Brocade Multi-Service IronWare Administration Guide (Supporting R05.6.00) - Manual (Page 124)

106 Multi-Service IronWare Administration Guide
53-1003028-02
Continuous system monitor overview
5
Continuous system monitor overview
Continuous system monitoring (Sysmon) is implemented to monitor the overall system’s health.
Sysmon is a system-wide, modular monitoring service. It monitors different system components of
a device to determine if those components are operating correctly.
Sysmon periodically monitors the system for defined event types such as errors on TM and FE
links. Sysmon runs as a background process. It has a default policy that controls what is monitored
and what actions will be taken if a fault is detected. Sysmon generates the following log outputs for
the monitoring information.
• Syslog
• Sysmon internal log
NOTE
Syslog reported Sysmon alarm messages should be reported to Brocade Technical Support.
Internal logs are generated to give more information to Brocade Technical Support when a problem
occurs. The existence of internal logs doesn’t mean the system is experiencing problems, or that
some actions need to be taken. If Sysmon detects a failure, it will report the failure by generating
the syslog messages. In some cases the failed device will be shutdown or isolated from the
system. In other cases the software may attempt to recover the failed device.
Overall system performance depends on how resources are utilized. Any shortage of resources
impacts the overall performance of a system. The system resource histogram feature provides
detailed information on how system resources are used. It collects information on task CPU usage,
buffer usage and memory usage and stores this information in internal memory.
Runtime diagnostics are a critical component of a networking system to provide maximum uptime
by detecting and isolating faults, and then recovering from them. A system runtime diagnostics
framework supports execution of diagnostic tests such as the port CRC error monitoring test. It
manages this background diagnostic test and provides mechanisms for taking corrective action.
Event monitoring
This section discusses the following topics:
• Event monitoring overview
• Event types
• Displaying event information
Event monitoring overview
The Sysmon monitors a number of event types periodically. Sysmon detects errors based on
polling and interrupt. Polling is reading specific hardware registers. Interrupt is an instantaneous
event detection by Sysmon. Sysmon continuously monitors management processor and interface
processors via polling and interrupt methods. Once a threshold is reached, Sysmon logs the event
in the internal Sysmon log and takes an action based on the event type. There are the following
action types:
• Syslog: Generates a message in the Syslog
• Shutdown link: Disables the link between the TM and the FE