| 摘要 |
A "virtual-chassis" topology network includes three or more "front-plane" switches and two or more "back-plane switches". Front-plane switches can be added to increase the number of ports available for connecting network segments; port count can be exchanged for bandwidth by adding more back-plane switches. The virtual-chassis topology relies on an asymmetrical trunk mode in which each front-plane switch is operating in trunk mode, while the back-plane switches are not. The number of ports per trunk equals the number of back-plane switches so that each front-plane switch is coupled to every back-plane switch (and vice versa). Each back-plane switch is only coupled to front-plane switches, while the untrunked ports of front-plane switches are available for links to network segments. In this topology, every node device (belonging to a segment connected to a front-plane port) is separated by at most three switches from any other node device (also belonging to a segment connected to a front-plane port). The topology is fault-tolerant in that, if any link between the front plane and the back plane is broken, or if an entire back-plane switch fails, the network can reroute packets to avoid the broken link or links. Without impairing performance, the topology can be expanded so that the number of ports available to network segments is almost the square of the number of ports per front-plane switch. The availability of the virtual-chassis topology allows a network user to acquire even a single switch with the confidence that it will remain useful as the network expands.
|
