This chapter provides an introduction to common dial networking scenarios used by service providers and enterprises.
Providing dial access means to set up one or more access servers or routers to allow on-demand connectivity for individual remote nodes or remote offices. The dial network solutions described in this chapter are based on business case scenarios. Depending on your business application, dial access has different implementations.
This chapter provides the following sections:
Service providers tend to supply public and private dial-in services for businesses or individual home users. Enterprises tend to provide private dial-in access for employees dialing in from remote LANs (such as a remote office) or individual remote nodes (such as a telecommuter). Additionally, there are hybrid forms of dial access---virtual private dial networks (VPDNs)---that are jointly owned, operated, and set up by both service providers and enterprises.
Figure 132 displays a common dial topology used by an Internet service provider (ISP). The central dial-in site is owned and controlled by the ISP, who only accepts dial-in calls. Enterprises and individual remote clients have no administrative control over the ISP's point-of-presence (POP).
Enterprises can provide bidirectional access services with remote LANs and one-way dial-in access for standalone remote nodes. Bidirectional access means that remote LANs can dial in to the enterprise, and the enterprise can dial out to the remote LANs. A remote LAN can be a large remote office or a small home office. A standalone remote node can be an individual PC that is dynamically assigned an IP address from the enterprise's modem pool. In most cases, an enterprise has complete administrative control over its local and remote devices. (See Figure 133.)
Service providers and enterprises both benefit from a hybrid dial solution called VPDN. Service providers offer virtually private access to enterprises by providing the dial-in access devices for the enterprise's use (for example, access servers and modem pools). In this solution, service providers construct the networking fabric for city-to-city dial connectivity for the enterprise. Enterprises provide only a home gateway router (with no attached modems) and a WAN connection to their service provider. VPDN dial solutions enable the enterprise to continue to maintain complete administrative control over its remote locations and network resource privileges. (See Figure 134.)
The hardware and software configuration designs for dial networks are derived from business operations needs. This section describes several of the most common business dial scenarios that Cisco Systems is supporting for basic IP and security services.
Refer to the scenario that best describes your business or networking needs:
Exponential growth in the remote access router market has created new addressing challenges for ISPs and enterprise users. Companies that use dial technologies seek addressing solutions that will:
Remote networks have variable numbers of end systems that need access to the Internet; therefore, some ISPs are interested in allocating just one IP address to each remote LAN.
In enterprise networks where telecommuter populations are increasing in number, network administrators need solutions that ease configuration and management of remote routers and provide conservation and dynamic allocation of IP addresses within their networks. These solutions are especially important when network administrators implement large dial-up user pools where ISDN plays a major role.
Use an IP addressing scheme that is appropriate for your business scenario as described in the following sections:
Additionally, here are some addressing issues to keep in mind while you evaluate different IP address strategies:
1. How many IP addresses do you need?
2. Do you want remote clients to dial in to your network and connect to server based services, which require statically assigned IP addresses?
3. Is your primary goal to provide Internet services to a network (for example, surfing the web, downloading e-mail, using TCP/IP applications)?
4. Can you conduct business with only a few registered IP addresses?
5. Do you need a single contiguous address space or can you function with two non contiguous address spaces?
This section describes two classic IP addressing strategies you can use to set up dial-in access. Classic IP addresses are statically or dynamically assigned from your network to each site router or dial-in client. The IP address strategy you use depends on if you are allowing remote LANs or individual remote clients to dial in.
A remote LAN usually consists of a single router at the gateway followed by multiple nodes such as 50 PCs. The IP address on the gateway router is fixed or statically assigned (for example, 220.127.116.11). This device always uses the address 18.104.22.168 to dial in to the enterprise or service provider network. There is also a segment or subnet associated with the gateway router (for example, 22.214.171.124 255.255.255.0), which is defined by the dial-in security server.
For individual remote clients dialing in, a specific range or pool of IP addresses is defined by the gateway access server and dynamically assigned to each node. When a remote node dials in, it receives an address from the specified address pool. This pool of addresses usually resides locally on the network access server. Whereas, the remote LANs have predefined or statically assigned addresses. The accompanying subnet is usually statically assigned too. (See Figure 135.)
Here are some advantages and disadvantages of manually assigning IP addresses:
Two of the key problems facing the Internet are depletion of IP address space and scaling in routing. The Easy IP (Phase 1) feature combines Network Address Translation (NAT) and PPP/Internet Protocol Control Protocol (IPCP). This feature enables a Cisco router to automatically negotiate its own registered WAN interface IP address from a central server and allows all remote hosts to access the global Internet using this single registered IP address. Because Easy IP uses existing port-level multiplexed NAT functionality within the Cisco IOS software, IP addresses on the remote LAN are invisible to the Internet.
Cisco IOS Easy IP solution is a scalable, standards-based, "plug-and-play" solution that is comprised of a combination of the following technologies:
Figure 136 shows an example of how Easy IP works. A range of registered or unregistered IP addresses are used inside a company's network. When a dial-up connection is initiated by an internal node, the router uses the Easy IP feature to rewrite the IP header belonging to each packet and translate the private address into the dynamically assigned and registered IP address, which could be borrowed from a service provider.
For a more detailed description of how Easy IP (phase 1) works, refer to the chapter "Configuring Easy IP" later in this document.
The Easy IP feature provides the following benefits:
For step-by-step configuration information on how to set up the Easy IP feature on a router or access server, refer to the chapter "Configuring Easy IP" later in this document.