On Wed, Nov 30, 2016 at 6:24 AM, László Sürü <[email protected]>
wrote:

> Hi Manuel,
>
> Yes, and the default destination port for a VXLAN-GPE tunnel should be
> 4790 then.
>
Does this matter yet? Meaning, is any of the gpe stuff really in use yet?
Typically we just use the vxlan-gpe vtep as a standard vxlan vtep. The 6633
port was just a way to have a different tunnel. Or the actual port number
really didn't matter.

>
> Best regards
> Laszlo
>
> -----Original Message-----
> From: Manuel Buil
> Sent: Tuesday, November 29, 2016 3:02 PM
> To: László Sürü <[email protected]>; Jiri Benc <[email protected]>;
> Yang, Yi Y <[email protected]>
> Cc: Jan Scheurich <[email protected]>;
> [email protected]; Georgios Paraskevopoulos (
> [email protected]) <[email protected]>; Georg
> Schmuecking <[email protected]>
> Subject: RE: Potential bug in NSH patch for OVS
>
> Hi Laszlo,
>
> It should use destination port 4789 which is the one defined for the
> default connectivity among computes. However, if the packet was ever in the
> SFC pipeline, it uses 6633.
>
> Regards,
> Manuel
>
> -----Original Message-----
> From: László Sürü
> Sent: Monday, November 28, 2016 5:54 PM
> To: Manuel Buil <[email protected]>; Jiri Benc <[email protected]>;
> Yang, Yi Y <[email protected]>
> Cc: Jan Scheurich <[email protected]>;
> [email protected]; Georgios Paraskevopoulos (
> [email protected]) <[email protected]>; Georg
> Schmuecking <[email protected]>
> Subject: RE: Potential bug in NSH patch for OVS
>
> Hi All,
> Sorry for the late reply, just catching up with this discussion.
>
> Regarding the tp_dst port, I've noted the above described misbehavior
> before, that is Instead of the IETF reserved 4790 destination port for
> VXLAN-GPE OpenStack Tacker  - if used for NFVI use case - uses the  hard
> coded 6633 destination port.
>
> Manuel, can this be the case here?
>
> Best regards
> Laszlo
>
> -----Original Message-----
> From: Manuel Buil
> Sent: Thursday, November 24, 2016 6:01 PM
> To: Jiri Benc <[email protected]>; Yang, Yi Y <[email protected]>
> Cc: Jan Scheurich <[email protected]>;
> [email protected]; Georgios Paraskevopoulos (
> [email protected]) <[email protected]>; Georg
> Schmuecking <[email protected]>; László Sürü <
> [email protected]>
> Subject: RE: Potential bug in NSH patch for OVS
>
> Hi Jiri,
>
> I think ovs pushes that metadata because the packet ingresses ovs through
> a vxlan-gpe port with exactly those parameters.
>
> -Manuel
>
> -----Original Message-----
> From: Jiri Benc [mailto:[email protected]]
> Sent: Thursday, November 24, 2016 3:24 PM
> To: Yang, Yi Y <[email protected]>
> Cc: Jan Scheurich <[email protected]>; Manuel Buil <
> [email protected]>; [email protected]; Georgios
> Paraskevopoulos ([email protected]) <[email protected]>;
> Georg Schmuecking <[email protected]>; László Sürü <
> [email protected]>
> Subject: Re: Potential bug in NSH patch for OVS
>
> On Thu, 24 Nov 2016 13:34:29 +0000, Yang, Yi Y wrote:
> > Vxlan implementation doesn't provide a field to change it, do you know
> > how we can change it? I search ovs source code, there isn't a field to
> > change tp_dst in tunnel header.
>
> Hmm, seems you're right.
>
> So, first thing that you need to figure out is why ovs is pushing
> "actions:set(tunnel(tp_src=49676,tp_dst=6633))" to the kernel when
> outputting to a tunnel that has a different port configured. Because this
> is the problem, not the kernel code.
>
> Btw, I just tried it and ovs doesn't do that for me. But I have most
> certainly a different configuration than you do.
>
>  Jiri
>
>
>
> _______________________________________________
> sfc-dev mailing list
> [email protected]
> https://lists.opendaylight.org/mailman/listinfo/sfc-dev
>
_______________________________________________
sfc-dev mailing list
[email protected]
https://lists.opendaylight.org/mailman/listinfo/sfc-dev

Reply via email to