Voice and data over frame relay

In our Aug. 2 newsletter, we began addressing the question of whether converging voice and data onto frame relay and other packet networks was putting too many of the proverbial eggs in a single basket, since a network failure could then affect even greater volumes of traffic. As a follow-up, here are a few more issues to consider.

First, we mentioned the possibility of using inverse multiplexing techniques on the local access loops to provide additional flexibility so there is some degree of throughput even if one of the loops fails - kind of like having four engines on a big jet. An additional concern if you choose to take this approach is to consider physical route diversity. Inverse multiplexing will help if there is an electronic failure that affects only one of the circuits. But if these circuits are multiplexed onto common equipment or physical facilities, your protection is greatly diminished. One cable cut (a.k.a. backhoe fade) and all of the circuits are out of business. So physical route diversity - having separate cable in separate conduits that are not close to each other - becomes a key issue. (Maybe this will be a driver for cable modem access to frame relay services?)

We also mentioned the possibility of provisioning multiple permanent virtual circuits (PVC) between critical network locations, and the fact that any data that is in one of the switches is lost during the cutover time. Still, it should not be necessary to provision additional PVCs. One of the major advantages of using frame relay is the automatic alternate routing inherent in most switches. Nevertheless, the automatic alternate routing doesn't seem to be universally (or uniformly) implemented by all service providers, and this is an important factor you should quiz your provider about. We'll cover this issue in detail in an upcoming series.

As far as the "lost" data is concerned, this is not something to be overly concerned about. There's always a certain amount of information that is lost (or discarded) in networks. That's why we have protocols. If the information that is lost happens to be part of a data stream, the upper-layer transport protocols (such as TCP or SNA) recognize loss and request retransmission. If the information happens to be voice, such as voice over IP using User Datagram Protocol (UDP) as the transport layer, the voice algorithms are designed to accommodate for a fairly high percentage of lost packets.

---

Steven Taylor, consultant and broadband packet evangelist, and Joanie Wexler, an independent networking technology editor and writer, team up to bring you this analysis and commentary. Taylor specializes in education and market analysis, and Wexler adds incisive reporting and research. For more detailed information on most of the topics discussed in this newsletter, connect to www.webtorials.com, the first Web site dedicated exclusively to market studies and technology tutorials in the Broadband Packet areas of Frame Relay, ATM, and IP. Feedback and additional topic ideas are welcome. Please contact taylor@webtorials.com or joanie_wexler@mindspring.com>

Convergence? Try voice over frame
Network World, 06/07/99

Multilink and voice over frame relay
Network World, 05/24/99

Add salt before you fry the egg
Network World, 11/16/98

Voice over frame relay audio primer
Network World Fusion

Archive of Network World on Frame Relay newsletters