VPN technology has matured to a level where it is believed to provide the same level of security and confidentiality afforded by private WANs, using the Internet as transport medium. It is much cheaper to deploy VPNs over the Internet than private WANs.
The service model of MPLS is the same as that of Frame Relay. However, MPLS enables service providers to offer richer services and affords many technical advantages. MPLS is the technology of choice over Frame Relay for private WAN deployments today. Frame Relay is far from dead though due to the large existing installed base and its simplicity for point-to-point WANs connecting the branch office to corporate headquarters. Frame Relay is also used in combination with MPLS to provide Layer 2 circuits to the nearest MPLS point of presence (POP). Therefore, despite all what you have heard about Frame Relay being obsolete, it will continue to be an important networking topic for some time at least.
Packet Switching versus Circuit Switching
WAN technologies can usually be categorized as either circuit-switching or packet-switching. A electrical circuit is a system of conductors (wires) forming a complete path around which a current can flow. The original telephone systems actually created an electrical circuit between two phones in order to carry the voice signal. The leased lines used for carrying data are also circuits, providing the ability to transfer bits as signals between two end points. In telecommunications terminology today, a circuit refers to the physical path between two end points providing the ability to send voice or data from one end point to the other.
Packet switching, as a technology, is more complex than circuit switching. The devices involved in packet switching have to do more than simply passing bits as signals from one end point to another. The devices in the service provider’s network have to be intelligent for packet switching. This is in contrast to circuit switching where devices in the service provider’s network simply have to carry signals without making sense of them. With packet switching, the devices read the bits sent by customers interpreting usually some form of address field in the packet header. The address field in the packet header is used by the devices to make choices, switching one packet to go in one direction and the next packet to go possibly in another direction to another device.
Circuit switching is an old but expensive technoloy, and it is what the traditional telephone network known as the public switched telephone network (PSTN) uses. Packet switching is more modern and may eventually replace circuit switching completely. Meanwhile we have to live in a world that is a hybrid of the two technologies.
We will now cover Frame Relay thoroughly describing terminology, protocol details, and configuration.
Frame Relay Concepts
Frame Relay is more complex a technology than point-to-point WAN links but also provides more features and benefits. Frame Relay networks are multiaccess networks, which means that more than two devices can connect to the network. This is similar to LANs where more than two devices can attach to the same network and any two devices can communicate directly. However, unlike LANs you cannot send a broadcast at data link layer over Frame Relay. Therefore, Frame Relay networks are termed as non-broadcast multiaccess (NBMA) networks.
Figure 12-7 presents a Frame Relay topology showing its most basic components.
Figure 12-7 Frame Relay Components
A Frame Relay network is made up of a large number of Frame Relay switches dispersed all over the coverage area of a Frame Relay service provider. This coverage area may span a country, region, or even the whole world. The switches are interconnected in a complex mesh topology. Some Frame Relay switches also terminate user circuits, in addition to connecting to other switches, and are called access switches. Other Frame Relay switches do not terminate user circuits, connecting to other Frame Relay switches only, and make the backbone of the Frame Relay network.
A leased line is installed between the router at a customer site and the nearest Frame Relay switch. This leased line is called the access link. In the context of Frame Relay, the router is the data terminal equipment (DTE) while the Frame Relay switch is the data circuit-terminating equipment (DCE). To ensure that the link is working DTE and DCE exchange regular messages with each other. These keepalive messages, along with other messages, are defined by the Frame Relay Local Management Interface (LMI) protocol. Please keep in mind that the terms DTE and DCE have different meanings in different contexts and the terms here are used in the context of Frame Relay.
The physical connectivity from a Frame Relay DTE router to the Frame Relay network is the access link. However, the end goal is to provide end-to-end connectivity between two DTE routers across the Frame Relay cloud. The logical end-to-end communications path between two DTE device is known as a virtucal circuit (VC). The provisioning of virtual circuits is responsibiliy of the service provider, and these predefined virtual circuits are also known as permanent virtual circuits (PVC). Frame Relay routers use the data link connection identifier (DLCI) as the Frame Relay address. DLCI identifies the VC over which the frame should travel.
Key Concept: Committed Information Rate (CIR) is the average rate, in bits per second, at which Frame Relay switch agrees to transfer data for a customer.
