THE BASIC GUIDE TO
|
| Table of Contents | Basic Guide Title Pages |
|
|
Chapter One |
In this chapter, we will introduce you to frame relay and how it works. In other words, we'll give you the basic gear you'll need to continue your exploration.
Basic Trail: Go!
The Basic Trail will give you a definition of frame relay as a technology. Then, we'll explore the networking trends that combined to create a market need for frame relay. Finally, we'll talk about the benefits of using frame relay in your network.
Advanced Trail: Go!
If you take the advanced trail, you'll find a comparison of the characteristics of frame relay and other network switching technologies, namely Time Division Multiplexing (TDM), circuit switching and X.25 packet switching. (For a more detailed comparison of frame relay processing and X.25 processing, see Chapter 2.)
View Points :
Figure 1: Frame relay network Go
Figure 2: Opens Systems Interconnection (OSI) Model Go
Table 1: Comparison chart of TDM circuit switching, X.25, and frame relay Go
Shortcut: Go!
If you have only a few minutes, take the shortcut. It will give you a quick overview of how frame relay was developed and the benefits of frame relay.
What is Frame Relay?
Frame relay is a high-speed communications technology that is used in hundreds of networks throughout the world to connect LAN, SNA, Internet and even voice applications.
Simply put, frame relay is a way of sending information over a wide area network (WAN) that divides the information into frames or packets. Each frame has an address that the network uses to determine the destination of the frame. The frames travel through a series of switches within the frame relay network and arrive at their destination.
Frame relay employs a simple form of packet switching that is well-suited to powerful PCs, workstations and servers that operate with intelligent protocols, such as SNA and TCP/IP. As a result, frame relay offers high throughput and reliability that is perfect for a variety of today's business applications.
A Quick Look at a Frame Relay Network
A frame relay network consists of endpoints (e.g., PCs, servers, host computers), frame relay access equipment (e.g., bridges, routers, hosts, frame relay access devices) and network devices (e.g., switches, network routers, T1/E1 multiplexers).
Accessing the frame relay network using a standard frame relay interface, the frame relay access equipment is responsible for delivering frames to the network in the prescribed format. The job of the network device is to switch or route the frame through the network to the proper destination user device. (See Figure 1)
Figure 1: Frame Relay Network
A frame relay network will often be depicted as a network cloud, because the frame relay network is not a single physical connection between one endpoint and the other. Instead, a logical path is defined within the network. This logical path is called a virtual circuit. No bandwidth is allocated to the path until actual data needs to be transmitted. Then, the bandwidth within the network is allocated on a packet-by-packet basis.
We'll be talking a lot more about virtual circuits -- both Permanent Virtual Circuits (PVCs) and Switched Virtual Circuits (SVCs) in the next chapter. And we'll also discuss how frames or packets are "relayed" across the network.
But, before we get too technical, let's turn our attention to how and why frame relay got its start.
Why was Frame Relay Developed?
From the beginning, frame relay was embraced enthusiastically by users because it was developed in response to a clear market need, namely the need for high speed, high performance transmission. Frame relay technology also made cost-effective use of widespread digital facilities and inexpensive processing power found in end user devices. Developed by and for data communications users, frame relay was simply the right technology at the right time. Let's explore the network trends that contributed to the development of frame relay.
As the 1980s came to a close, several trends combined to create a demand for and enable higher speed transmission across the wide area network:
Need for Increased Speed
Today, rapid storage and retrieval of images for interactive applications is as common as transmitting full screens of text was in the 1970s and 1980s. Early graphics applications users who were accustomed to rapid information transfer over their LANs expected similar response times when transmitting data over the wide area. Since peak bandwidth requirements for graphics were substantially higher than for text transactions, increased bandwidth and throughput were clearly needed if response time expectations were to be met.
Dynamic Bandwidth Requirements
This type of LAN user required high bandwidth in bursts, followed by periods of idle time. "Bursty" traffic, as we call it, is well-suited for statistical sharing of bandwidth, which is a characteristic of frame relay technology.
Smarter Attached Devices
As networking requirements were changing, computing power was changing as well. Decreasing cost of processing power resulted in the proliferation of intelligent PCs and powerful workstations and servers all connected by LANs. These new end-user devices also offered the possibility of performing protocol processing, such as error detection and correction. This meant that the wide area network could be relieved of the burden of application layer protocol processing -- another perfect fit for frame relay.
