1. Support for more granular alerts

SYNTAX The syntax for the enable and configure (change) options has been changed as follows to support the

decoupling of congestion and latency alerts:

bottleneckmon --enable [-alert | -alert=congestion | -alert=latency | -noalert ] [ -cthresh

congestion_threshold] [-lthresh latency_threshold] [ -time seconds] [-qtime seconds]

bottleneckmon --config [-alert | -alert=congestion | -alert=latency | -noalert] [-cthresh

congestion_threshold] [-lthresh latency_threshold] [-time seconds][-qtime seconds]

[[slot]port_list]

DESCRIPTION Modify the description section (second paragraph on page 70) as follows:

When bottleneck monitoring is enabled on a switch and you specify alerts, the command triggers an

SNMP and a RASlog alert when the ports on the configured switch experience a bottleneck. Another

alert is sent after the condition resolves. You can configure alerts for latency, for congestion, or both.

Modify the description of the alert configuration in the operand section as follows:

-alert Enables alerts when configured thresholds are exceeded on the ports that are

enabled for bottleneck monitoring. This option enables both congestion and

latency alerts. The alerting mechanism is by RASlog and SNMP traps. This

operand is optional; if omitted, no alert is assumed. When -alert is specified, you

can optionally specify thresholds for congestion, or latency, or both. If -alert is not

specified and you try to specify additional parameters, the command fails with an

appropriate message.

-alert=congestion

Enables congestion alerts only. You can optionally specify a congestion threshold

for this type of alert. Latency thresholds are not valid.

-alert=latency

Enables latency alerts only. You can optionally specify a latency threshold for this

type of alert. Congestion thresholds are not valid.

-cthresh congestion_threshold

Specifies the severity threshold for congestion that triggers an alert. The

threshold indicates the percentage of one-second intervals affected by the

bottleneck condition within the specified time window. The threshold is

expressed as the equivalent fraction between 0 and 1. The default value is

0.8. This parameter is valid only for congestion alerts.