Chapter 1, Introducing Microsoft Windows 2000 Networking

|1|

Chapter 1, Lesson 1

The OSI Reference Model

|1|

|1| 1. Introducing the OSI Reference Model

A. The languages that networked computers use to communicate are called protocols.

1. The computers must use common protocols.

B. The communication between computers on a network occurs at many levels.

1. The developers of networking protocols often split the essential communication functions into multiple layers, which are implemented by separate protocols.

2. The combination of protocols needed to provide the various levels of communication is collectively known as a protocol stack.

3. For communication to occur, the corresponding layers of the stack on the two computers must run the same protocols at each layer of the stack.

C. The most common generalized representation of the protocol stack is a seven-layer construction called the Open Systems Interconnection (OSI) reference model.

1. The OSI model is defined in a document called “The Basic Reference Model for Open Systems Interconnection.”

a. This document is the result of two separate projects conducted by the International Organization for Standardization (ISO) and the Comité Consultatif International Téléphonique et Télégraphique (CCITT).

(1) The CCITT is now known as the Telecommunications Standardization Sector of the International Telecommunication Union (ITU-T).

b. The projects were combined in 1983, and the resulting document was published as ISO 7498 by the ISO and as X.200 by the CCITT.

|2| D. The OSI reference model splits the networking protocol stack into seven layers.

|3| 1. The top of the model represents an application running on the computer, and the bottom of the model represents the network medium that connects the computers.

2. Each layer contains specific functions that contribute to the communications process that enables an application on one computer to send data to an application on another computer.

3. The protocols used by the networked computers perform these functions, and they complement one another to create the protocol stack.

E. The protocols operating at adjacent layers of the stack communicate by providing services to, and receiving services from, one another.

1. Data originating at an application at the top of the stack is passed down through the layers until it reaches the network medium.

2. After the data is transmitted over the medium to the destination computer, it is passed up through the layers of the protocol stack to the corresponding application at the top.

3. The corresponding layers of the protocol stack on each computer must run the same (or compatible) protocols for communications to take place.

|1|

Note Several of the labs in this course involve the capture and analysis of network traffic using the Network Monitor application included with Windows 2000. To effectively analyze the contents of the captured network packets, you (students) must understand the data encapsulation process.

|4| F. Understanding data encapsulation

1. Data encapsulation is the process by which the protocols operating at the various layers of the OSI model package the information they receive from the layer above.

2. A protocol receiving data from the layer above it encapsulates the data by adding its own information, in the form of a new header (and in some cases, a footer as well).

a. The header consists of fields that contain information specific to the encapsulating protocol that the corresponding protocol on the destination computer will read.

b. The information received from the layer above follows the header and becomes the payload in the unit of data created by the protocol.

3. When the data unit is passed down to the layer below it, another protocol encapsulates it again.

|5| 4. The header and payload from the protocol above combine to become the payload of the data unit created by the new protocol when it attaches its own header, as shown in Slide 5.

5. The end result of the data encapsulation process is a data packet consisting of the original information generated by the application at the top of the stack, with several protocol headers (and, in some cases, one footer) attached to it.

a. The data packet is the protocol data unit that is transmitted over the network to another computer.

(1) When the packet arrives at its destination, the protocol at each layer of the stack reads the header information added by the corresponding protocol on the source computer and processes the payload data accordingly.

|6| (2) This process enables the protocols at each layer of the stack on one computer to communicate with their counterparts on another computer, as shown in Slide 6.

|1|

7| 2. The Physical Layer

A. Provides the interface between the computer and the network medium that carries data from one computer to another

1. In most cases the network medium is a cable, such as a copper local area network (LAN) or telephone cable.

2. Networks can also use various types of fiber-optic cable or wireless media, such as radio waves and infrared transmissions.

B. In a computer, the physical layer takes the form of a network interface adapter (also known as a network interface card, or NIC) or a modem.

