Below is an illustration of how events and alarms are generated in the SDM and passed from the SDM to the third-party NMS as Northbound traps.
There are two ways in which events are generated in the SDM:
- Through SDM operations such as device discovery, threshold events, status polling, or custom events.
- Through traps. A trap is an unsolicited notification sent to the SDM by the SNMP agent in the network element.
The SDM collects and displays traps. The SDM logs traps with a time and date signature so that technicians in the headend can analyze the logs for fault patterns. The SDM also forwards SNMP traps to other network management stations (NMS).
Although all traps are passed to the SDM, they can be filtered; however, doing so affects all users of the SDM and prevents the data for the filtered traps from reaching the events.
Events are stored in the SDM database and can be viewed in the Network Events window. See Events for detailed information.
When there is an event that indicates a failure condition, alarms are generated. Alarms can be viewed and acted upon in the SDM. Filters can also be set up to sort alarms. See View All Alarms for detailed information.
The correlation of events with alarms is done using the failure object of the event object: for all events with the same failure object, there will only be one alarm, and the severity of the alarm will be the severity of the latest event with the same failure object. This is illustrated below.