End-user equipment was becoming more sophisticated in its ability to recognize errors and retransmit packets at the same time as digital facilities were reducing error rates within the network. In addition, industry-standard high layer protocols, such as TCP/IP, added intelligence to end-user devices.
Without the overhead associated with error detection and correction, frame relay could offer higher throughput than other connectivity solutions, such as X.25.
Higher Performance
More LANs in general and Internet Protocol (IP) LANs in specific, fueled the need to internetwork LANs across the wide area network, another factor that drove the growth of public frame relay services.
Some users tried to solve the internetworking challenge by simply hooking LAN bridges or routers together with dedicated lines. This approach worked for simple networks, but as complexity increased, the drawbacks became apparent: higher transmission costs, lower reliability, limited network management and diagnostics, and hidden inefficiencies.
It soon became apparent that a better approach to LAN internetworking was to connect bridges and routers into a reliable, manageable WAN backbone designed to make the best use of facilities and offer the high performance users demanded.
Frame relay technology offered distinct advantages for the wide area network. First, it was a more efficient WAN protocol than IP, using only five bytes of overhead versus 20 for IP. In addition, frame relay was easily switched. IP switching was not widely available in the WAN, and IP routing added unnecessary delays and consumed more bandwidth in the network.
Widespread Digital Facilities
As the public telecommunications infrastructure migrated from analog facilities to high quality digital facilities, bandwidth availability increased and error rates decreased. The error correcting capabilities of X.25 and SNA, which were developed to cope with the inherent errors of analog lines, were no longer necessary in digital wide area networks.
In the Beginning
While telecommunications managers contemplated the task of how to manage growing user requirements and increased network complexity, frame relay was being conceived in Bell Labs as part of the ISDN specification. Soon, frame relay evolved into a network service in its own right.
In 1990, four companies collaborated to refine the frame relay specification. "The Gang of Four," as they were known, later formed the Frame Relay Forum, which was incorporated in 1991. Since its inception, the Frame Relay Forum has grown to more than 300 members, evidence of widespread acceptance of frame relay as the method of choice for high-speed networks.
We'll discuss the work of the Frame Relay Forum in more detail in Chapter 4. Now, let's focus on the benefits of frame relay as a technology.
Banking on Frame Relay
The success of new technology is often dependent upon compelling economic reasons for implementation. Users of frame relay have found that it provides a number of benefits over alternative technologies:
1. lower cost of ownership
2. well-established and widely-adopted standards that allow open architecture and plug-and-play service implementation
3. low overhead combined with high reliability
4. network scalability, flexibility and disaster recovery
5. interworking with other new services and applications, such as ATM
Cost of Ownership
Frame relay provides users a lower cost of ownership than competing technologies for a number of reasons:
It supports multiple user applications, such as TCP/IP, NetBIOS, SNA and voice, eliminating multiple private lines to support different applications at a single site.
It allows multiple users at a location to access a single circuit and frame relay port, and it efficiently uses bandwidth, thanks to its statistical multiplexing capability.
Because only a single access circuit and port are required for each site, tremendous savings are often realized in recurring costs of transmission facilities.
Customers realize a significant reduction in hardware, such as the number of router cards and DSU/CSUs required, reducing up-front costs and ongoing maintenance costs when compared with point-to-point technologies.
Standards
Well established, widely-adopted standards are key to equipment interoperability and efficient use of capital acquisition funds. With frame relay, users can relax knowing that frame relay standards are in place both in the United States and around the world. This ensures that equipment and services provided today will be functional for the long run, with constantly evolving standards to support new applications and meet dynamic marketplace needs.
Low Overhead and High Reliability
By using only two to five bytes of overhead, frame relay makes efficient use of each frame. This means that more of the frame relay bandwidth is used for sending user data and less for overhead. Bandwidth utilization of frame relay is nearly equivalent to leased lines and better than numerous other technologies, such as X.25 or IP switching.
Network Scalability, Flexibility and Disaster Recovery
To the end user, a frame relay network appears straightforward: one user simply connects directly to the frame relay cloud. A frame relay network is based on virtual circuits which may be meshed or point-to-point, and these links may be permanent or switched. (See Chapter 2 for more details.)
