From OpenNMS

Node status monitoring

Nodes are monitored by basic ICMP ping. If a response is not received within a configurable timeout, with a configurable number of retries, then the node is considered "down", and appropriate events passed into the OpenNMS event system.

Path Outages

For any node, the IP address of an upstream device can be specified - if a node is found to be "down", this "path outage" node is checked as well. If the upstream node is also down, then the "Node Down" event is suppressed, and a "Path outage" event substituted instead. Chains of devices can be configured to represent the path between the OpenNMS server and the monitored nodes. Careful configuration of this feature ensures that, for example, if a link to a remote site goes down, your NOC operators will not be flooded with numerous unnecessary "Node Down" events.