Configuring Fa0/1 on SwitchA for fullduplex mode and configuring Fa0/1 on SwitchD to autonegotiate duplex settings could cause a switching loop. A switch port that has been manually configured to use full duplex or half duplex mode does not respond to a port that is attempting to autonegotiate duplex settings. When the autonegotiating port receives no reply, it will use the default duplex settings. A port configured to transmit at 100 Mbps defaults to half duplex mode, and a port configured to transmit at 1000 Mbps defaults to full duplex mode. Therefore, Fa0/1 on SwitchD will use half duplex mode, causing a duplex mismatch with Fa0/1 on SwitchA.

Duplex mismatches can cause collisions, alignment errors, and intermittent connectivity. You can detect a duplex mismatch by monitoring a switch for %CDP-4-DUPLEX-MISMATCH error messages. Additionally, you can issue the show interfaces interface command, which displays interface counter information. If you see an abnormal increase in frame check sequence (FCS) errors and alignment errors on a half duplex port, you should suspect a duplex mismatch. An abnormal increase in FCS errors and runts on a full duplex port is also an indicator of a duplex mismatch.

When SwitchD is in half duplex mode, it performs carrier sense to determine whether the link is clear before sending packets. However, SwitchA does not perform carrier sense, because it is configured for full duplex mode; this is what causes intermittent connectivity problems with a duplex mismatch. When SwitchA sends a high volume of traffic to SwitchD, the Spanning Tree Protocol (STP) bridge protocol data units (BPDUs) from SwitchD could be lost due to collisions. If SwitchA does not receive those BPDUs from SwitchD, SwitchA will assume that there is a loss of connectivity to SwitchD, unblock port Fa0/2, and forward all packets to SwitchB, which could cause a switching loop.

Manually configuring Fa0/1 on SwitchA and SwitchD to use the same duplex mode would not cause a switching loop. Configuring both switch ports for full duplex mode would enable both ports to send and receive data simultaneously. Configuring both switch ports for halfduplex mode would enable only one port to send data at a time? however, communication could still occur, albeit slowly.

Configuring Fa0/1 on SwitchA for half duplex mode and configuring Fa0/1 on SwitchD toautonegotiate duplex settings would not cause a switching loop. Fa0/1 on SwitchD would not receive a duplex autonegotiation response from SwitchA, so SwitchD would default to halfduplex mode. Both switch ports would then be using the same duplex mode, thereby enabling communication between the two ports.

Configuring Fa0/1 on SwitchA and SwitchD to autonegotiate duplex settings would notcause a switching loop. If both sides of a link were configured to autonegotiate duplex settings, they would negotiate full duplex mode if both ports support full duplex operation.

If either side of the link does not support full duplex operation, the ports would negotiate half duplex mode.