[c-nsp] Input (overrun) errors on PA-GE / 7204VXR

[Graham Wooden]

Hi Pete,

I am not pushing line rate on the card; it's maybe doing 20Mb at any given
time.
I think my problem is the type of traffic; it's mostly SIP/RTP traffic and
I am afraid   those small packets is what is causing the issues.  I went
ahead and tweaked the buffers and cleared the counters.  I am going to let
it run some time and review again.

I maybe looking at a NPE-G1 upgrade Š

-graham

On 5/19/12 11:38 PM, "Pete Lumbis" <alumbis at gmail.com> wrote:

>I'm not 100% sure, but I don't think the PA-GE can actually handle
>line rate. This is the only thing I found to back that up,
>https://supportforums.cisco.com/docs/DOC-2613
>
>
>
>On Sat, May 19, 2012 at 4:51 PM, Graham Wooden <graham at g-rock.net> wrote:
>> Hi Folks,
>>
>> It seems that I am overloading the traffic ingress into a PA-GE
>>interface on
>> one of my internet edge routers,  a 7204VXR
>> (NPE-400/512MB/advipservicesk9-mz.122-33.SRE4) without too much effort.
>>
>> 30 second counters shows the traffic not to be too bursty (as well as
>>the 5
>> minute graphs), especially when the traffic leaving the other side are
>>the
>> two FE interfaces (upstreams) on the I/O controller. Combined on those
>>two
>> are maybe do about 20Mb each at different times.
>>
>> Anything else I can look at, besides writing this off to be an issue of
>>"too
>> much traffic inbound"? A NPE-400 with the PA-GE can't handle this low
>>volume
>> of traffic?  Or do need to look at the physical layer first (which
>>seems to
>> be OK).
>>
>> The other end of the 1000LX is a C6509 (SUP720-3b), with the interface
>> sitting on a WS-X6516-GBIC.  That interface shows zero issues, and has
>>input
>> flow-controll off (output is on).
>>
>> I am trying not to spend the money on a NPE-G1, which seems way overkill
>> because the level of traffic beside so little.  Here is some output
>>from the
>> 7204VXR.
>>
>> EBR1-ATL1#show running-config interface gi2/0
>> Building configuration...
>>
>> Current configuration : 208 bytes
>> !
>> interface GigabitEthernet2/0
>>  description CR1-ATL1
>>  ip address nn.nn.nn.10 255.255.255.248
>>  ip flow ingress
>>  negotiation auto
>>  service-policy output SetDSCPVOIP
>>  hold-queue 768 in
>>  hold-queue 512 out
>> end
>>
>> EBR1-ATL1#show interfaces gi2/0
>> GigabitEthernet2/0 is up, line protocol is up
>>  Hardware is WISEMAN, address is 0007.4f08.f438 (bia 0007.4f08.f438)
>>  Description: CR1-ATL1
>>  Internet address is nn.nn.nn.10/29
>>  MTU 1500 bytes, BW 1000000 Kbit/sec, DLY 10 usec,
>>     reliability 255/255, txload 1/255, rxload 1/255
>>  Encapsulation ARPA, loopback not set
>>  Keepalive set (10 sec)
>>  Full Duplex, 1000Mbps, link type is auto, media type is LX
>>  output flow-control is XON, input flow-control is XON
>>  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 4w2d
>>  Input queue: 1/768/0/0 (size/max/drops/flushes); Total output drops: 0
>>  Queueing strategy: fifo
>>  Output queue: 0/512 (size/max)
>>  5 minute input rate 490000 bits/sec, 363 packets/sec
>>  5 minute output rate 3180000 bits/sec, 447 packets/sec
>>     1664722579 packets input, 2698235750 bytes, 0 no buffer
>>     Received 0 broadcasts (0 IP multicasts)
>>     0 runts, 0 giants, 0 throttles
>>     47686 input errors, 0 CRC, 0 frame, 47686 overrun, 0 ignored
>>     0 watchdog, 2086826 multicast, 0 pause input
>>     2081482879 packets output, 3723382484 bytes, 0 underruns
>>     0 output errors, 0 collisions, 0 interface resets
>>     0 unknown protocol drops
>>     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
>> EBR1-ATL1#
>>
>>
>> _______________________________________________
>> cisco-nsp mailing list  cisco-nsp at puck.nether.net
>> https://puck.nether.net/mailman/listinfo/cisco-nsp
>> archive at http://puck.nether.net/pipermail/cisco-nsp/