Print

Resolving Host Names

A host is any computer that makes a service or data available to other computers, over a network or the Internet. For systems running TCP/IP, any computer can provide data or services to other computers running the protocol, so for all practical purposes, ‘host’ is just another name for a networked computer.

All computers on a TCP/IP network identify themselves with an IP address, which is four groups of eight binary digits, separated by periods. We humans are notoriously poor at remembering long strings of binary digits, so for our convenience each of these 8-digit strings, called octets, are translated into decimal numbers of up to three digits. This gives us the IP address as we normally see it, in the form of a number such as 192.168.0.2.

This format is called dotted decimal, and it’s a step in the right direction but still a bit much for those of us who have to look twice to dial a 7-digit phone number. The other problem with IP addresses is that they usually change when whenever a computer is moved, and even more often in a network that uses dynamic addressing. For this reason systems are usually given a name that makes sense to the users, such as Sales02, Server1, or Yoda. This name is the Host name, and it will normally stay the same when the system is moved around the network.

There’s nothing wrong with having a system named Yoda, but to do anything on the network that name must be mapped to the IP address that is recognized by the other computers, and this process is called hostname resolution. Without hostname resolution, networks would stay small and the Internet would not exist at all. In a Windows network, there are three main ways to configure hostnames.

Local Hostname

For a small peer-to-per network (up to 5 systems), each computer can be given a hostname and the systems will sort it out among themselves. To change the hostname, right-click on My Computer, choose Properties and select the Network Identification tab. Depending on the version of Windows and the configuration, administrator rights may be required to change the hostname.

HOSTS file

Windows Internet Naming Service (WINS)

A text file that must be edited is obviously impractical in a network that uses dynamic addressing, and can be tedious even in a larger network with static addressing. WINS has become very popular because it builds the name resolution database automatically. The WINS database resides on a server, and when a WINS client system starts up it sends a name resolution request to the server. The server will check to see if the hostname is in use by anyone else on the network, and if not the new host and its IP address are added to the table. The server will also assign an IP address if dynamic addressing is being used. If the hostname is already in the database, the WINS server will query that system to see if it’s still in use. If it doesn’t answer up after three tries, the newcomer gets it.

Domain Name Service (DNS)

For larger, more complex networks, and specifically for the Internet, an even more powerful method of name resolution is provided by DNS. This uses a hierarchal naming system that can handle millions of computers. Internet addresses are the example most people are familiar with. If you send an e-mail to sales@micro2000.com from anywhere in the world, DNS servers will resolve it to the correct IP address so it can be routed to us.

The top-level domain in this example is .com, so the message gets routed first to one of the .com servers which looks through its millions of .com listings to find micro2000. There, it gets resolved to the IP address for the system hosting the micro2000 domain. Once it’s routed to our domain server, it gets further resolved and sorted to the right department. The message will pass through additional DNS servers before it gets here, and these will build their own resolution tables of the traffic that comes their way. In the future, messages to micro2000.com will not have to go all the way to the top-level domain to get resolved.

Domain Name Service can also be used in a local network, with each of the DNS servers having its own Zone of Authority. The network is divided into domains, and each domain will have one primary domain controller with authority over that domain, which means it will have a table of all the systems in that domain, and their IP addresses. The domain can also have secondary controllers with a copy of the primary controller’s name resolution table. If this sounds complicated, it is, and DNS should be installed and configured by a qualified network administrator. For the workstations though, all that is needed is to add the hostname and IP address of the DNS server (domain controller) with a zone of authority that includes the workstation. After that, the workstation can access other systems on the network rather painlessly.

In addition to the primary and secondary DNS servers, a network may have what’s called a caching server. This one does not have a zone of authority. It simply records the traffic coming through and saves the results in its own table. This data is saved for a specified period called the Time to Live, or TTL. As long as the data is in the table of the caching server, it can resolve an address without calling on the next higher DNS server.

This may be more than most of us need to know about computer names and addresses, but next time you send an e-mail, visit a web site or copy a file from another system on your local network, at least you’ll have a better idea of what’s going on under the hood.

Resolving Host Names by  Micro2000.

Newsletter-Feed-22

This site looks much better in a browser that supports current web standards, but it is accessible to any browser. Download one now