Examples of Name Resolution
When a network-connected machine begins a conversation with another machine, the first thing it needs to do to contact the target is to map the IP address of the target machine to a Media Access Control (MAC) address. Broadcasting Address Resolution Protocol (ARP) packets so that the two machines can trade MAC addresses and then communicate at the MAC layer of the OSI model accomplishes this. This also enables the two machines to resolve the IP address of their target to a MAC address. Listing 3.1 shows the output of an arp -a command showing the mapping of MAC addresses to IP addresses.
Listing 3.1 Listing the ARP Cache
c:\> arp -a Interface: 192.168.0.200 on Interface 0x1000003 Internet Address Physical Address Type 192.168.0.1 00-40-05-1e-30-61 dynamic 192.168.0.3 00-20-c5-e2-6a-fc dynamic 192.168.0.5 00-10-4b-cc-ee-92 dynamic
Another type of name resolution is mapping a NetBIOS name to an IP address. This occurs frequently on Windows computers because many Windows computers use NetBIOS over TCP/IP (NetBT) to perform file and print sharing. This occurs through broadcast, Windows Internet Name Service (WINS), or a static lmhosts file. Listing 3.2 shows the output of an nbtstat -c command showing the mapping of IP addresses to NetBIOS names.
Listing 3.2 Displaying the NetBIOS Name Cache
c:\> nbtstat -c Local Area Connection: Node IpAddress: [192.168.0.200] Scope Id: [ ] NetBIOS Remote Cache Name Table Name Type Host Address Life [sec] ------------------------------------------------------------ WOOD <20> UNIQUE 192.168.0.1 592 GOLD <20> UNIQUE 192.168.0.5 587
The main reason for name resolution, however, that we focus on in this book is the process of finding a Domain Controller. If DCs are incapable of finding each other on the network to communicate, or if clients can't find DCs to log on to, Active Directory becomes useless. Many interrelated services are required for a functioning Active Directory, and all of them can be found on a DC. Kerberos is used for authentication; LDAP is used to find, insert, update, and delete objects; and Global Catalog servers are used to find objects in this and other AD trees in the forest. All these services are automatically installed on an Active Directory DC when the Windows 2000 server is promoted to a DC. These services must be found by a client many times during a session but particularly during the logon process. Additionally, DNS is used not only to find a DC, but also to find a physically close DC that is a member of the client's AD domain—rather than just any DC at random.For more information on Kerberos, see "Kerberos." For more information on installing Domain Controllers, see "Promoting a Server to Domain Controller."
DNS is also one of the first troubleshooting areas that should be addressed when an AD begins to function improperly. DNS is absolutely critical to the location of services and objects within an AD. AD replication and logon use DNS to find other services on the network they need.For more information on finding services, see "Finding a Domain Controller."
Unlike Windows NT, which can use any number of tools and services to locate a DC, Windows 2000 Domain Controllers find each other through DNS and DNS alone. Although other name resolution methodologies can be used by client computers, the default and preferred method is DNS.