AppleTalk Addressing and Address Structure

 

While TCP/IP ”which is examined in Chapter 4, "TCP/IP Basics" ”is described as an open protocol, AppleTalk is a proprietary protocol controlled by Apple Computers. AppleTalk has a unique network address structure and a unique naming methodology for network services.

This section explores the AppleTalk network address structure that all clients (also called workstations) and servers must have to communicate within an AppleTalk internetwork.

The AppleTalk network address is a 24-bit address consisting of two distinct components ”namely, a 16-bit network portion and an 8-bit node address. The network portion identifies a LAN or WAN segment, while the node address identifies a workstation or server. The two components are usually written together as network.node using decimal notation. For example, the address 52.6 identifies workstation or server 6 in network 52. Unlike TCP/IP, which has multiple levels of address hierarchy and summarization, AppleTalk is limited to these two levels. The DDP coordinates address administration within the AppleTalk network in addition to providing connectionless delivery of AppleTalk packets.

Network addresses for LAN and WAN segments are determined by the network administrator in the same way that TCP/IP subnets are assigned by the administrator to identify a network segment. AppleTalk identifies two different types of network addressing methods for LAN and WAN segments, AppleTalk Phases 1 and 2. In AppleTalk Phase 1, network segments are identified by a single network number.

In AppleTalk Phase 2, network segments are identified by a cable-range that corresponds to one or more logical network numbers. A cable-range is either one network number or a contiguous sequence of several network numbers specified by a starting and ending network number in the format start-end . For example, the cable-range 100 “100 identifies a logical network that has the single network number 100, while the cable-range 50 “64 identifies a logical network that spans 15 network numbers, from 50 through 64.

Each device in an AppleTalk network needs a node number with which to communicate with other devices. Unlike network protocols that require the network administrator to assign node or host addresses, an AppleTalk device determines its node address dynamically. As with the network portion of the address, AppleTalk Phases 1 and 2 have different requirements that control the selection of the node address during the negotiation process.

AppleTalk Phase 1 network segments may have up to 254 node addresses ”127 are reserved for workstations, and 127 are reserved for servers. Each workstation or server in the Phase 1 network segment must have a unique node number. In AppleTalk Phase 1, a logical network segment could support only 127 AppleTalk hosts . This proved to be a scalability issue that was solved in AppleTalk Phase 2.

AppleTalk Phase 2 network segments have two classifications for node addresses, which are called extended and nonextended. On a nonextended Phase 2 network segment, 253 node numbers can be associated with a single network address on the segment. Each server or workstation is assigned a unique node address in the range 1 “253. Extended Phase 2 network segments also allow for the assignment of node addresses in the range 1 “253. However, because multiple network numbers may exist on the segment (via the cable-range), each workstation or server is assigned a unique combination network.node address. The difference between extended and nonextended addresses might seem a bit subtle. In a nutshell , an extended network can support multiple network numbers, and a nonextended network can support only a single network address.

Note

Phase 2 nonextended network segments are usually either LocalTalk networks or WAN segments. LocalTalk is Apple's first implementation of networking at the data link and physical layers that uses the telephone cable as the physical transport and carrier sense multiple access collision detect (CSMA/CD) at the data link layer. LocalTalk and AppleTalk Phase 1 were developed for workgroup applications. AppleTalk Phase 2 resulted from the need to enhance the scalability of the AppleTalk protocol to support deployment on an enterprise-wide scale. Because many of the same characteristics are shared between AppleTalk Phase 1 and nonextended Phase 2 network segments, you can think of a Phase 1 network segment as simply a nonextended Phase 2 network segment.

Cisco routers have never supported LocalTalk, although WAN segments may be addressed in AppleTalk Phase 1 style. We recommend, however, that Phase 2 addressing be used exclusively when configuring Cisco devices for the sake of consistency, clarity, and flexibility.


