About the Event Pipeline

Any incident information that appears in your incident views first travels through the event pipeline. The event pipeline guarantees that the incident data is analyzed in chronological order.

Note Not all information that travels through the pipeline results in an incident.

If an incident does not meet the criteria for an event pipeline stage, it is ignored and passed to the next stage in the pipeline. The following table describes the event pipeline stages.

NNMi Event Pipeline Stages
Event PipelineStages Description
SNMP Trap Receiver

Accepts all SNMP traps.

Tip See About the Trap Service Stagesfor information about Trap Service stages that occur before the Event Pipeline stages begin.

Incident Receiver

Accepts all incident information that comes from the NNMi Causal Engine. See The NNMi Causal Engine and Incidents.

Note The incident information that is received includes any Custom Correlation configurations.

Geo Incident Receiver Accepts all incident information that comes from Global or Regional Managers.
Type Enforcer

Determines if a configuration exists for this trap, event, or incident.

If the incident configuration exists, the type enforcer begins to populate the incident fields according to the configuration. Examples of the incident fields that are populated include Severity, Origin, Category, and Correlation Nature. If an incident configuration is disabled or does not exist for the incident, NNMi drops the incident.

Resolver

Drops the trap if the Source Object or Source Node is not in the topology, unless the "Discard Unresolved SNMP Traps and Syslog Messages" check box is unchecked.

Determines if the incident's Source Node or Source Object (such as interface or card) matches an object in the NNMi database.

If available, the Resolver populates the incident with the most current Source Node and Source Object attribute values.

Customization

Checks for any of the following incident configurations in the order listed:

  • Suppression
  • Enrichment
  • Dampening
Store Bulk Collects incidents and stores them. NNMi stores this information in bulk, using a pre-defined time period or number of incidents, whichever occurs first. The default time period is 3 seconds. The default number of incidents is 300.If you send a trap and subsequent traps do not occur on the network for a period of time after the trap is sent, NNMi waits up to 30 seconds before persisting new incident or trap information.
Notification Notifies other process and services about a new incident.
Pairwise Checks for any current pairwise configurations for the incident.
Rate Checks for any current rate configurations for the incident.
Dedup Checks for any current deduplication configurations for the incident.
Relate Performs any additional Causal Engine correlations, including Custom Correlations, and cancels the incident when applicable.
Actions Performs any automatic actions that the NNMi administrator has configured to be run for one or more incidents. See Using Actions to Perform Tasks for more information.
Rba Requires Network Node Manager iSPI Network Engineering Toolset Software (NNM iSPI NET) and requires installation of a Diagnostic Server.

The NNM iSPI NET Diagnostics stage checks whether Diagnostics should be run on the current incident and submits a execution request to run the Diagnostics report on the device. See RbaManager.