How To Run a Successful Network Topology Discovery
Traverse is able to determine the layer-2/layer-3 network topology (which device is connected to which other devices) automatically when performing a network discovery.
The discovery engine in Traverse uses SNMP to collect various information from routers and switches in the network, including ARP/MAC address tables, routing tables, interface numbering, etc to build the topology. The more information Traverse is able to collect, the more accurate and complete the dependency information will be. Therefore it is essential that the routers, switches, firewalls, wireless access points as well as other devices are configured to support SNMP queries.
There are a number of pre-requisites that will ensure an effective network discovery session:
Before starting the network discovery, ensure that SNMP enabled devices are configured to allow queries from the DGE. Some devices limit SNMP queries from a fixed number of IP addresses (sometimes categorized as a "management station"). Other devices may implement an access-list/firewall rule that prevent SNMP queries on UDP port 161. If necessary, consult with network/security administrators to authorize such access.
On the same note, verify that the devices in the specified subnet are reachable by ICMP ping. In the first stage of the discovery session, Traverse will perform a (staggered) "ping sweep" to determine which IP addresses are active. If there is a firewall/router between the DGE and target network, ensure that ICMP ping (echo) requests will be allowed from the DGE.
Provide all the SNMP community strings used within your network in the inital step of a discovery session. Each community string should be specified in a line by itself. The discovery engine will automatically determine which community string is application to each device.
Taking these extra steps will ensure that Traverse is able to determine the physical and logical topology of various nodes.