C. The physical layer defines how to encode data into signals suitable for transmission over the network medium.

1. Depending on the medium involved, the physical layer protocol describes how to convert the binary data generated by the computer into electrical voltages, pulses of light, radio transmissions, or whatever form is required.

D. The physical layer protocol defines the properties of the network medium itself, such as the types of cables used to build the network and how they should be installed.

1. In the case of LAN protocols, these elements are essential because the physical layer is coupled with the data-link layer protocol.

2. Data-link layer protocols such as Ethernet and Token Ring perform functions that rely on precise timing of the signal transmissions.

a. They cannot operate successfully on a physical network with cables that are too long or that do not conform to the specifications the protocol requires.

3. Although Ethernet and Token Ring are primarily associated with the data-link layer, they also include physical layer parameters that define the types of networks they can use.

E. The installation of the physical layer is increasingly left to specialized contractors.

1. Network administrators who must know the intricacies of Transmission Control Protocol/Internet Protocol (TCP/IP) and other upper-layer protocols to do their jobs do not necessarily have to know everything about the physical layer configuration.

2. Most LANs use unshielded twisted pair (UTP) cable, just like that of telephone systems, and cabling contractors frequently install data network cabling when they install the telephone network.

F. A network administrator must have a working knowledge of the physical layer standards used to build the network so that he or she can troubleshoot simple problems.

|8| 3. The Data-Link Layer

A. Responsible for the final packaging of the application data before it is transmitted over the network medium

B. A data-link layer protocol receives data units from network layer protocols and encapsulates them for the final time.

1. On a LAN, the data-link layer protocol applies both a header and a footer to form a frame.

a. The frame is the envelope used to carry the data to other computers on the network.

C. Like a paper envelope transported by the postal service, the data-link layer protocol header contains the addresses of the computer sending the data and the computer receiving it.

1. These addresses are found in two of the fields that make up the data-link layer protocol header.

2. The manufacturer hard-codes the 6-byte addresses used by the data-link layer protocols on a LAN into the network interface adapters.

a. These addresses are known as hardware addresses or media access control (MAC) addresses.

b. Each manufacturer is assigned a 3-byte value called an organizationally unique identifier (OUI) by the Institute of Electrical and Electronics Engineers (IEEE).

c. The addresses for the adapters produced by a particular manufacturer consist of its OUI plus a unique 3-byte identifier assigned by the manufacturer.

(1) This ensures that every adapter on a network has a unique hardware address.

|9| D. Data-link layer protocols can also perform a number of other functions, including the following:

1. Media access control (MAC)

a. The process by which a computer gains access to a shared network medium

b. On most LANs, all the computers are connected to a common baseband medium.

(1) A baseband medium is one that can carry only one signal at a time, as opposed to a broadband medium, which can carry many signals at once.

c. On a baseband network, only one computer can transmit data at a time.

(1) If two computers were to transmit simultaneously, their signals would collide, causing the data to be lost.

d. The data-link layer protocols used on LANs have a MAC mechanism that defines a method for preventing, minimizing, or recovering from collisions.

e. Token Ring networks use a MAC mechanism called token passing, in which a small packet called a token circulates around the network.

(1) Only the computer that has the token can transmit its data.

(2) Because there is only one token, only one computer at a time can transmit data, which prevents collisions.

f. Ethernet networks use a different MAC mechanism called Carrier Sense Multiple Access with Collision Detection (CSMA/CD).

(1) Computers using CSMA/CD listen to the network, and if it is free, transmit their data.

(a) CSMA/CD is not as sure a method for preventing collisions as token passing; collisions are expected to occur on Ethernet networks.

(2) CSMA/CD reduces the number of collisions that would otherwise occur and also enables the computers to detect the collisions and compensate for them by retransmitting their data.

g. CSMA/CD is one of the primary reasons why data-link layer LAN protocols are so closely associated with physical layer standards.

(1) If an Ethernet network uses the wrong cable or the cable is installed incorrectly, the computers cannot effectively detect collisions and data is lost.

(2) Data loss reduces the network’s efficiency by forcing the protocols at the higher layers to detect the missing packets.

