Recently I’ve setup a monitoring tool called Intermapper to monitor the Nortel switches I have at work. The tool polls the SNMP traffic information from the switches, and also picks up some basic error notifications received on it’s ports.
On the trunk ports that are being uplinked to the core, I noticed that SNMP has been reporting packet discards, more precisely “RX Discards” = XX/min. As I have etherchannel setup with two uplinks, both uplinks are displaying the same packet discard behavior. I know this is common on lower end Cisco Catalyst switches, when the interfaces experiences “microbursts” due to small shared buffers. However, this isn’t the case as it is constantly reporting large amount of packet discards on only trunk interfaces and regardless on how much traffic is traversing the ports.
After doing some research, I found out that this is a case of “VLAN configuration mismatch” on one end of the trunk. The receiving end of the Nortel switch is simply discarding tagged packets that are ending up on the switch which it is not configured for, hence explaining why both etherchannel trunk interfaces are displaying this error message.
Once I have removed the VLAN that don’t belong on the trunk from the sender switch, the error message immediately disappeared.