Table of Contents

NetWare Protocols

Background
Technology Basics
Media Access
Network Layer
Transport Layer
Upper-Layer Protocols

NetWare Protocols

Background

NetWare is a network operating system (NOS) and related support services environment created by Novell, Inc. and introduced to the market in the early 1980s. Then, networks were small and predominantly homogeneous; local-area network (LAN) workgroup communication was new; and the idea of a personal computer (PC) was just becoming popular.

Much of NetWare's networking technology was derived from Xerox Network Systems (XNS), a networking system created by Xerox Corporation in the late 1970s. For more information on XNS, see Chapter 22, "Xerox Network Systems."

By the early 1990s, NetWare's NOS market share had risen to between 50 and 75 percent. With over 500,000 NetWare networks installed worldwide and an accelerating movement to connect networks to other networks, NetWare and its supporting protocols often coexist on the same physical channel with many other popular protocols, including TCP/IP, DECnet, and AppleTalk.

Technology Basics

As a NOS environment, NetWare specifies the upper five layers of the OSI reference model. It provides file and printer sharing, support for various applications such as electronic mail transfer and database access, and other services. Like other NOSs such as the Network File System (NFS) from Sun Microsystems, Inc. and LAN Manager from Microsoft Corporation, NetWare is based on a client-server architecture. In such architectures, clients (sometimes called workstations) request certain services such as file and printer access from servers.

Originally, NetWare clients were small PCs, while servers were slightly more powerful PCs. As NetWare became more popular, it was ported to other computing platforms. Currently, NetWare clients and servers can be represented by virtually any kind of computer system, from PCs to mainframes.

A primary characteristic of the client-server system is that remote access is transparent to the user. This is accomplished through remote procedure calls, a process by which a local computer program running on a client sends a procedure call to a remote server. The server executes the remote procedure call and returns the requested information to the local computer client.

Figure 19-1 illustrates a simplified view of NetWare's best-known protocols and their relationship to the OSI reference model. With appropriate drivers, NetWare can run on any media-access protocol. The figure lists those media-access protocols currently supported with NetWare drivers.

Figure 19-1 : NetWare and the OSI Reference Model

Media Access

NetWare runs on Ethernet/IEEE 802.3, Token Ring/IEEE 802.5, Fiber Distributed Data Interface (FDDI), and ARCnet. For information on Ethernet/IEEE 802.3, see Chapter 5, "Ethernet/IEEE 802.3." For information on Token Ring/IEEE 802.5, see Chapter 6, "Token Ring/IEEE 802.5." For information on FDDI, see Chapter 7, "Fiber Distributed Data Interface." NetWare also works over synchronous wide-area network (WAN) links using the Point-to-Point Protocol (PPP). PPP is discussed in more detail in Chapter 9, "Point-to-Point Protocol."

Attached Resource Computer Network (ARCnet) is a simple network system that supports all three primary cable types (twisted-pair, coaxial, and fiber-optic) and two topologies (bus and star). It was developed by Datapoint Corporation and introduced in 1977. Although ARCnet has not attained the popularity enjoyed by Ethernet and Token Ring, its low cost and flexibility have resulted in many loyal supporters.

Network Layer

Internet Packet Exchange (IPX) is Novell's original network-layer protocol. When a device to be communicated with is located on a different network, IPX routes the information to the destination through any intermediate networks. Figure 19-2 shows the IPX packet format.

Figure 19-2 : IPX Packet Format

The fields of the IPX packet are as follows:

• Checksum---A 16-bit field that is set to 1s.
  

• Packet length---A 16-bit field that specifies the length, in bytes, of the complete IPX datagram. IPX packets can be any length up to the media maximum transmission unit (MTU) size. There is no packet fragmentation.
  

• Transport control---An 8-bit field that indicates the number of routers the packet has passed through. When the value of this field reaches 15, the packet is discarded under the assumption that a routing loop might be occurring.
  

• Packet type---An 8-bit field that specifies the upper-layer protocol to receive the packet's information. Two common values for this field are 5, which specifies Sequenced Packet Exchange (SPX), and 17, which specifies the NetWare Core Protocol (NCP).
  

• Destination network, destination node, and destination socket---Specify destination information.
  

• Source network, source node, and source socket---Specify source information.
  

• Upper-layer data---Contains information for upper-layer processes.
  

Although IPX was derived from XNS, it has several unique features. From the standpoint of routing, the encapsulation mechanisms of these two protocols are the most important difference. Encapsulation is the process of packaging upper-layer protocol information and data into a frame. For Ethernet, XNS uses standard Ethernet encapsulation, whereas IPX packets are encapsulated in Ethernet Version 2.0 or IEEE 802.3 without the IEEE 802.2 information that typically accompanies these frames. Figure 19-3 illustrates Ethernet, standard IEEE 802.3, and IPX encapsulation.