2. Protocol identification

a. The process by which the data-link layer protocol identifies the protocol that generated the payload carried in the packet

b. Computers often run multiple protocols at the network layer, all of which share a single data-link layer protocol.

c. As the network layer protocols pass their data down to the data-link layer, the data-link layer protocol creates a field in the header containing a code that specifies which network layer protocol generated the data in the payload.

(1) This code allows the data-link layer protocol on the computer receiving the packet to identify the network layer protocol to which it should pass the incoming data.

3. Error detection

a. Error detection can be performed at several different layers of the OSI model.

b. The data-link layer protocols used on LANs are unique among the layers in that they include a footer as well as a header.

c. The footer consists of a field called a Frame Check Sequence (FCS), which contains a cyclical redundancy check (CRC) value calculated by the sending computer on the contents of the entire frame.

d. On receipt of the packet, the destination computer performs the same calculation and compares its results to those in the FCS field.

(1) If the results match, the packet has been transmitted without error.

(2) If the results do not match, the receiving system discards the packet.

e. There is no error correction at the data-link layer.

(1) The protocol can detect damaged packets and discard them, but it does not retransmit them.

(a) Retransmission is left to the protocols operating at the upper layers of the OSI model.

E. The data-link layer protocols used on LANs are the most complex of the protocols at this layer.

F. Other data-link layer protocols are much simpler than the LAN protocols (Ethernet and Token Ring). Examples of these less complex protocols are

1. The Serial Line Internet Protocol (SLIP), which consists only of a single byte that follows each packet transmitted over a connection

a. SLIP is used for point-to-point connections between two computers only, so there is no need for addresses or a MAC mechanism.

b. SLIP has fallen into disuse because it is too simple and lacks features provided by other protocols.

2. The Point-to-Point Protocol (PPP)

|10| 4. The Network Layer

A. The network layer is primarily responsible for the end-to-end communications between computers located on different networks.

1. By contrast, data-link layer protocols are used to transmit data between two computers that are directly connected by a LAN or a dedicated link.

a. The destination address in a data-link layer protocol header always identifies a computer on the local network.

2. The network layer protocol also encapsulates data using a header that contains Source and Destination Address fields, but these addresses identify the ultimate source and destination of the packet.

3. The network layer protocol is also responsible for routing packets, fragmenting them (when necessary), and identifying the protocol that generated the data in the packet.

|11| B. Network layer addresses

1. The most common network layer protocol, the Internet Protocol (IP), contains its own independent system of addresses.

2. Novell NetWare’s Internetwork Packet Exchange (IPX) protocol uses the hardware addresses coded into the computers’ network interface adapters.

3. The NetBIOS Extended User Interface (NetBEUI) protocol provided with Windows uses NetBIOS (Network Basic Input/Output System) names as network layer identifiers.

C. The network layer protocol is also responsible for routing the packet to its destination.

|1|

1. A packet traveling through an internetwork to a specific destination is passed from router to router.

2. Each router is responsible for sending the packet on its way over the most efficient path to the destination.

3. Internetworks typically have built-in redundancy that provides multiple paths to a given destination.

a. A router knows these paths and transmits packets over the path that can get them to the destination in the shortest time or that can use the fewest intermediate routers.

4. When a router processes an incoming packet, it strips off the data-link layer protocol header and footer and uses the information in the network layer protocol header to determine the best route to the destination.

a. The router then reencapsulates the packet for transmission to its next destination.

5. Because routers function as the interfaces between networks, the packets may be reencapsulated using the same or a different data-link layer protocol.

D. Fragmentation

1. Fragmentation is necessary when a router connects two networks that support different-size packets.

a. When a packet arrives at a router over a Token Ring network, for example, the packet can be up to 4500 bytes long.

b. If the router determines that it must transmit the packet over an Ethernet network, it has a problem, because Ethernet only supports packets up to 1500 bytes long.

c. To resolve this problem, the router splits the packet into several fragments, each no larger than the maximum transmission unit (MTU) size for the outgoing network.

d. Each fragment is identified by a code that indicates its place in the packet.

e. The router then transmits each fragment in a separate packet.

f. When the destination computer receives all the fragments, it reassembles them into the original packet.

