[j-nsp] Distributed PPM and LACP always goes into Queue 3. host-outbound-traffic knob has no effect -- bug?

[Masood Ahmad Shah]

Yeah host-outbound-traffic should change the distributed protocol handler
sourced traffic, however a firewall filter that uses the forwarding-class
and dscp actions to specify the override values on loopback will only
affect the RE sourced traffic and not the distributed protocol handler
sourced. Highly likely a bug then.


On Sat, May 16, 2015 at 4:44 PM, Huan Pham <drie.huanpham at gmail.com> wrote:

> Thanks Masood,
>
> This seems to be version specific and it is a bug on 11.4R7.5 (on MX5 I
> tested to be specific)
>
> On 12.3R8.7 I do not encounter this problem. The queue that (one hop) BFD
> is put on can be changed with "host-outbound-traffic" command (but still
> cannot be changed with lo0 firewall outbound filter).
>
> Thanks again,
>
> Huan
>
>
> On 15 May 2015, at 11:48 am, Masood Ahmad Shah <masoodnt10 at gmail.com>
> wrote:
>
> AFAIK host-outbound configuration or lo0 output filter will NOT influence
> the PFE generated traffic. Only the output interface filter can match the
> PFE generated traffic.
>
> Cheers,
> Masood
>
> On Fri, May 15, 2015 at 10:22 AM, Huan Pham <drie.huanpham at gmail.com>
> wrote:
>
>> Hi list,
>>
>> I've tested in the lab and confirm that distributed PPM (e.g. one hop BFD)
>> and LACP on MX does not honour host-outbound-traffic class of service nor
>> outbound RE-reclassificaiton filter. This traffic always gets into queue
>> 3.
>> Depending on your design, this behaviour could be a problem, especially if
>> your queue 3 is not designed for critical traffic.
>> Is is a bug? Is there any way to move this control traffic to a different
>> queue?
>>
>> Thanks very much in advance.
>>
>> Huan
>> _______________________________________________
>> juniper-nsp mailing list juniper-nsp at puck.nether.net
>> https://puck.nether.net/mailman/listinfo/juniper-nsp
>>
>
>