The performance of most LANs exceeds the capacity available in the wide area. This section asesses the performance of 10 and 100 Mbps LAN technology using Ethernet switches and hubs. The theoretical capacity of various topologies is summarised below:
Theoretical Maximum Capacity of Various Topologies | ||||
Medium | Mode | Max Server Rate | Max Client Rate | Max Util * |
10 Mbps | Shared | 10 Mbps | 10 Mbps | 2-10 Mbps ** |
Switched | 10 Mbps | 10 Mbps | 10 Mbps | |
100 Mbps | Shared | 100 Mbps | 100 Mbps | 2-80 Mbps ** |
Switched | 100 Mbps | 100 Mbps | 80 Mbps | |
100/10 Mbps | Switched | 100 Mbps | 10 Mbps | 10 Mbps |
155/10 Mbps | Switched | 155 Mbps | 10 Mbps | 10 Mbps |
* Maximum utilisation assuming full duplex operation.
** Maximum utilisation depends on number of nodes simultaneously attempting transmission.
The following delivery network topologies are examined:
The performance of the Local Area Network will be measured at ERG (Ethernet
and WAN measurements) and Forester Hill sites (ATM and Ethernet measurements).
The following LAN measurements were performed using the ERG JTAP Test Network. The throughput perfromance for UDP packets from both 100 Mbps and 10 Mbps interfaces were measured using the performance measurement tools developed for the JTAP project. The measured performance are reported in this section. These results are compiled from the preliminary measurements. They will be further analysed and extended in the near future.
Host interface performance is to be analysed for the delivery PCs and SUN Ultra servers.
Testing-Home; Tools; lan-Tests; wan-Tests; MediCAL Home