2. Depending on the configuration of the internetwork, a single packet might be fragmented more than once on the way to the destination.

a. The packet is split into fragments, and then those fragments are split into smaller fragments by another router.

b. The fragments are not reassembled until they reach the packet’s final destination.

E. Protocol identification at the network layer works the same way as it does at the data-link layer, and for the same reason.

1. Computers can often use multiple transport layer protocols with a single network layer protocol.

2. The network layer protocol header contains a code specifying which transport layer protocol generated the data carried inside the packet.

|12| 5. The Transport Layer

A. Provides functions that complement those of the network layer protocol and help to get the data to the destination in a timely and efficient manner

B. Uses connection-oriented and connectionless protocols

|13| C. Connection-oriented protocols

1. A connection-oriented protocol is one in which the source and destination computers exchange a series of messages before they transmit any application data.

a. These messages establish a connection, which confirms that both computers exist, are functioning properly, and are ready to receive and transmit data.

b. Connection-oriented communications are usually associated with guaranteed delivery, in which the computer receiving data returns acknowledgments to the sender on a regular basis, confirming that it has received the data without errors.

c. Connection-oriented protocols, such as the Transmission Control Protocol (TCP), are typically used to transmit large amounts of data that require extreme accuracy.

(1) For example, when transmitting a program file over a network, every bit must be transmitted correctly, or the file will not execute properly.

(2) By contrast, a video stream can survive the loss of some bits in transit; the only effect is a momentary lapse in the quality of the display.

|14| 2. When transmitting large amounts of data, a connection-oriented transport layer protocol must split the data stream into pieces called segments, which can fit into single packets.

a. Segmentation is a lot like the fragmentation process that occurs at the network layer, but be careful not to confuse the two.

(1) Fragmentation is performed by the intermediate routers that the packets pass through on their way to their destination.

(2) Segmentation occurs only at the source of a data transmission.

(3) Segments can also be fragmented by routers during their journey.

b. The sending computer splits the stream into segments of an appropriate size and packages each segment in a separate packet.

c. The transport layer protocol header contains a code that identifies each segment so that the destination computer can reassemble them into the original data stream.

|1|

3. Connection-oriented protocols generate a great deal of network traffic overhead, both in the form of

a. Additional messages used to establish connections and acknowledge transmissions

b. Additional header fields

|15| 4. Most connection-oriented protocols operating at the transport layer also include a function called flow control.

a. Flow control is the ability of the receiving computer to transmit information that instructs the sender to modify its transmission rate.

b. The network interface adapter in every networked computer has a memory buffer used to hold incoming packets that are waiting to be processed.

c. When the memory buffer in the network adapater of a computer receiving a data transmission approaches fullness, the computer can send flow control information to the sender, ordering it to slow down its transmission rate.

d. In TCP, flow control is implemented as a field in the protocol header that specifies the number of packets the computer can receive.

(1) The receiving computer modifies the value of this field in the acknowledgment messages it returns to the sender.

(2) When the field’s value decreases, the sender knows that the receiver’s buffer is filling up faster than it can process the incoming packets, and it slows its transmission rate accordingly.

(3) As the field’s value increases, the sender can ramp up the transmission rate.

|16| D. Connectionless protocols

1. Connectionless protocols, such as UDP, are much simpler than connection-oriented protocols and have much lower overhead.

a. IP, at the network layer, is also a connectionless protocol.

2. Do not transmit connection establishment messages before they transmit application data and do not use packet acknowledgments

3. Have no segmentation or flow control

4. Usually are not suitable for transmitting large amounts of data requiring extreme accuracy

a. For the most part, computers use connectionless protocols for quick exchanges that consist of a single request and reply, such as Domain Name System (DNS) and Dynamic Host Configuration Protocol (DHCP) transactions.

