Network connections

Network connections are logical interfaces between software (such as protocols) and hardware (such as modems or network adapters). Network connections can be seen in the Network Connections window, Connections are prioritized and are normally configured with various types of protocols, services, and client software.

Network protocols are network languages used for computer-to computer communication. For example, Windows networks, UNIX networks, and the Internet all rely on the TCP/IP network protocol for basic communication. In Windows, connections can communicate with foreign hosts only by using network protocols that are installed on the local computer and bound to that connection. TCP/IP(version 4) is installed and bound by default to every connection. However, the NWLink protocol must be manually installed, configured, and bound to connections requiring compatibility with Novell NetWare networks that do not use TCP/IP. (NWLink is the Microsoft implementation of the Internetwork Packet Exchange/Sequenced Packet Exchange [IPX/SPX] protocol, which is native to NetWare.) In addition, the

AppleTalk protocol must be manually installed and bound to connections requiring compatibility with Apple networks that do not use TCP/IP.

TCP/IP is actually a group of protocols referred to as a stack or suite. This protocol stack includes Address Resolution Protocol (ARP), Internet Protocol (IP), Transmission Control Protocol (TCP), User Datagram Protocol (UDP), Domain Name System (DNS), Hypertext Transfer Protocol (HTTP), and many others.

Network services are programs that provide features, such as quality of service, to hosts or protocols on a network. Some network services that can be readily installed and bound to network connections. Additional services can be installed from the Windows Server 2003 Setup disk or from a third-party source.

Network clients are programs that allow a computer to connect to a network operating system. For example, by installing Client Service For NetWare and binding the service to a particular connection, you can connect to NetWare networks.

Addressing is the practice of maintaining a coherent system of addresses within your network so that all computers can communicate. In a network, each host normally distinguishes itself by using a specific network address. For example, IP version 4 provides a method for computers with 4-byte addresses to communicate with each other. Addresses must be configured so that the first part of the address (the network ID) matches that of the other computers on the local network, or subnet. To allow such computers to communicate with hosts on different subnets, you must connect the subnets by using routers (such as the Routing And Remote Access service of Windows Server 2003). Addresses can be configured manually, distributed automatically through the use of a DHCP server, or self-configured.

Name Resolution Most networks use a naming system so that people can refer to computers by name

instead of by address. Name resolution is the process of translating a computer name into an address, and vice versa.

Because Windows can use two different naming systems, NetBIOS and DNS, Windows networks support two name resolution systems. NetBIOS is native to older Microsoft networks, and today it is used primarily for compatibility with legacy features and systems.

DNS is the native naming system of the Internet and all Windows operating systems released since Microsoft Windows 2000.

To resolve NetBIOS names, Microsoft networks can send broadcast queries to all systems on the same network segment or send requests to a WINS server. To resolve DNS (host) names, Microsoft networks rely on the DNS protocol and DNS servers. To function properly, both of these name resolution services must be configured and supported by a knowledgeable network administrator.

Network Computer Groups In Windows, computers can be grouped into workgroups or domains.

A workgroup is a simple grouping of resources intended to help users find such resources as printers and shared folders. By default, computers in Windows workgroups use the NetBIOS naming system to name computers and resolve those names. NetBIOS is used with associated protocols, such as Common Internet File System (CIFS).an extension of the Server Message Block (SMB) protocol. To provide file sharing, security for network shares, and network browsing features. However, no centralized security or management features are available.

A domain is a collection of computers, defined by a network administrator, that share a common directory, security policies, and relationships with other domains. Security and directory information are stored in domain controllers within each domain.

Active Directory In Windows Server 2003 networks, domains are created in and supported by Microsoft

Active Directory directory service. Active Directory is a distributed database and directory service that is replicated among all domain controllers on the network. The Active Directory database stores information about network objects including domains, computers, users, and other objects. The distributed nature of Active Directory gives network users access to permitted resources anywhere on the network by using a single logon process. It also provides a single point of administration for all network objects. The term domains is used to refer both to groupings of computers in Active Directory and to hierarchical name suffixes such as microsoft.com in DNS. Remember that Active Directory domains and DNS domains are separate entities governed by separate systems. However, to simplify administration, Active Directory domains and their member computers are normally assigned names that match DNS names. In this way, the ActiveDirectory namespace and the DNS namespace overlap.

