Installation Guide

240 Brocade MLX Series and Brocade NetIron XMR Installation Guide
53-1004195-03
Using alarms to collect and monitor device status
• System: Interface Ethernet 3/1, state down -fabric connectivity down
• System: Interface Ethernet 3/1, state up –fabric connectivity up
Using alarms to collect and monitor device status
Beginning in Brocade NetIron R05.3.00, the software keeps two logs; one of hardware status currently available to the system, and
another of hardware status history. The current alarm log keeps only entries for current information; when a hardware status is no longer
valid, the entry is cleared. The alarm history log keeps a record of hardware statuses even after the status has changed. The alarm history
log enables you to quickly determine trouble areas in a system. For example, by accessing the history, you can quickly determine if a
problem is occurring too frequently and might require action.
Each hardware status entry is called an alarm and is classified by severity assigned by the software. The software categorizes alarms in the
following levels:
• Critical - A condition that will cause damage to the system. A condition that causes a traffic outage on multiple ports.
• Major - A condition that causes traffic outage on single ports or might cause damage to the system.
• Minor - A condition that should be investigated but will not damage the system.
By default, all hardware status alarm levels of major severity and higher are logged, though you can configure the status alarm levels sent
to both alarm logs (current and history). You can use the show alarm command to view the current status on a device, or a logged history
of hardware alarms. To change the levels of alarms sent to the alarm logs, refer to “Configuring Alarm History Buffer Size” on page 240.
The alarms are specific to hardware status, whereas the syslog records information for software events. Alarms can also be configured on
very specific terms such as a failed temperature sensor on a single interface module.
To take advantage of this feature, you should first set the alarm history buffer size. This is optional, but you have the option to limit how
many entries are stored in the alarm history so you can free up space for other resources. Refer to “Configuring Alarm History Buffer Size”
on page 240 for more information.
Next, you should configure the severity of alarms for each device you want logged. For detailed information, refer to “Configuring alarm
logging” on page 241. Once you have configured your alarm logging, you can display alarms in the current alarm log and alarm history
log using the show alarm command, as described in Table 43 on page 242.
Configuring Alarm History Buffer Size
The history buffer size is configurable. The default buffer size is 400 entries, but it can be configured to list between 100 and 3000
entries using the alarm history command.
For example, to configure the alarm history log size to 100 entries, enter the following command:
Brocade# alarm history 100
Syntax: [no] alarm history n
where n is the number of log entries you want to store in the alarm history log, between 100 and 3000.
To reset the alarm history back to the default buffer size, use the alarm history command with the no operand. For example, to set the
buffer size back to the default entry size from 100, enter the following command:
Brocade# no alarm history 100