(1) The reply message functions as a tacit acknowledgment, and the request is easily retransmitted if no reply is received in a timely manner.

b. Streaming audio and video applications can use connectionless protocols because they do not require bit accuracy.

5. Because they do not have the many complex features of connection-oriented protocols, connectionless protocols have much smaller headers.

a. The UDP header is only 8 bytes, as opposed to 20 bytes for TCP.

E. All transport layer protocols have a header field that provides protocol identification.

1. Protocol identification at the transport layer is used to identify the application layer protocol that generated the data carried in the packet (because there are no separate session or presentation layer protocols).

a. This process completes the path up through the protocol stack that the data takes at the destination computer.

F. Both connection-oriented and connectionless protocols can also perform error detection and correction, using a system of transmitted CRC values like those used in the data-link layer.

1. The difference is that the transport layer performs end-to-end error detection, but the data-link layer only checks for errors on local network transmissions.

2. Transport layer protocols can correct errors, as well as detect them, by transmitting to the sender a list of packets that need to be retransmitted.

3. Transport layer protocols can perform two types of error correction.

a. Correction of signaled errors occurs when a protocol at another layer (such as the data-link layer) informs the transport layer protocol that a specific packet needs to be retransmitted, and the transport layer protocol corrects the error without detecting it on its own.

b. Unsignaled errors are those that the transport layer detects on its own and corrects, without the aid of another protocol.

|1|

|17| 6. The Session Layer

A. The lower four layers of the OSI reference model (the physical, data-link, network, and transport layers) are concerned with getting data from one computer to another over the network.

1. The protocols operating at these layers are designed to overcome the obstacles that might prevent various types of data from being transmitted properly.

B. The session, presentation, and application layers are not concerned with data transmission; they assume that the lower-layer protocols can get data from one computer to another, intact and on time.

1. The session, presentation, and application layers have less obvious boundaries between them.

2. There are no session and presentation layer protocols; the functions attributed to these layers are incorporated into the application layer protocols.

C. The session layer performs many functions that aid in the exchange of messages between two computers, which is called a dialog.

1. Dialog separation inserts a bookmark-like device called a checkpoint into a dialog stream, which enables the communicating computers to perform an action at the same point in the dialog.

2. Dialog control regulates the communications between the two computers through one of two transmission modes:

a. Two-Way Alternate (TWA) mode. Only one computer can transmit at a time.

b. Two-Way Simultaneous (TWS) mode. Either computer can transmit at will.

(1) TWS mode presents problems that the session layer must address, such as whether a reply message was generated before or after the most recently transmitted request.

|18| 7. The Presentation Layer

A. Is relatively simple, when compared to the session layer

B. Provides a translation service that enables different types of computers to communicate

1. Applications on each computer platform generate network access requests using their own native syntax, which might be different from the syntax used by the application on the destination computer.

a. The syntax can incorporate several elements, include bit-coding formats such as American Standard Code for Information Interchange (ASCII) and Extended Binary Coded Decimal Interchange Code (EBCDIC), compression standards, and encryption algorithms.

2. Before the data generated by the application reaches the transport layer, the computer converts it from its native syntax (called an abstract syntax) to a transfer syntax, suitable for transmission over the network.

3. The computer receiving the data then translates the incoming information, this time converting the transfer syntax to the application’s own abstract syntax.

|19| 8. The Application Layer

A. Application layer protocols form the top of the computer’s protocol stack.

1. An application running on a computer uses an application layer protocol to request access to a resource located elsewhere on the network.

a. For example, a Web browser uses the Hypertext Transfer Protocol (HTTP) to generate requests for home pages hosted by Internet servers.

2. The application layer protocol is also the final destination for the data passed up through the stack on the receiving computer.

B. There are many application layer protocols, and each is designed to provide highly specialized services required by a particular application or type of application.

1. In some cases, the protocol is virtually indistinguishable from the application itself, as in the case of the File Transfer Protocol (FTP).

