[nsp] Wrong interface counters on 2948G-L3?

From: sthaug@nethelp.no
Date: Thu Jan 04 2001 - 09:55:32 EST


We have a Cat 2948G-L3 connected via a Gigabit link (LX GBIC) to a 3524XL.
The interface counters on the two Gigabit interfaces that are connected,
are inconsistent. E.g. after having reset the counters at about the same
time:

On the 3524XL:

GigabitEthernet0/2 is up, line protocol is up
  Hardware is Gigabit Ethernet, address is 0004.271b.c51a (bia 0004.271b.c51a)
  MTU 1500 bytes, BW 1000000 Kbit, DLY 10 usec,
     reliability 255/255, txload 1/255, rxload 1/255
  Encapsulation ARPA, loopback not set
  Keepalive not set
  Auto-duplex (Full), link type is autonegotiation, media type is LX
  output flow-control is on, input flow-control is off
  ARP type: ARPA, ARP Timeout 04:00:00
  Last input 00:00:00, output 00:00:00, output hang never
  Last clearing of "show interface" counters 00:04:56
  Queueing strategy: fifo
  Output queue 0/40, 0 drops; input queue 0/75, 0 drops
  30 second input rate 2280000 bits/sec, 351 packets/sec
  30 second output rate 1048000 bits/sec, 292 packets/sec
     114033 packets input, 92864753 bytes, 0 no buffer
     Received 330 broadcasts, 0 runts, 0 giants, 0 throttles
     0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored
     0 watchdog, 330 multicast, 0 pause input
     97533 packets output, 46148470 bytes, 0 underruns
     0 output errors, 0 collisions, 0 interface resets
     0 babbles, 0 late collision, 0 deferred
     0 lost carrier, 0 no carrier, 0 pause output
     0 output buffer failures, 0 output buffers swapped out

On the 2948G-L3:

GigabitEthernet49 is up, line protocol is up
  Hardware is xpif_port, address is 0002.bad7.6907 (bia 0002.bad7.6907)
  MTU 1500 bytes, BW 1000000 Kbit, DLY 10 usec, rely 177/255, load 1/255
  Encapsulation ARPA, loopback not set, keepalive set (10 sec)
  Full-duplex, 1000Mb/s, 1000Base-LX/LH, Auto-negotiation
  ARP type: ARPA, ARP Timeout 04:00:00
  Last input 00:00:00, output 00:00:00, output hang never
  Last clearing of "show interface" counters 00:04:56
  Queueing strategy: fifo
  Output queue 0/40, 0 drops; input queue 0/75, 0 drops
  30 second input rate 1064000 bits/sec, 226 packets/sec
  30 second output rate 2215000 bits/sec, 208 packets/sec
     75179 packets input, 46122114 bytes, 0 no buffer
     Received 0 broadcasts, 0 runts, 0 giants, 0 throttles
     0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored, 0 abort
     0 watchdog, 97436 multicast
     0 input packets with dribble condition detected
     69551 packets output, 92910728 bytes, 0 underruns
     0 1-collision pkt cntrs, 0 more-collision pkt cntrs,
     0 excessive-collision pkt cntrs
     44496 output errors, 0 collisions, 0 interface resets
     0 babbles, 0 late collision, 0 deferred
     0 lost carrier, 0 no carrier
     0 output buffer failures, 0 output buffers swapped out

Particularly interesting are the following counters (lines with only
zero counters removed):

     114033 packets input, 92864753 bytes, 0 no buffer
     Received 330 broadcasts, 0 runts, 0 giants, 0 throttles
     0 watchdog, 330 multicast, 0 pause input
     97533 packets output, 46148470 bytes, 0 underruns
vs.
     75179 packets input, 46122114 bytes, 0 no buffer
     0 watchdog, 97436 multicast
     69551 packets output, 92910728 bytes, 0 underruns
     44496 output errors, 0 collisions, 0 interface resets

Notice that 69551 packets output + 44496 output errors on the 2948G-L3
is about the same (114047) as 114033 packets input on the 3524XL. The
3524XL shows *no* input errors.

Similarly, 97533 packets output on the 3524XL is about the same as
the 97436 multicasts received on the 2948G-L3 (and the 2948G-L3
consistently shows more multicast packets input than total packets
:-)

I believe the interface counters on the 2948G-L3 are simply wrong, and
this link is fine. Anybody else seen similar problems with the counters
on the 2948G-L3?

Steinar Haug, Nethelp consulting, sthaug@nethelp.no



This archive was generated by hypermail 2b29 : Sun Aug 04 2002 - 04:12:24 EDT