Because of this structure, frame relay is more easily scalable than a fixed point-to-point network. This means that additions and changes in a network are transparent to end users, giving telecommunications managers the flexibility to modify network topologies easily and scale networks as applications grow and sites are added.
This inherent flexibility lends itself equally well to the provision of alternate routes to disaster recovery sites, which are, in many cases, transparent to the end user.
Interoperability with New Applications and Services
Compared with using point-to-point leased lines, frame relay suits meshed networks and hub and spoke networks equally well. This means that frame relay easily accommodates new applications and new directions of existing networks, for example, SNA migration to APPN.
In addition, frame relay standards have been developed to interwork with newly evolving services such as ATM. As new applications emerge and/or bandwidth requirements increase, networks can gracefully migrate to the appropriate technology without stranding existing network equipment.
If you're an experienced hiker, we've provided a little more of a challenging trail in this section. Here, you'll read why frame relay offers advantages over other technologies.
Frame Relay: The Right Mix of Technology
Frame relay combines the statistical multiplexing and port sharing features of X.25 with the high speed and low delay characteristics of TDM circuit switching. Defined as a "packet mode" service, frame relay organizes data into individually addressed units known as frames rather than placing it in fixed time slots. This gives frame relay statistical multiplexing and port sharing characteristics.
Unlike X.25, frame relay completely eliminates all Layer 3 processing. (See Figure 2.)
Figure 2: Open Systems Interconnection (OSI) Model
Only a few Layer 2 functions, the so-called "core aspects," are used, such as checking for a valid, error-free frame but not requesting retransmission if an error is found. Thus, many protocol functions already performed at higher levels, such as sequence numbers, window rotation, acknowledgments and supervisory frames, are not duplicated within the frame relay network.
Stripping these functions out of frame relay dramatically increases throughput (i.e., the number of frames processed per second for a given hardware cost), since each frame requires much less processing. For the same reason, frame relay delay is lower than X.25 delay, although it is higher than TDM delay, which does no processing.
In order to remove this functionality from the frame relay network, end devices must ensure error-free end-to-end transmission of data. Fortunately, most devices, especially those attached to LANs, have the intelligence and processing power to perform this function.
Table 1 summarizes the characteristics of TDM circuit switching, packet switching, and frame relay.
|
TDM Circuit Switching
|
X.25 Packet Switching
|
Frame Relay
|
|
|
Time-slot multiplexing
|
yes
|
no
|
no
|
|
Statistical (virtual circuit) multiplexing
|
no
|
yes
|
yes
|
|
Port sharing
|
no
|
yes
|
yes
|
|
High speed (per $)
|
yes
|
no
|
yes
|
|
Delay
|
very low
|
high
|
low
|
Table 1: Comparison of TDM circuit switching, packet switching and frame relay
Frame relay uses a variable length framing structure, which, depending on user data, ranges from a few to more than a thousand characters. This feature, similar to X.25, is essential for interoperability with LANs and other synchronous data traffic, which requires variable frame size. It also means that traffic delays (although always lower than X.25) will vary, depending on frame size. Some traffic types do not tolerate delay well, especially variable delay. However, frame relay technology has been adapted to carry even delay-sensitive traffic, such as voice.
As the 1980's came to a close, several network trends combined to create a need for a new form of wide area network switching:
This new wide area switching technology required high speed, low delay, port sharing, and bandwidth sharing on a virtual circuit basis. While existing TDM circuit switching and X.25 packet switching had some of these characteristics, only frame relay offered a full complement. These characteristics make frame relay an ideal solution for the bursty traffic sources found in LAN-WAN internetworking.
Frame relay provides a number of benefits over alternative technologies:
Frame relay offers users the ability to improve performance (response time) and reduce transmission costs dramatically for a number of important network applications. In order to be effective, frame relay requires that two conditions be met:
1. the end devices must be running an intelligent higher-layer protocol
2. the transmission lines must be virtually error-free
Other wide area network switching technologies, such as X.25 packet switching and TDM circuit switching, will remain important where line quality is not as good, when the network itself must guarantee error-free deliver, or when the traffic is intolerant of delay.
END CHAPTER ONE.
GO TO CHAPTER TWO
