Previous Table of Contents Next


Chapter 17
Frame Relay Testing and Training

Steve Greer, Peter Luff, and Sean Yarborough

Because it provides efficient, cost-effective transfer of bursty, bandwidth-intensive applications, frame relay is now the main access protocol for wide area network communications. This chapter discusses the advantages and disadvantages of frame relay, as well as the reasons why training and testing are important.

INTRODUCTION

Frame relay — a network access protocol for high-speed, bursty data applications — has been a data communications buzzword for the past five years, but only recently has it begun to serve its purpose. According to Vertical Systems Group, a consulting group in Dedham MA, the number of U.S. public frame relay network subscribers, which was only 590 in the early 1990s, is projected to increase to 8,210 by the late 1990s. The main catalyst for this explosion is the need for high-speed, LAN interconnection.

Frame relay networks provide companies with cost-effective and efficient transfer of such bursty bandwidth-intensive applications as file transfer, E-mail, graphics, and imaging applications. Frame relay decreases the cost of LAN interconnection through the use of statistical multiplexing, which allows many subscribers to utilize the bandwidth on a single circuit. The cost savings is a direct result of the reduction in the use of dedicated leased lines, which tend to be underutilized because bursty transmission is sporadic.

LANs have become so ubiquitous that subscribers expect to be able to communicate across multiple interconnected LANs just as easily as when they communicate over a single LAN. Frame relay addresses LAN interconnection extremely well (i.e., up to T1 speeds), so subscribers can enjoy the same quality data transmission over the WAN that they have come to expect over the LAN.

It is important that service providers and subscribers understand Frame relay technology and its advantages and disadvantages. For example, Frame relay has no inherent error correction. It assumes that a network successfully transmits data and avoids errors. As a result, users need to test the network for errors and comprehensively train employees on Frame relay technology to reduce errors. This chapter briefly reviews the features of Frame relay technology and explains some of the important tests to perform on Frame relay devices and the network.

ADVANTAGES OF PACKET-SWITCHING FRAME RELAY NETWORKS

The main advantage of frame relay is that it is a packet-switching technology. Packet-switching networks send data from source to destination based on each packet’s unique destination address. Once the data is packetized, it can be statistically multiplexed.

Statistical multiplexing allows many subscribers to share the same bandwidth by assuming that not all subscribers will be using the bandwidth at the same time. This avoids high-cost, point-to-point connections such as dedicated leased lines that employ circuit switching. Leased lines are expensive because they are rented and dedicated for exclusive use 24-hours a day, seven days a week. With leased lines, subscribers pay for bandwidth whether it is being used or not. Statistical multiplexing provides multiple data connections through the network simultaneously, and no single customer pays for exclusive privileges. This results in a significant cost advantage over circuit-switched networks.

A second advantage of Frame relay is that its variable-length frames and its low overhead provide excellent network throughput and low delay of data. The variable-length frames allow Frame relay to encapsulate protocols well. Frame relay is protocol independent, so its payload can carry a variety of higher-layer LAN protocols, such as the IP. Because the network does not concern itself with error correction and flow control, overhead in the Frame relay network is low. Therefore, the network uses most of its resources switching user data.

Frame relay realizes another advantage through the CIR. The CIR represents the data traffic level that the network plans to support under typical network conditions; it is agreed upon by the service provider and the subscriber. The advantage of a CIR for the customer is that once it is agreed upon, the service provider should be capable of transmitting at or below the CIR. For example, if an average of 56K-bps throughput is required between two sites, then the CIR should be equal to or greater than 56K bps. Typically, data traffic sent below the CIR passes through the network at a high priority, and data sent in excess of the CIR has a lower priority. This low-priority traffic is the first to be dropped when subscribers create network congestion. The CIR serves as the basis of a well-planned network as well as a billing mechanism.

Finally, frame relay has worldwide industry support from manufacturers, standards organizations, and service providers. This is important because it ensures that there is a high level of interoperability between devices of different manufacturers in different countries.

DISADVANTAGES OF FRAME RELAY

Although frame relay carries data more efficiently because it has no inherent error correction utilities, this lack of frame management can be detrimental. Frames can be errored because of transmission impairments in the network (i.e., a bit error corrupts a frame, which is then dropped by the network).

In Frame relay, the error correction is left up to the user’s intelligent devices (i.e., a router), which can discover discrepancies and request a retransmission. In addition, the reliable deployment of technologies such as T1, DDS, and fiber optics decreases the need for error correction within the network.

Network problems can occur when subscribers exceed the CIR. As previously mentioned, the CIR is the amount of data traffic, agreed upon by the service provider and the subscriber, the network is planned to support under normal network conditions. When too many subscribers exceed the CIR, then a situation may develop in which the network becomes congested and begins to drop frames to alleviate the congestion.

For example, 20 banks are all connected to a Frame relay network. During the day, they all transmit under their CIR for routine electronic communication. At 2:00 p.m., however, all of these banks send their daily transactions to another location for processing, which causes network congestion. If only one bank transmits over the CIR, network congestion probably will not occur. It is the aggregate of many banks transmitting over their CIR that causes congestion and dropped frames.

Frame relay frames are variable in length; therefore, they cause variable delay in the network. For example, a short frame can be switched quickly by the Frame relay network. However, longer frames take longer to process and switch. This creates a variable transit time between long and short frames. Integrated data, voice, and video applications cannot be delayed; therefore, these applications are not best suited to Frame relay. They are better suited to integrated switched digital networks (ISDN) or cell relay technology, such as ATM.


Exhibit 1.  The Frame Relay Frame.


Previous Table of Contents Next

Copyright © CRC Press LLC