Updating Thresholds

Thresholds can originate from many sources and can be designated per OID, per OID/DEVICE, or per OID/DEVICE/INSTANCE.

Examples;

<CiscoCpuUtil>
   <thresholds>
        high = 75   # per oid
   </thresholds>
   <r-waupacaco.wiscnet.net>
        <thresholds>
                high = 85 # per device
        </thresholds>
        <utilization>
                high = 85 # or explicitly list the instance
        </utilization>
   </r-waupacaco.wiscnet.net>
</CiscoCpuUtil>

• the manual threshold file. Can be OID, OID/DEVICE or OID/DEVICE/INSTANCE

  [@server data]$ cat /usr/local/fido/etc/fido_manual_thresholds.config

<jnxYellowAlarmCount>
   <r-peer-2.net.wisc.edu>
     <thresholds>
        # re1 active
        not_equal_to = 1
     </thresholds>
   </r-peer-2.net.wisc.edu>
</jnxYellowAlarmCount>
     

• manually from the snmp poll config file. Can be OID, OID/DEVICE or OID/DEVICE/INSTANCE. Typically per OID.

  <hrSystemProcesses>
     desired_state = "within threshold bounds"
     threshold_units = "processes"
     <manual_thresholds>
          high = 0
     </manual_thresholds>
     ...
     ...
     

• OR from an SNMP poll. Automatically per OID/DEVICE/INSTANCE

     
  <iemStatusProbeCurrentTemp>
     ...
     ...
     <threshold>
       high = return $snmp_data{$device}{'data'}{'iemConfigProbeHighTempThreshold'}{$instance}
       low  = return $snmp_data{$device}{'data'}{'iemConfigProbeLowTempThreshold'}{$instance}
     </threshold>
    

The './update_thresholds.pl' script does the following

• Collects and merges thresholds based on $update_config{'threshold_file_glob'}
• Loads in the snmp poll config file and sees which tests actually want to use thresholds
• Copies/modifies threshold values based on configuration requests
• Creates the full thresholds file