[c-nsp] xconnect problem between C7606

Sigurbjörn Birkir Lárusson sigurbjornl at vodafone.is
Wed Oct 15 11:00:39 EDT 2014


I've had this issue with two LAN cards on a 7600 before with a
bog-standard EoMPLS circuit (with a slightly raised MTU from the default
1500).  In that case the microcode was "eating" (seemingly at random) some
MPLS packets from the customer who was purchasing the circuit, that caused
packet loss on the circuit and slow throughput.

It was solved by placing a service-policy on the interface on both sides
(ingress only) that did nothing (set dscp to dscp), somehow putting the
policy in place fixed things with no explanation provided.

Kind regards
Sibbi

On 15.10.2014 14:14, "Holger L" <cisco at entrap.de> wrote:

>On Wed, October 15, 2014 15:47, jure brkljacic wrote:
>> Is this service policies a dummy one,permiting everything?
>>
>> Can you be pleaae so kind and share the service policy configuration.
>
>policy-map IN-10MB
> class class-default
>  police cir 10000000
>policy-map OUT-10MB
> class class-default
>  shape average 11000000
>
>interface GigabitEthernet4/17
> service instance 98 ethernet 98
>  service-policy input IN-10MB
>  service-policy output OUT-10MB
>
>interface GigabitEthernet4/19
> service instance 98 ethernet 98
>  service-policy input IN-10MB
>  service-policy output OUT-10MB
>
>After configuring this policies IPv6 is working again. Anyway, I just had
>a webex with TAC regarding this issue and they think some of our network
>controllers might be broken. On the 76-ES+T-40G are 4 network controllers
>connected to 10 ports each. After moving our setup to the last controller,
>IPv6 was able to pass. Yours might be broken too..
>
>Bye,
>Holger
>
>_______________________________________________
>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/




More information about the cisco-nsp mailing list