NMIS includes a powerful thresholding feature, for comparing the collected performance data against the configured thresholds.  

For details on configuring thresholds please refer to Basic and Advanced Thresholds in NMIS8

Standard List of Thresholds in NMIS

This is a summary of the thresholds and the associated vendors.

Threshold NameEventVendor
availableProactive Interface AvailabilityCommon for all Vendors
calls_utilProactive Calls UtilisationCisco
ccpuProactive CPUCisco
cpuProactive CPUCisco (the most common for Cisco devices)
cpuUtilProactive CPUAlcatel, Zyxel
cpu_cpmProactive CPUCisco
env_tempProactive TempCisco, Zyxel
hrsmpcpuProactive CPUMicrosoft
jnx_bufferProactive Buffer UtilisationJuniper
jnx_cpuProactive CPUJuniper
jnx_heapProactive Heap UtilisationJuniper
jnx_tempProactive TempJuniper
mem-procProactive Memory FreeCisco
memUtilProactive Memory UtilisationAlcatel, Zyxel
modem_deadProactive Dead ModemCisco
modem_unavProactive Modem UtilisationCisco
pkt_discards_inProactive Interface Discards Input PacketsCommon for all Vendors
pkt_discards_outProactive Interface Discards Output PacketsCommon for all Vendors
pkt_errors_inProactive Interface Error Input PacketsCommon for all Vendors
pkt_errors_outProactive Interface Error Output PacketsCommon for all Vendors
reachableProactive ReachabilityCommon for all Vendors
responseProactive Response TimeCommon for all Vendors
ssCpuRawIdleProactive CPU IO Idlenet-snmp (Linux, Solaris, etc)
ssCpuRawSystemProactive CPU IO Systemnet-snmp (Linux, Solaris, etc)
ssCpuRawUserProactive CPU IO Usernet-snmp (Linux, Solaris, etc)
ssCpuRawWaitProactive CPU IO Waitnet-snmp (Linux, Solaris, etc)
util_inProactive Interface Input UtilisationCommon for all Vendors
util_outProactive Interface Output UtilisationCommon for all Vendors

Tracking Which Threshold to Change

If you are receiving events for a threshold and you are wanting to tune/modify those thresholds, the best way to do this is to first get the name of the node and the event.  Using the table above will help you to determine if this event is a vendor specific event or if it is common for all vendors.

In the case of an event which is common for all vendors, you can follow the documentation Basic and Advanced Thresholds in NMIS8 to modify the levels accordingly.

If the thresholds are vendor specific you will need to identify which "threshold name" to modify, the hardest one to find is the threshold for CPU, to determine which one, access the model for the node in question, and look in the model for how CPU is managed, for Cisco Routers it is in the system -> nodehealth section, this will include the names of the thresholds, e.g. the Section=nodehealth includes the relevant threshold variables at the bottom.

In this example the threshold name to modify is just "cpu" and the process is described in Basic and Advanced Thresholds in NMIS8.

  • No labels