[c-nsp] non-existing input errors on 6500/SXI...?

[Gert Doering]

Hi,

On Fri, Oct 21, 2011 at 02:17:32PM -0400, John Gill wrote:
> Are the errors incrementing, or are they going up/down over time?  

Slowly increasing, never going down.

> There are a couple of bugs with counter outputs where they are 
> read/initialized incorrectly.

Oh?  Never been hit by that one, but indeed, that would be exciting :-)

> Can you look at "show int counters trunk" for wrong encap?  

Cisco-M#sh int g1/9 count trunk

Port            TrunkFramesTx  TrunkFramesRx  WrongEncap
Gi1/9             65779332406    49937765599           0

> The 2960G 
> uses DTP by default, so it may just be those frames hitting the 
> interface.  

I assumed something like that, but if I understand the 6500 right,
it also has DTP on-by-default:

Cisco-M#sh int g1/9 acc
GigabitEthernet1/9 SW: sp1/switch6:g0/14 (sp1)
                Protocol    Pkts In   Chars In   Pkts Out  Chars Out
                   Other          0          0     323191   29733572
           Spanning Tree      11113     711232  261042841 25060112736
                     CDP     161772  101431044     179775   84853449
                     DTP     323501   19410060          0          0

mmmh.  Now that's funny, sending out "other" packets but receiving
DTP just fine...  but yeah, that's a symmetric counter bug - the 
2960 has the same weirdness...

Switch6#sh int g0/14 acc
GigabitEthernet0/14 SW: sp1/cisco-m:gi1/9 (sp1)
                Protocol    Pkts In   Chars In   Pkts Out  Chars Out
                   Other          0          0     300953   18057180
           Spanning Tree   39125257 2504016448       2113     135232
                     CDP      26657   11195940      23994   15044238
                     DTP      47934    2876040          0          0

(so where's the 47000 DTP packets coming from if the other end
never sent a single one?)

Since both sides are set to "switchport mode trunk" unconditionally,
I'll disable DTP on both sides (switchport noneg) and see whether it 
changes anything...

... some 5 later: no, didn't fix anything:
Cisco-M#sh int g1/9 | inc err
     33 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored
     304072 packets output, 142939079 bytes, 0 underruns

> Switchport trunk encapsulation on the 2960G should be set to 
> be dot1q.

My 2960G has no other encapsulations anyway...  

Switch6(config-if)#swi trunk ?
  allowed  Set allowed VLAN characteristics when interface is in trunking mode
  native   Set trunking native characteristics when interface is in trunking
           mode
  pruning  Set pruning VLAN characteristics when interface is in trunking mode


> Also, check L3 interfaces (vlan interfaces or no switchport interfaces) 
> for anything regarding input queue drops:

This box has been up since 36 weeks, so quite a number of input flushes
have accumulated over time.  I've cleared all counters, and all
"Input" (and "input error") counters are still zero, while the 
"input errors" on gi1/9 are again at 50...


> sh int | inc is up|Input
> 
> show queueing interface gi1/9 could also show queuing drops on input.

Nothing there:

  Packets dropped on Receive:
    BPDU packets:  0
    que           dropped        30-s bytes        peak bytes    5-mins avg bps          peak bps  [cos-map]
    ----------------------------------------------------------------------------------------------------
    1                   0                 0                 0                 0                 0   [0 1 2 3 4 5 6 7 ]


there errors are hiding really well... :-)

gert
-- 
USENET is *not* the non-clickable part of WWW!
                                                           //www.muc.de/~gert/
Gert Doering - Munich, Germany                             gert at greenie.muc.de
fax: +49-89-35655025                        gert at net.informatik.tu-muenchen.de
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 305 bytes
Desc: not available
URL: <https://puck.nether.net/pipermail/cisco-nsp/attachments/20111021/b556dc23/attachment-0001.pgp>