2. In other cases, the protocol is a separate entity that many different applications can use.

|20| C. Some of the most commonly used application layer protocols are as follows:

1. HTTP. Used by Web browsers and servers to exchange home page requests and replies

2. FTP. Used to transfer files between computers and perform basic file management tasks

3. DNS. Used to send host name resolution requests and replies to and from DNS servers

4. Dynamic Host Configuration Protocol (DHCP). Used to automatically configure the TCP/IP client on network computers

5. Simple Mail Transfer Protocol (SMTP). Used by e-mail clients to send messages and to exchange e-mail messages between servers

6. Simple Network Management Protocol (SNMP). Used to gather performance information about specific network components and transmit it to a central network management console

9. Using the OSI Model in the Real World

|1|

Note The OSI model exists primarily as a learning tool and a guide for referencing network communications processes, not as a model for creating protocols.

A. Originally, the OSI reference model was intended to be a guideline for the creation of a new set of networking protocols that would conform precisely to its seven layers.

1. Those protocols never appeared in a commercial form.

B. The protocols that networks use today are not analogous to the model, for several reasons.

1. The majority of the protocols currently in use, such as TCP/IP and Ethernet, were developed before the OSI model existed and were constructed using their own architectural models.

a. These protocols, therefore, tend to have functions that overlap into multiple OSI model layers.

2. Some of the layers, particularly the session and presentation layers, really do not warrant a separate protocol.

a. Creating a separate protocol would make the network communications process unnecessarily complex.

|21| C. The typical protocol stack used by an application running on a networked computer consists of four protocols.

1. The application layer protocol interacts with the application that needs to communicate with another computer on the network, and typically includes the functions attributed to the presentation and session layers.

2. The boundaries of the network and transport layers are the most accurate found in the model.

a. Virtually all computers use two separate protocols that conform quite rigidly to the functions of the network and transport layers.

3. The functions of the data-link and physical layers are often combined in a single protocol that encompasses all their functions.

D. The OSI model’s value is in its usefulness as a teaching and reference tool.

1. Although there are no separate session or presentation layer protocols, for example, the functions attributed to those layers are essential to network communications.

Chapter 1, Lesson 2

Installing and Binding Windows 2000 Networking Components

|1|

|22| 1. Introducing the Windows 2000 Networking Components

A. The networking functions defined by the OSI reference model are largely realized in Windows 2000 by four components: network interface adapters, protocols, clients, and services.

|23| 1. The protocol stack on a Windows 2000 computer consists, from the bottom up, of a network interface adapter, one or more protocols, one or more clients, and optionally, a collection of services.

|24| B. Network interface adapters

1. A network interface adapter in a computer running Windows 2000 typically consists of a NIC and the device driver that the computer needs to communicate with it.

a. These components perform the physical and data-link layer functions of the OSI reference model.

2. A network interface adapter does not have to be a NIC.

a. When you use a modem or other wide area network (WAN) communications device to connect to a network at a remote location, the WAN device itself functions as a network interface adapter and is functionally interchangeable with a NIC.

3. A computer with a single network interface adapter can handle the data traffic of multiple protocol modules operating above it.

a. The packets generated by the various protocols are combined and transmitted over the single network medium, a process called multiplexing.

4. A computer can also have multiple network interface adapters connecting it to different networks.

a. The most common configuration is a computer with a NIC connecting it to a LAN and a WAN connection to the Internet or another remote network.

b. However, one computer can also have multiple NICs installed, enabling it to function as a router that passes data between two networks.

|25| 5. When the computer has two or more network interface adapters, you can do one of the following:

a. Configure both adapters to handle the traffic generated by different protocols (enabling you to use one NIC to connect to a TCP/IP network and the other to connect to a NetWare network running IPX, for example)

b. Have both adapters handle all the installed protocols

|26| C. Protocols

1. Computers use many protocols when communicating on a network, but often these protocols are grouped together into cooperative groups, commonly called protocol suites.

a. For example, TCP/IP is a protocol suite that includes not only TCP and IP but also many other protocols operating at various layers of the OSI model.

