User's Manual

76-4
SDRAM (Synchronous Dynamic Random Access Memory) and NVRAM (Non Vulnerable Random Access
Memory) is provided inside the switch as two part of the log buffer zone, The two buffer zone record the log
information in a circuit working pattern, namely when log information need to be recorded exceeds the buffer
size, the oldest log information will be erased and replaced by the new log information, information saved in
NVRAM will stay permanently while those in SDRAM will lost when the system restarts or encounter an power
failure. Information in the log buffer zone is critical for monitoring the system operation and detecting
abnormal states.
Note: the NVRAM log buffer may not exist on some switches, which only have the SDRAM log buffer
zone.
It is recommended to use the system log server. By configuring the log host on the switch, the log can
be sent to the log server for future examination.
76.7.1.2 Format and Severity of the Log Information
The log information format is compatible with the BSD syslog protocol, so we can record and analyze the log
by the systlog (system log protect session) on the UNIX/LINUX, as well as syslog similar applications on PC.
The log information is classified into eight classes by severity or emergency procedure. One level per value
and the higher the emergency level the log information has, the smaller its value will be. For example, the
level of critical is 2, and warning is 4, debugging is leveled at 7, so the critical is higher than warnings which no
doubt is high than debugging. The rule applied in filtering the log information by severity level is that: only the
log information with level equal to or higher than the threshold will be outputted. So when the severity
threshold is set to debugging, all information will be outputted and if set to critical, only critical, alerts and
emergencies will be outputted.
Follow table summarized the log information severity level and brief description. Note: these severity levels
are in accordance with the standard UNIX/LINUX syslog.
Table 76-1 Severity of the log76-4information
Severity
Value
Description
emergencies
0
System is unusable
alerts
1
Action must be taken immediately
critical
2
Critical conditions
errors
3
Error conditions
warnings
4
Warning conditions
notifications
5
Normal but significant condition
informational
6
Informational messages
debugging
7
Debug-level messages
Right now the switch can generate information of following four levels
 Restart the switch, mission abnormal, hot plug on the CHASSIS switch chips are classified critical
 Up/down interface, topology change, aggregate port state change of the interface are notifications
warnings