[c-nsp] 3560 mtu miss-match causing output drops

[mb at adv.gcomm.com.au]

Quoting Peter Rathlev <peter at rathlev.dk>:

>
> Drops on account of MTU wouldn't show up as errors in your end.
Only way
> to see them was to look for "giant frames" on the other end.
>
> If your drop counter increases you have to look at why. Look at
"show
> interface Gi0/X counters errors". If they're OutDiscards, look at
"show
> platform port-asic stats drop Gi0/X". Maybe adjust the QoS model.
(Even
> with QoS disabled it uses queueing.)

  Yes - they are outdiscards:

  #sh interface gigabitEthernet 0/1 counters errors

Port        Align-Err     FCS-Err    Xmit-Err     Rcv-Err  UnderSize 
OutDiscards
Gi0/1               0           0           0           0         
0      3489015

Port      Single-Col  Multi-Col   Late-Col  Excess-Col 
Carri-Sen      Runts     Giants
Gi0/1              0          0          0           0         
0          0          0

  #sh platform port-asic stats drop gigabitEthernet 0/1

  Interface Gi0/1 TxQueue Drop Statistics
    Queue 0
      Weight 0 Frames 0
      Weight 1 Frames 0
      Weight 2 Frames 0
    Queue 1
      Weight 0 Frames 5387809
      Weight 1 Frames 251
      Weight 2 Frames 0
    Queue 2
      Weight 0 Frames 139
      Weight 1 Frames 0
      Weight 2 Frames 0
    Queue 3
      Weight 0 Frames 0
      Weight 1 Frames 0
      Weight 2 Frames 0

  The drops only occur when there is egress data - Minimal
traffic(i.e 10Mb/sec) I see counters increase
> If the ISP link is 100 Mbps, the other links are 1 Gbps and traffic
is
> primarily in their direction you might've run into the much
> discussed-about buffering on the 3560 platform.

  Yes - As stated above, it is only egress traffic that appears to
cause the drops - but we start to see the drops with only minimal
traffic running over the link?

  Thanks for your assistance.


-------------------------------------------------------------------------
This e-mail was sent via GCOMM WebMail http://www.gcomm.com.au/