Brocade Communications Systems RFS6000 System Reference Manual page 19

Serviceability
A special set of Service CLI commands are available to provide additional troubleshooting
capabilities for service personnel (access to Linux services, panic logs, etc.). Only authorized users
or service personnel are provided access to the Service CLI.
A built-in Packet Sniffer enables service personnel and users to capture incoming and outgoing
packets in a buffer.
The controller also collects statistics for RF activity, Ethernet port activity etc. RF statistics include
roaming stats, packet counters, octets tx/rx, signal, noise SNR, retry, and information for each
Client.
Tracing / logging
Log messages are well-defined and documented system messages with various destinations. They
are numbered and referenced by ID. Each severity level group, can be configured separately to go
to either the serial console, telnet interface, log file or remote syslog server.
Trace messages are more free-form and are used mainly by support personnel for tracking
problems. They are enabled or disabled via CLI commands. Trace messages can go to a log file, the
serial console, or the current tty.
Log and trace messages are interleaved in the same log file, so chronological order is preserved.
Log and trace messages from different processes are similarly interleaved in the same file for the
same reason.
Log message format is similar to the format used by syslog messages (RFC 3164). Log messages
include message severity, source (facility), the time the message was generated and a textual
message describing the situation triggering the event. For more information on using the controller
logging functionality, see
Process monitor
The controller Process Monitor checks to ensure processes under its control are up and running.
Each monitored process sends periodic heartbeat messages. A process that is down (due to a
software crash or stuck in an endless loop) is detected when its heartbeat is not received. Such a
process is terminated (if still running) and restarted (if configured) by the Process Monitor.
Hardware abstraction layer and drivers
The Hardware Abstraction Layer (
hardware/platform specific data. Drivers include platform specific components such as Ethernet,
Flash Memory storage and thermal sensors.
Redundancy
Using the controller redundancy, up to 12 controllers can be configured in a redundancy group
(and provide group monitoring). In the event of a controller failure, an existing cluster member
assumes control. Therefore, the controller supported network is always up and running even if a
controller fails or is removed for maintenance or a software upgrade.
The following redundancy features are supported:
Brocade Mobility RFS6000 and RFS7000 System Reference Guide
53-1001858-01
"Configuring system logging"
HAL)
provides an abstraction library with an interface hiding
Software overview
on page 502.
1
5