Abstract

Testing Remote Ethernet Bridges

This May 21, 1994 Data Communications lab test, conducted by Kevin Tolly of The Tolly Group and David Newman of Data Communications, evaluated remote Ethernet bridges designed to reduce WAN costs by compressing LAN traffic across low-speed links. The article focused on whether vendors’ compression claims held up under realistic testing and whether these products could make remote Ethernet bridging practical over 56Kbit/s links and other low-bandwidth WAN services.

The lab tested six remote Ethernet bridge products from Magnalink Division of Telco Systems, Microcom, Newport Systems Solutions, RAD Network Devices, Retix, and 3Com. The products were evaluated for compression effectiveness, latency, WAN-link support, filtering, SNMP management, ease of configuration, documentation, technical support, and overall suitability for branch-office connectivity. The test bed used two 10Mbit/s Ethernet LANs connected by a simulated 56Kbit/s WAN link, with traffic generated and measured using tools from LANquest Labs, Network General, Wandel & Goltermann, and other test equipment.

The results were encouraging for organizations trying to stretch low-speed WAN links. All six products compressed text traffic by at least 3:1, and several delivered much higher effective throughput. Magnalink’s Series 3000 Compression Bridge/Router produced the strongest compression results, reaching nearly 7:1 on a 56Kbit/s link. Newport’s LAN2LAN/MPR was close behind, also approaching 7:1 in the text-compression tests. RAD Network Devices’ Remote Compression Bridge exceeded 4:1 and stood out for its simple “plug-and-play” behavior. Retix, 3Com, and Microcom also compressed traffic, though with lower effective throughput in the primary comparison.

Magnalink’s Series 3000 Compression Bridge/Router received a Data Communications “Tester’s Choice” designation. The article praised its exceptional compression performance and its ability to force tougher test conditions because its results were so strong. RAD Network Devices’ Remote Compression Bridge also received a “Tester’s Choice” designation. Although it lacked some of the advanced features found in more expensive products, the RCB was highlighted for its easy installation, strong value, and better-than-4:1 compression.

The article also emphasized that compression ratio alone was not the full story. Some products introduced more latency as they buffered traffic, and management features varied widely. The lab examined whether network managers could change WAN frame size, change link speed, disable compression, configure filters, and alter LAN MAC-address settings remotely. Products also differed in support for MAC, SAP, SNAP, NetBIOS, and custom filters. These controls mattered because filtering unnecessary LAN traffic could be as important as compression in reducing WAN utilization.

Overall, the lab concluded that remote Ethernet bridges could deliver real WAN savings and practical branch-office connectivity when properly selected and configured. The best products combined strong compression, low administrative burden, useful filtering, and reliable management. For companies using low-speed links, these bridges offered a practical way to improve effective throughput without immediately upgrading WAN services.

Solutions Tested

Magnalink Division of Telco Systems — Series 3000 Compression Bridge/Router
Microcom — MBR/6003
Newport Systems Solutions — LAN2LAN/MPR
RAD Network Devices — Remote Compression Bridge (RCB)
Retix — 4810 LAN Bridge
3Com — NetBuilder