NET 121b: Essentials of Networking

Chapter 13: Troubleshooting Network Connectivity

Objectives:

This chapter discusses several utilities that may be used in diagnosing network problems. The topics of this chapter are:

  1. Network troubleshooting utilities
  2. Telnet
  3. File Transfer Protocol (FTP)
  4. Troubleshooting
Concepts:

The text discusses several utilities found on most Windows and UNIX workstations..

  • ARP - ARP stands for Address Resolution Protocol. In standard Ethernet networks, machines may communicate inside the network with their MAC addresses. Communications across networks are more likely to use IP addresses. An ARP cache is a table that lists the IP addresses and MAC addresses of devices on a network. This table is consulted to change from one kind of addressing to the other. For example, I have just issued the command
    ARP -a
    to my workstation. It has responded with the contents of its ARP cache: its own IP and MAC addresses, and those of my default router.
  • hostname - This command will respond with the name of your device in your domain.
  • IPCONFIG - shows useful information on Windows NT and later machines, like the IP address, default router, and subnet mask. More information is shown if the command is entered as
    • IPCONFIG /all
    • IPCONFIG /release will release the currently held IP address to the DHCP server that gave it
    • IPCONFIG /renew will obtain a new lease from the DHCP server for an IP address
  • WINIPCFG - Like a light version of IPCONFIG, found on Windows 95, 98, and Me computers.
  • IFCONFIG - a UNIX command that can be used to view or configure the network interface settings for a workstation
  • nbtstat - The name of this utility is NetBIOS over TCP/IP Statistics. Not very enlightening. You need to know that your computer will typically hold the names and IP addresses of several devices in memory. Sometimes those devices go offline, and others come online. This may make it desirable to check what is in memory:
    nbtstat -a
    or tell the computer to reload this information from standard sources:
    nbtstat -R
    (Note that the case of the letter R in the command above is required to be capital.)
    nbtstat IP_address This version lets you check the tables in memory of the device specified by the IP address
  • NETSTAT - Can be used to view the status of current connections using TCP, UDP, ICMP, and IP. The status messages are a bit cryptic, so you will want to keep a reference for them handy when using this command.
  • NSLOOKUP - This can be used to report the IP address of a DNS name. It does not send a ping to the named device. The example in the text explains that this command checks what is stored in your DNS server about the name in question. A response to the command may take this format:
    Server: server name
    Address: IP address
    Name: DNS name
    Address: IP address
    The first pair of responses are about the DNS server on your network. The second pair are about the DNS name you are looking up. When I tried this with nslookup microsoft.com, I received two IP addresses in the line about Microsoft's server. Not unexpected, since a busy network will have more than one server responding to requests.
  • DIG - used on UNIX and Windows platforms, but must be installed on a Windows platform. The link provided goes to an IBM site that tells us DIG stands for Domain Information Groper. It digs information out of a DNS server. Think of it as a tool for troubleshooting DNS servers and services.
  • PING - can be issued on a command line, and has an extensive list of options. Usually, the options are unnecessary. You can ping the address 127.0.0.1, which stands for the IP stack on the machine you are using. You may also want to ping the IP address you think you have, with your network cable unplugged. Pinging the local loopback proves you have a working IP stack. Pinging your actual address, when unplugged, proves you have that address.
    Be aware that you can ping an IP address or a DNS name.
  • TRACERT (Trace Route) - This command will show how long each link in a route takes, as well as showing links that fail to pass packets to the next link. Successful transfers of data will report the total time to the destination. You can limit the trace to a specific number of hops with the command
    tracert -h hop_limit
    where hop_limit is a number.
  • ROUTE - all devices on an IP network have routing information tables in their memory. ROUTE allows you to view that information, and to modify that information. The reason you would want to modify it is you need to do so when routing tables are static and they need to be changed.
Telnet

Telnet is utility that allow your computer to connect like a dumb terminal to a host running a Telnet daemon. As we discussed in class, a daemon is a program, typically on a server, that runs in the background until it is needed or called by another device. It is typically used to connect to a UNIX server or a mainframe. Windows and Novell servers do not support Telnet connections to themselves.

File Transfer Protocol (FTP)

FTP is used to copy files to and from devices on an IP network. Numerous versions of FTP utilities exist for Windows, UNIX, and other platforms. The text states that it is important to correctly specify whether a file being transferred is text or binary. In practice, treating all files as binary files usually works.

Troubleshooting

The text presents a series of story problems about troubleshooting with these utilities. We will discuss them in class.