Let’s now formally define some important Frame Relay terms before moving forward:
◈ Virtual circuit (VC) is a logical communications path that is used by frames travelling between DTEs.
◈ Permanent virtual circuit (PVC) is a permanently defined virtual circuit. PVC is analogous to a point-to-point leased line in concept.
◈ Switched virtual circuit (SVC) is set up dynamically when needed. An SVC is analogous to a dial-up connection in concept.
◈ Data terminal equipment (DTE) is a networking device like a router used by a customer to connect to the Frame Relay network of a service provider. The DTE typically resides at the customer site and is frequently referred to as customer premises equipment (CPE).
◈ Data circuit-terminating equipment (DCE) are the Frame Relay access switches that terminate customer access links and reside in the service provider network. The term DCE is also considered to mean data communication equipment by many.
◈ Access link is the leased line between the DTE (router) and DCE (Frame Relay switch).
◈ Access rate (AR) is the speed at which the access link is clocked. The access rate does not necessarily have to match the CIR. However in order to fully utilize the CIR, the access rate must be equal to or higher than the CIR.
◈ Committed information rate (CIR) is the speed at which the bits can be sent over a VC, according to the service contract between the Frame Relay service provider and its customer.
◈ Data link connection identifier (DLCI) is a Frame Relay address present in the header of every Frame Relay frame. DLCI is significant over a single hop only and different DLCI values may be used on different hops along a VC for the same packet.
◈ Non-broadcast multi-access (NBMA) is a network on which broadcasts are not supported but more than two device can be connected to the same network.
◈ Local Management Interface (LMI) is the protocol used between a DCE and DTE to manage the connection. LMI involves messages to establish SVCs, status messages for PVCs, and keepalives to mention a few.
Virtual Circuits
Frame Relay is a cost-effective alternate to point-to-point leased lines to build enterprise WANs. In the absence of Frame Relay, enterprises wishing to connect offices worlwide would have to lease very expensive international leased circuits to connect LANs through routers. A Frame Relay network is owned by a service provider offering services to companies that want to connect its locations to each other. Frame Relay virtual circuits act like point-to-point leased lines for the customer while providing significant cost benefits as compared with leased lines.
A virtual circuit (VC) spans the access links at the two ends as well as the Frame Relay network. For example, you can see two VCs in figure 12-8, one between R1 and R3 and the other between R2 and R3. Bold and grayed dashed lines have been used to represent VCs. You should keep in mind that the Frame Relay network is owned and operated by a service provider and is shared by many customers of the same service provider. Yet virtual circuits provisioned by the service provider for a certain customer create the illusion of a point-to-point dedicated circuit. Also the traffic from different customer is kept separate and Frame Relay networks built around this model are considered sufficiently secure.
Originally, when the world was moving from expensive private leased lines to the co-operative model of Frame Relay, customers were concerned about bandwidth because of the contention within the Frame Relay cloud with other customers for available capactity. In order to address these concerns, Frame Relay uses a concept of committed information rate (CIR). Each VC has a CIR, which is a guarantee by the provider that a particular VC would get that much bandwidth. So you can migrate from a private leased line to Frame Relay with a CIR equal to the leased line bandwidth.
Frame Relay service model requires one access from each site to the Frame Relay service provider, regardless of the number of sites to be interconnected. This is not the case if you want to build a WAN using private leased lines. In that case you would need N*(N-1) leased lines where N is the number of sites you are trying to connect. For example, if you have 3 sites you would require 3*(3-1)=6 leased lines, while for 10 sites the number of leased lines required steps up to 10*(10-1)=90 leased lines. This solution simply does not scale to large deployments. Though the access links required to connect a site to nearest Frame Relay point of presence (POP) are still private leased lines, but they are shorter and fewer.
When a Frame Relay network is designed, there may not be a VC between any pair of sites. If there is a PVC between any two sites, it is called a full-mesh topology. When not all pairs of sites have a direct PVC, it is called a partial-mesh topology. In most practical scenarios, partial mesh is used as not all customer sites typically need to connect to all other sites. For example, global enterprises typically use a star-topology which is a special case of partial-mesh topology. In a star topology a large number of remote branch offices are connected to the data center to access the resources including data and applications.
Posts
0 comments:
Post a Comment