Handbook of Local Area Networks, 1998 Edition:LAN Interconnectivity Basics
Click Here!
Search the site:
ITLibrary
ITKnowledge
EXPERT SEARCH
Programming Languages
Databases
Security
Web Services
Network Services
Middleware
Components
Operating Systems
User Interfaces
Groupware & Collaboration
Content Management
Productivity Applications
Hardware
Fun & Games
EarthWeb sites
Crossnodes
Datamation
Developer.com
DICE
EarthWeb.com
EarthWeb Direct
ERP Hub
Gamelan
GoCertify.com
HTMLGoodies
Intranet Journal
IT Knowledge
IT Library
JavaGoodies
JARS
JavaScripts.com
open source IT
RoadCoders
Y2K Info
Previous
Table of Contents
Next
CIDR Fundamentals
The technique of hierarchical routing was used in IP from the beginning. Each data link subnetwork was viewed as a segment, and the addressing information about all the hosts on that subnetwork was aggregated into a single network number. Therefore, initially the IP routing system had a single level of hierarchy. Introduction of subnets added another (second) level of hierarchy. By using subnets, a segment that consists of several data link subnetworks could be formed, and these subnetworks would be assigned subnet numbers out of a common network number. Then it would be possible to aggregate addressing information for all the hosts on all these subnetworks into a single network number. Aggregation at the level of network numbers was fairly constrained by forcing the aggregated information to fit either into a Class C, or Class B, or Class A network. This constraint had little or no justification other than preserving the notion of network number classes.
CIDR improved scalability of the Internet routing system by preserving routing information aggregation at the level of individual data link subnetworks, and by replacing aggregation at the level of network numbers with two levels of aggregation—at the level of individual subscribers, and at the level of individual providers. Therefore, CIDR put in place three levels of hierarchical address information aggregation—subnetwork, subscriber, and provider.
To enable hierarchical address information aggregation at the level of subscribers and providers, CIDR deprecated the notion of network number and network number classes. With CIDR, an IP address structure consists of two components—an address prefix part and a host number part, with no predetermined boundaries between the prefix and the host number parts. The host number part remained the same as in the original address structure. Rather than forcing top-level aggregates and aggregates at the level of individual subscribers to fall into one of the three network number classes, CIDR allows aggregated addressing information to be expressed as an address prefix of variable-length.
Subnetwork-level Address Information
Very similar to the subnet model, with CIDR each data link subnetwork is assigned a unique address prefix. In this way, CIDR retains the ability to aggregate addressing information at the level of individual data link subnetworks. In that sense, a router that interconnects a subnetwork with other subnetworks acts as an aggregation point for addresses of all the hosts on that subnetwork.
Individual Subscriber-level Information
To provide aggregation of addressing information at the level of individual subscribers, CIDR address allocation recommends that each subscriber would be given a single contiguous block of IP addresses (a single address prefix). The subscriber would use these addresses for its internal address assignment. A router at the border of the subscriber (e.g., a router that interconnects the subscriber with its provider) would aggregate the addressing information for all the nodes within the subscriber into a single address prefix. Therefore, the router would act as an aggregation point for addresses of all the hosts within the subscriber. As a result of aggregation at the subscriber’s boundaries, routers outside the subscriber would be able to maintain just a single route that would allow to forward to all the destinations within the subscriber.
Provider-level Information
To provide aggregation of addressing information at the level of providers, CIDR address allocation recommends that each provider would be given a single contiguous block of IP addresses (a single address prefix). The provider would then subdivide this block into smaller subblocks (each covered by an address prefix), and give these subblocks to its subscribers. This way, routers within the provider would maintain individual routes for all the subscribers of the provider. When advertising addressing information to other providers, however, routers within the provider would aggregate the addressing information associated with these subscribers into a single address prefix.
Therefore, a router that interconnects the provider with other providers would act as an aggregation point for all the addresses of the subscribers of that provider. As a result, routers within a provider would not need to maintain individual routes for the subscribers in other providers. Instead, the routers would maintain individual routes just for each provider. It is precisely this feature of CIDR that enables the Internet routing system to scale less than linearly with the number of subscribers.
To illustrate the aggregation levels used by CIDR, an example is shown in Exhibit 3-2-3. Hosts H1, H2, and H3 are connected to Ethernet 1, and are on a common IP subnet. Router R1 acts as an aggregation point (at the data link subnetwork level) for the addresses of these hosts—R1 advertises to R2 a single route that covers addresses of all the hosts on Ethernet 1. Likewise, router R2 acts as an aggregation point (at the data link subnetwork level) for the addresses of hosts H7, H8, and H9, and advertises to R1 a single route that covers addresses of all the hosts on Ethernet 3.
Exhibit 3-2-3. Aggregation Levels Used By CIDR
If addresses of all the hosts within subscriber A are assigned out of a single block (single address prefix), R2 acts as an aggregation point (at the subscriberís level) for subscriber A—R2 advertises to R3 a single route that covers addresses of all the hosts within A. This way other routers within Provider X do not need to maintain routes to individual subnets within A, but could instead maintain a single route.
If the address block used by A is taken out of the Provider XUs block, routers that interconnect X with other providers (R4 and R5 in this example) act as aggregation points at the providerís level—R4 and R5 advertise to R6 and R7 a single route that covers addresses of all the destinations for all the XUs subscribers. This way routers in Providers Y and Z do not need to maintain routes to individual subscribers of X, but could instead maintain a single route.
Previous
Table of Contents
Next
Use of this site is subject certain Terms & Conditions.
Copyright (c) 1996-1999 EarthWeb, Inc.. All rights reserved. Reproduction in whole or in part in any form or medium without express written permission of EarthWeb is prohibited.
Please read our privacy policy for details.