2. Because using TCP/IP requires most or all of the protocols in the suite, Windows 2000 (and most other operating systems) install them as a single entity.

a. When you install the Internet Protocol (TCP/IP) software module on a computer running Windows 2000, you are actually installing the entire TCP/IP protocol suite and a variety of related tools and utilities.

b. “Installing a protocol” in Windows 2000 means installing a single protocol module, which technically implements multiple protocols and applications.

3. TCP/IP is the default protocol used by Windows 2000.

a. TCP/IP’s functionality falls just above the network interface adapter and spans the network and transport layers, even reaching as far up as the application layer in some of its components.

b. The primary TCP/IP protocols are IP at the network layer and TCP and UDP at the transport layer, but the suite also includes many application and network layer protocols.

|27| 4. Windows 2000 also includes two other protocols that are roughly analogous to TCP/IP in their functions.

a. NWLink is the Microsoft version of the IPX protocols developed by Novell for use with its NetWare operating system.

(1) Like TCP/IP, IPX is a protocol suite that consists of the IPX protocol itself, which operates at the network layer, and multiple transport layer protocols, including Sequenced Packet Exchange (SPX) and Network Core Protocol (NCP).

(2) NWLink is primarily intended for connecting computers running Windows to NetWare networks, but it is also possible to use it for communication between computers running Windows.

b. NetBEUI, the third of the general use protocol modules included with Windows 2000, was the original default networking protocol for the Windows operating systems.

(1) Ideal for small LANs, NetBEUI requires no manual configuration and is completely self-adjusting.

(2) Unlike TCP/IP and IPX, however, NetBEUI is not routable, which means that it is not suitable for use on internetworks.

5. The network adapter drivers used by Windows 2000 conform to the Network Device Interface Specification (NDIS).

a. The boundary between the adapter driver at the data-link layer and the protocol modules at the network layer on a computer running Windows 2000 is called the NDIS boundary layer.

(1) This boundary enables the computer to use different protocol modules interchangeably with the same network interface adapter

6. You can install one, two, or all three of the protocol modules supplied with Windows 2000 and choose which ones to use for specific purposes.

|28| D. Clients

1. The transport driver interface (TDI) at the top of the Windows 2000 protocol modules performs a function similar to that of the NDIS boundary layer.

2. Above the TDI are the Windows 2000 client modules, which can use any of the protocols installed on the computer.

3. The Client for Microsoft Networks module provides basic Windows network file and print services to the computer, enabling applications to access files and printers on network computers just as if they were installed in the local machine.

4. For NetWare connectivity, Windows 2000 also includes Client Service for NetWare (in Microsoft Windows 2000 Professional) or Gateway Service for NetWare (in Microsoft Windows 2000 Server).

a. These client modules are based on a component called a redirector, which evaluates resource access requests and determines whether the requested resource is located on the local machine or on the network.

(1) If the resource is on the network, the redirector passes the request to the appropriate protocol, starting it on its way down the stack to the network medium.

|29| E. Services

1. In Windows terminology, a service is a program that runs continuously on a computer, waiting to satisfy requests for particular functions.

a. For example, on a computer running Windows 2000 and DNS Server, the DNS Server program runs as a service that loads when the computer starts and is ready to service requests from DNS clients at all times.

2. Windows 2000, particularly in its server versions, includes a large collection of services that provides networking functions.

