Bridge and Switch Forwarding Facts
Both bridges and switches build a forwarding database. The database is a list of Data Link (MAC) addresses and the port used to reach the device. Bridges and switches can automatically learn about devices to build the forwarding database. A network administrator can also program the device database manually. Bridges and switches use the following process to dynamically build the forwarding database:
- The process begins by examining the source address of an incoming packet. If the source address is not in the forwarding database, an entry for the address is made in the database. The port it came in on is also recorded.
- The destination address is then examined.
- If the destination address is in the database, the packet is forwarded to the appropriate port if the port is different than the one on which it was received.
- If the destination address is not in the database, the packet is sent out all ports except for the one on which it was received. This is known as flooding.
- A broadcast packet is forwarded (flooded) to all ports except the one on which it was received.
Transparent bridges forward packets only if the following conditions are met.
- The frame contains data from the layers above the Data Link layer.
- The frame's integrity has been verified through a valid Cyclic Redundancy Check (CRC).
- The frame is not addressed to the bridge.
How switches forward packets depends on the switch type. The following table compares the different methods the switch uses to forward packets (some Cisco switches support all three methods).
|
Method |
Characteristics |
|
Store-and-forward |
Store-and-forward switches:
|
|
Cut-through |
Cut-through switches:
|
|
Fragment-free |
Fragment-free switches:
|
Note: Newer switches can monitor each port and determine which switching method to use. They can automatically change to store-and-forward if the number of errors on a port exceeds a configurable threshold.