Remote access connections must be configured for users who connect to a Windows network from a nonlocal site. The two basic methods for remote access include direct dial-up to a network computer and virtual private networks. For dial-up access, you must not only configure a server to answer incoming calls, but you must also configure authentication, access permissions, and encryption requirements. VPNs enable private connections to cross a public network such as the Internet. These network connections require a different set of configuration procedures for authentication, encryption, and security.

Network Address Translation (NAT) is a method of allowing computers internal to your network that have been given nonpublic addresses to communicate with computers on the Internet. When you configure NAT to be used with your network infrastructure, this setup affects the addressing scheme of your network. Internet Connection Sharing (ICS) is a simple implementation of NAT included with recent Windows operating systems. 

Certificate Infrastructure are used for public key cryptography, which is an important security element in Windows Server 2003 networks. Certificates and public key cryptography are used in many Windows features, such as the Secure Sockets Layer (SSL), the Internet Protocol Security (IPSec) protocol (which encrypts IP communications), smart cards, and the Encrypting File System (EFS, which secures files on a network). The certificate infrastructure supported in Windows Server 2003 networks integrates with the Public Key Infra structure (PKI) system: a system of digital certificates, certification authorities, and other registration authorities that authenticate each party involved in an electronic transaction.

Default TCP/IP Settings, You can view TCP/IP settings configured for any connection by opening the Internet Protocol (TCP/IP) Properties dialog box. To open this dialog box, open a connection’s properties dialog box, select Internet Protocol (TCP/IP) from the list of network components, and click Properties. Figure 1-10 shows the Internet Protocol (TCP/IP) Properties dialog box associated with the default LAN connection. The IP address of a computer is assigned automatically in a default Windows installation. For a new computer on a network, or for a network on which no DHCP server has been configured, the computer assigns itself an IP address in the range of 169.254.0.1.169.254.255.254. This address is available through a feature called Automatic Private IP Addressing (APIPA).

Automatic Private IP Addressing APIPA is an addressing feature for simple networks that consist of a single network segment. Whenever a computer running Windows Server 2003 has been configured to obtain an IP address automatically, and when no DHCP server or alternate configuration is available, the computer uses APIPA to assign itself a private IP address in the range of 169.254.0.1.169.254.255.254.

To determine whether Automatic Private IP Addressing is currently enabled and active, enter ipconfig /all at a command prompt. The resulting text identifies your IP address and other information. If the Autoconfiguration Enabled line reads Yes and the IP address is in the 169.254.0.1.169.254.255.254 range, Automatic Private IP Addressing is active. This automatic addressing feature works only for computers on a network segment that cannot obtain an IP address through other means. If a DHCP server later becomes

available to a host that has assigned itself an APIPA address, the computer changes its IP address to one obtained from the DHCP server. Computers using APIPA addresses can communicate only with other computers using APIPA addresses on the same network segment; they are not directly reachable from the Internet. Note also that through APIPA, you cannot configure a computer with a DNS server address, a default gateway address, or a WINS server address. If you want a computer to obtain an address automatically and also want to specify a default gateway, DNS server, or WINS server when no DHCP server is available, you can do so by using an alternate configuration. APIPA is available on any computer running Windows 98, Microsoft Windows Millennium Edition (Me), Windows 2000, Windows XP, or Windows Server 2003. 

Disabling APIPA If you want to ensure that APIPA will not be used, you can either configure an alternate address in the connection’s IP properties, or disable the automatic addressing feature by editing the Registry. Note that to disable APIPA for one adapter and to disable APIPA for all adapters requires you to edit different Registry keys.

To disable APIPA on a single adapter by editing the Registry, complete the following steps:

1. Use the Registry Editor Regedit.exe to add the Registry entry IPAutoconfigurationEnabled with a value of 0 (REG_DWORD data type) in the following subkey: HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\Tcpip\Parameters\ Interfaces\interface

2. Restart the computer.

To disable APIPA for multiple adapters by editing the Registry, complete the following steps:

• Set the value of the IPAutoconfigurationEnabled entry to 0 (REG_DWORD data type) in the following Registry subkey: HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\Tcpip\Parameters
• Restart the computer.