3. By default, Windows 2000 installations include services that provide basic networking functionality, such as the following:

a. Server. Enables the computer to share its files and printers with other systems on the network

b. Workstation. Makes it possible for applications running on the computer to access resources on other network systems

c. Messenger. Enables administrators and applications to send and receive messages

d. Browser. Compiles and maintains a list of the resources on the network

e. Netlogon. Enables the computer to locate the domain controller on the network and log on to a domain

4. Windows 2000 also includes many optional services that you can install with the OS or at any time afterward, such as the following:

a. Dynamic Host Configuration Protocol (DHCP). The combination of a service and a protocol that enables a computer running Windows 2000 Server to automatically assign IP addresses and other configuration parameters to the TCP/IP clients on a network

b. Domain Name System (DNS). A distributed Internet service that enables computers on a network to resolve host names into the IP addresses needed for TCP/IP communications

c. Windows Internet Name Service (WINS). A Network Basic Input/Output System (NetBIOS) name server, this is a LAN-based service that enables computers to resolve NetBIOS names into the IP addresses needed for TCP/IP communications.

d. Microsoft Certificate Services. Enables you to create and manage the certificate authorities (CAs) that issue digital certificates

(1) Digital certificates are electronic credentials that certify the online identities of individuals, organizations, and computers.

e. Routing and Remote Access Service (RRAS). A service that enables a computer running Windows 2000 Server to function in a variety of communications roles, including LAN router, remote access server, virtual private network (VPN) server, and network address translation (NAT) server

f. Internet Information Services (IIS). A group of services that enables a computer running Windows 2000 to function as a Web, FTP, or news server

|30| 2. Installing Windows 2000 Networking Components

A. To participate on a network, a computer running Windows 2000 must have, at the very least, a network interface adapter, a protocol, and a client installed.

B. Installing a network interface adapter

1. Most of the computers and NICs manufactured today conform to the Plug and Play standard, which automates the installation of the network interface adapter and its device driver.

2. To manually install an adapter, you use Add/Remove Hardware on Control Panel.

3. For every network interface adapter installed in a Windows 2000 computer, a Local Area Network icon appears in the Network And Dial-Up Connections window, which provides access to the configuration information for the interface.

4. You can also create additional connections in this window that use modems and other WAN devices as the network interface adapters by clicking the Make New Connection icon to launch the Network Connection Wizard.

C. When you install a network interface adapter in a computer running Windows 2000, or when the Windows 2000 setup program detects one during the operating system installation, the system installs a basic default protocol stack configuration consisting of the following components:

1. A device driver for the network interface adapter

2. Client for Microsoft Networks

3. The Internet Protocol (TCP/IP) module

4. The File and Printer Sharing for Microsoft Networks service, which enables other computers on the network to access the system’s files and printers.

D. To configure the networking components, or to install additional components, do the following:

1. Open the Properties dialog box for a particular connection by right-clicking its icon in the Network And Dial-Up Connections window.

2. Select Properties from the pop-up menu.

|31| a. The Local Area Connection Properties dialog box identifies the network interface adapter in the Connect Using box, and this dialog box also contains a list of the networking components installed on the computer.

3. Use the Install and Uninstall buttons to add or remove clients, protocols, and services.

4. Select a networking component from the list and click Properties to open a dialog box in which you can configure the properties of that component.

a. The controls in the dialog box vary, depending on the component you have selected.

|32| 3. Binding Windows 2000 Networking Components

A. By default, when you install a networking component, such as a client or protocol module, it is automatically bound to all the other components above and below it.

1. For example, if you install the NWLink protocol module on a computer with two network interface adapters, both adapters will be configured to use NWLink.

B. You can control the bindings of the various components from the Properties dialog box for each connection.

1. The check boxes next to the components in the Properties dialog box indicate which components are bound to the adapter used by that connection.

2. Clear the check box next to a component to unbind it from that connection.

C. Example: a computer running Windows 2000 with both a network interface adapter and a modem installed

1. You create a new connection so that you can use the modem to connect to the Internet, and you install the NetBEUI protocol for local network communications.

a. By default, the Internet Protocol (TCP/IP) and NetBEUI protocol modules are both bound to the NIC and the modem.

b. However, the modem cannot use NetBEUI when accessing the Internet, and the NIC cannot use TCP/IP on the local network because all the other computers are running NetBEUI only.

2. To conserve system resources and enhance performance, you can unbind TCP/IP from the NIC adapter and unbind NetBEUI from the modem connection.