As mentioned previously, the node address is negotiated dynamically at the time an AppleTalk device boots or is reset. AARP is responsible for negotiating node addresses for devices on a network segment. Dynamic address assignment is accomplished using a very simple algorithm. Any time an AppleTalk device is rebooted and attempts to attach to the network, it checks to see whether a network address has been previously assigned to it. If so, the device sends out an AARP packet to verify that the address is still valid and has not been claimed by another node on the network segment. If available, the address is used, and the node begins normal network operations. If the address has been claimed, the node sends out a series of additional AARP packets proposing a new node address until a valid address is found. Figure 5-2 depicts the address negotiation process.

Figure 5-2. AppleTalk Node Address Selection Process

graphics/05fig02.gif

To enhance the user 's interaction with the AppleTalk network, Apple decided that users should be shielded from knowing the specifics of network and node addressing. Rather than knowing that workstation 5 in network 10 wants to communicate with server 8 in network 20, the user needs to know only device names. Apple created a naming scheme that allows for the logical grouping of workstations and for the assignment of individual names to individual workstations and servers. The term used for a logical collection of workstations or servers is a zone .

Zones can be defined for any logical characteristic of an organization, such as its distinct operations, departments, and geographical locations. For example, a company might create a Marketing Zone, a Sales Zone, and an Engineering Zone, all of which might cross multiple geographies. Alternatively, a company might have a New York City Zone that encompasses all the organizational functions for the identified geographical area. The selection and assignment of zone names is completely at the discretion of the network administrator. To accommodate logical grouping across multiple physical LAN or WAN segments, the administrator can apply the same zone to multiple networks. Additionally, a network segment can be assigned multiple zone names to accommodate the different logical groups that can have network resources attached to that segment.

In contrast to zone names, which are determined by the network administrator, the names for individual workstations and servers are determined by the user or administrator of that device. An individual may name a workstation John's Mac or Godzilla, while a server administrator may name a server after its function, such as Finance or Publications. These names, along with the zone in which they reside, are registered with the network shortly after device startup by the NBP.

NBP associates AppleTalk names and device attributes with addresses. It orchestrates the name-binding process, including name registration, name confirmation, name deletion, and name lookup. When names are registered in NBP, the earlier example of workstation 10.5 wanting to communicate with server 20.8 might be expressed as follows : John's Mac in the New York Zone wants to communicate with server Finance in the Accounting Zone. As you can see, NBP enables users to refer to network resources by names, much like the domain name service (DNS) of TCP/IP.

Note

Cisco IOS devices use the name assigned with the global command hostname to register with NBP. A Cisco router registers itself with NBP as type ciscoRouter. NBP associations can be viewed by using the IOS EXEC command show appletalk nbp , which is examined later in the section "Verifying AppleTalk Connectivity and Troubleshooting."


While zone name assignment is not part of the network address, it is an integral part of the proper operation of an AppleTalk network. Proper configuration of AppleTalk on routers requires that zones be assigned in addition to network numbers or cable-ranges.

Table 5-1 summarizes the differences between the various network and node numbering requirements.

Table 5-1. AppleTalk Phase 1 and Phase 2 Capabilities
Capability AppleTalk Phase 1 AppleTalk Phase 2
Networks, nodes, and zones    
Number of logical networks (cable segments) 1 65279
Maximum number of devices 254 [1] 253 [2]
Maximum number of end nodes 127 No limit on end nodes; 253 total node limit
Maximum number of servers 127 No limit on servers; 253 total node limit
Number of zones in which a network can exist 1 1 (nonextended); 255 (extended)
Media-level encapsulation    
Nonextended network Does not apply Yes
Extended network Does not apply Yes
Cable addressing Does not apply; uses network numbers Single network number (nonextended); cable range of 1 or more (extended)

[1] Node numbers 1 and 255 are reserved

[2] Node numbers 0, 254, and 255 are reserved



Cisco Router Configuration
Cisco Router Configuration (2nd Edition)
ISBN: 1578702410
EAN: 2147483647
Year: 1999
Pages: 116

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net