There are three types of auto discovery jobs that can be run on a network:

  • Ping Sweep

  • SNMP Sweep

  • Multi-hop Discover 

  • Ping-Sweep – Takes a range of IPs.  Pings each device twice to see if the device is active, then runs an SNMP-Sweep on each of the responding devices. 

  • SNMP-Sweep – Takes a range of IPs.  (Note that SNMP's are not needed to discover Cisco devices.)  This SNMP Sweep discovery type takes a range of IPs and attempts to discover each device IP.  It uses an SNMP query first.  If the SNMP string is incorrect, or the device just does not respond, it also attempts to login to the device, and try to discover it with account information. 

  • Multi-hop Discover – Takes the IP of a core seed router and a range.  Pulls route and arp table information from the seed router, takes each IP in this list, and repeats the process until it exhausts the list of unique IPs in the range.  A heavy load can be seen on routers in large networks, so this discovery type should not be used except as last resort in environments where users have no of what exists in their network.

Depending on the type of auto Discovery completed, details must be entered defining the Auto Discovery job. These details are located on four tabs.

  • Properties

  • Seed Addresses

  • Ranges

  • Credentials


Do not use SNMP v3 Credentials for Auto Discovery if you have PIXes.

The Seed Addresses tab is only available when running a Multi-hop discovery.

Once devices are located, they are placed into the Auto Discovered network as managed devices, and system jobs are run in the scheduler to pull the configurations and hardware specs for the new devices.

Although devices enter a network as managed, they can be reclassified. The three device classifications available in Network Configuration Manager are:


Indicates devices that are associated with networks. Managed devices reside in the central repository, and are under the control of authorized network users.


Indicates devices that have been discovered, but are not managed by a network and flagged, so that they are not rediscovered in subsequent Auto Discovery runs. All revision history about a device is lost when it is placed into an unmanaged state.


When a device is removed from all networks, it's state is changed to Unclassified. Unclassified devices cannot be managed until placed back into a network. The device retains all revision history while in an unclassified state.