4.1 Concepts of discovery
NNMi's default setting for discovering only routers and switches enables you to focus your network management on the critical or most important devices. In other words, target the backbone of the network first. In general, avoid managing end nodes (for example, personal computers or printers) unless the end node is identified as a critical resource. For example, database and application servers might be considered critical resources.
NNMi provides several ways to control the devices to be discovered and included in the NNMi topology. Your discovery configuration can be very simple, quite complex, or anywhere in between, depending on how your network is organized and what you want to manage with NNMi.
- Important note
-
NNMi does not perform any default discovery. You must configure discovery before any devices appear in the NNMi topology.
Each discovered node (physical or virtually hosted) counts toward the license limit, regardless of whether NNMi is actively managing that node. The capacity of your NNMi license might influence your approach to discovery.
For details about the configuration for discovering many nodes, see NNMi Help.
Status monitoring considerations might also influence your choices. By default, the State Poller only monitors interfaces connected to devices NNMi has discovered. You can override this default for some areas of your network, and you can discover devices beyond the limits of your responsibility. (For details about State Poller, see Chapter 5. NNMi State Polling.)
NNMi provides two primary discovery configuration models:
-
List-based discovery
![[Figure]](GRAPHICS/ZUENG033.GIF)
Uses a list of seeds to explicitly tell NNMi exactly which devices to add to the database for monitoring.
-
Rule-based discovery
Tells NNMi which areas of your network and device types to add to the database, gives NNMi a starting address in each area,
and then lets NNMi discover the defined devices.
You can use any combination of list-based and rule-based discovery to configure the devices for NNMi to discover. Initial discovery adds these devices to the NNMi topology, and then spiral discovery routinely rediscovers the network to ensure that the topology remains current.
NNMi uses tenants to support networks containing overlapping address domains. Overlapping address domains might exist in the static network address translation (NAT), dynamic network address translation (NAT), or port address translation (PAT) area in the network management domain. To handle such a network, NNMi uses seeded discovery to place overlapping address domains in different tenants. For details, see NNMi Help.
- Important note
-
If you plan to configure multi-tenancy, configure tenants before initiating network discovery.
- Organization of this section