I have a 3725 Cisco router connected via 2 FastEthernet interfaces to 2 Gigagbit interfaces on a 3750 switch stack. The 3725 is running c3725-advipservicesk9-mz.124-12.bin and the switch is running c3750-ipservicesk9-mz.122-53.SE1.bin, we are seeing incrementing input errors only ( no other errors ) on the 3725 router on both the FE interfaces and incrementing output drops on one of the Gigabitethernet interface on the 3750 switch stack.
Steps taken so far:
Hard coded both sides to full and speed 100
Checked the replaced the cabling between the switch and router
Checked the hardware on the switch , the asic is not dropping any packets
Put a hold queue 2000 out on the switch stack gigabit ethernet interface to buffer packets but that does not help, still getting output drops
It is highly unlikely that both FE interfaces on the 3725 router are bad at the same time.
Any suggestions would be appreciated at this time.
The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk
In no event shall Author be liable for any damages whatsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.
I'm wondering whether you might just be pushing too much traffic too quickly at your 3725. The 3725 is only rated at 100 to 120 Kpps. Two things you might try, set the links to 10 Mbps or, if supported and not enabled, enable router ingress flow-control. Either might help with the 3725's input errors.
For the 3750 switch's egress drops, here too using ingress flow control might throttle the sender.
If you have time-sensitive traffic, you might not want to use flow control. In that case, either you need to accept the drops as normal for oversubscription and/or attempt to identify the high rate sources and reduce their bandwith to the network.