Troubleshooting Routing

Routing problems are notoriously difficult to resolve, particularly when routers that are external to your organization ”such as on the Internet ”are used. As an administrator, though, you can verify that your configuration has been set up correctly and also demonstrate that the routers within your organization are working as expected. This section outlines some of the utilities that can be used, such as ping , netstat , and traceroute .

Verifying the Configuration

You should confirm that the system being used as a router is configured correctly. Start by running the ifconfig -a command to verify that the network interfaces have been configured correctly. If the interfaces are marked UP and RUNNING , then also check the IP addresses, the network masks, and the broadcast addresses. If the network mask is incorrectly set, the broadcast address will also be wrong. You can rectify this by checking the /etc/inet/ netmasks file.

You should also verify that the /etc/inet/ hosts and /etc/inet/networks files contain the correct IP addresses and names , as well as the contents of each network interface configuration file ”that is, those files starting with /etc/hostname.< xxx > , where < xxx > represents the specific interface.

ping

The ping command is extremely useful for testing the reachability of other hosts on the network. Use this command to check that you can reach another system on the local network ”this demonstrates connectivity to the network. You should also ping the default gateway. If this fails, then the default gateway could be down, but also check the /etc/defaultrouter file to make sure you have entered the correct IP address or hostname.

netstat

The netstat command is used with the -r option to view the routing table entries. It is often preferable, when diagnosing problems, to use the netstat -rn command, so that only IP addresses are displayed, rather than host and network names.

traceroute

Use the traceroute command to follow the path to the remote host you are trying to reach. The traceroute command sends three probes at 5-second intervals to each router that is encountered along the path. If any of these return asterisks , then it means there has been no response from the router, which can assist greatly with identifying where the problem is located. The following example traceroute output shown demonstrates a problem:

 ultra10# traceroute www.mobile-ventures.net Tracing route to mobile-ventures.net [209.67.50.203] over a maximum of 30 hops:   1     9 ms     8 ms     8 ms  10.25.0.1   2    13 ms     8 ms     8 ms  gsr01-st.blueyonder.co.uk [62.30.65.1]   3    12 ms    12 ms    12 ms  172.18.6.69   4    13 ms    14 ms    12 ms  tele1-azt-pos.telewest.net [194.117.136.2]   5    13 ms    13 ms    12 ms  pos50402hsd-gsr2-linx.cableinet.net [194.117.154.190]   6    12 ms    12 ms    11 ms  zcr1-so-5-0-0.Londonlnt.cw.net [166.63.222.37]   7    85 ms    83 ms    85 ms  *   *   *   8    80 ms    78 ms    80 ms  *   *   * ... 

This output demonstrates that the trace was successful through the routers provided in the U.K. by my ISP (entries 1 thru 6) and that the problem occurs when the next router following the international gateway in London is probed (entry 7). This information could be useful to the ISP if you raise a trouble ticket to report the problem, but it clearly shows that the problem is not on your own local network.