📄 thresholds.pod
字号:
=head1 NAMEThreshold=head1 DESCRIPTIONThe thresholds routine runs whenever you like, it process the collected statistics in the RRDs and compares the numbers to stored thresholds in this tableand if exceeded raises an event for that device. The thresholds use the device 'role' to weight the events.The threshold routine is typically run every 15mins from cron. The most recent 15min average of the snmp variable is calculated fromthe stored rrd values, and compared to the threshold policy table.Yu can set a policy of all zeros, to effectively disable thresholding for a device, or class of devices. This means that for a device, or a classof devices based on role and type, you can turn off proactive events. For example, useful for switches which may have a high port churn, asusers turn off hosts at night, so the procative interface availibility event can be disabled, but still keep proactive interface utilisation, etc=head2 threshold=over 4=item *reachablereachable is the % of the reachability of a device, for the past 15 mins.Will raise event 'Proactive Reachability Threshold'=item *responseresponse is a threshold for the ping response time of a device in ms, as average over the past 15mins.Will raise event 'Proactive Response Time Threshold'=item *cpu15 mins average CPU time.Will raise event 'Proactive CPU Threshold'=item *Memmem is the % buffer and % I/O memory used, against the total router memory. so a high % used is not a good thing.Will raise event 'Proactive Memory Threshold'=item *nonucastnonucast is the average number of broadcast packets seen in the last 15 mins - useful for trapping broadcast storms. Should not peak to more than 5% of the ucast packets.Will raise event 'Proactive Interface Input NonUnicast Threshold ifDescr'Will raise event 'Proactive Interface Output NonUnicast Threshold ifDescr'=item *utilutil is line utilisation expressed as a % of the interface bandwidth, either default or configured with a band statement on theinterface. On a serial line, anything over 60% ( 15min average) will cause a router queue and that is not a good thing.Will raise event 'Proactive Interface Input Utilisation Threshold ifDescr'Will raise event 'Proactive Interface Output Utilisation Threshold ifDescr'=item *int_availint_avail is a 15min summary of the availability (up) of a individual interface.Will raise event 'Proactive Availability Threshold Interface ifDescr'=item *availavail is the 15 min summary of the total device interface availability , as a % of up against total (i.e) % Availability = 100 * total device interface 'operStatus' / device 'operCount'Will raise event 'Proactive Interface Availability Threshold'=item *modemmodem is the number of dead modems in a chassis.Will raise event 'Proactive Dead Modem Threshold"=item *modem_utilmodem_util is the number of modems unavailable in a chassis.Will raise event 'Proactive Modem Utilisation Threshold'=back=head2 role=head2 type=head2 node=head2 interface=head2 critical=head2 fatal=head2 major=head2 minor=head2 warning
⌨️ 快捷键说明
复制代码
Ctrl + C
搜索代码
Ctrl + F
全屏模式
F11
切换主题
Ctrl + Shift + D
显示快捷键
?
增大字号
Ctrl + =
减小字号
Ctrl + -