Hi Arashmid, 

Yes, it needs to be clarified. Here I think that in terms of GTP-U over IPv6 
case, gNB could be signaled a remote IPv6 endpoint address to which the SRGW 
binds a SID with appropriate payload type, such as IPv4/IPv6/Ethernet. In 
enhanced mode the SID doesn’t need to be allocated per session basis.

Wrt supporting unstructured PDU type, we may need to clarify which NH type to 
be used to carry over IPv6.

Cheers,
--satoru

> 2018/04/10 17:45、Arashmid Akhavain <arashmid.akhav...@huawei.com>のメール:
> 
> Hi Pablo,
>  
> We might need to add some clarification w.r.t state reduction as well and 
> explain this in the context of different modes.
>  
> Cheers,
> Arashmid
>  
> From: dmm [mailto:dmm-boun...@ietf.org] On Behalf Of Pablo Camarillo 
> (pcamaril)
> Sent: 10 April 2018 03:27
> To: Kentaro Ebisawa <ebike...@gmail.com>; dmm <dmm@ietf.org>
> Subject: Re: [DMM] Next Header in End.M.GTP6.D 
> (draft-ietf-dmm-srv6-mobile-uplane-01)
>  
> Hi Kentaro,
>  
> You are right. For each PDU session type you will have a different instance 
> of an End.M.GTP6.D SID. We can add a note in the next revision of the draft 
> in case this is not clear.
>  
> Thank you.
>  
> Cheers,
> Pablo.
>  
> From: dmm <dmm-boun...@ietf.org> on behalf of Kentaro Ebisawa 
> <ebike...@gmail.com>
> Date: Wednesday, 4 April 2018 at 11:08
> To: dmm <dmm@ietf.org>
> Subject: [DMM] Next Header in End.M.GTP6.D 
> (draft-ietf-dmm-srv6-mobile-uplane-01)
>  
> Hi,
>  
> In "6.2. End.M.GTP6.D", how would SR Gateway know if user packet inside GTP 
> is IPv4 or IPv6?
> I think this info is required to set newly added SRH.NextHeader field in step 
> 3~5.
>  
> My understanding is that information would be provided by the control plane, 
> and if that's correct, should SR Gateway have 2 (or more) SIDs corresponding 
> to the type of user packet and also inform gNB to use different SIDs 
> (dstAddr) based on user packet type?
> One could also have SID per TEID, but I'm not sure if that's a good design 
> choice considering the effort to reduce state in network nodes.
>  
> Thanks,
> -- 
> Kentaro Ebisawa <ebike...@gmail..com>
> _______________________________________________
> dmm mailing list
> dmm@ietf.org
> https://www.ietf.org/mailman/listinfo/dmm

_______________________________________________
dmm mailing list
dmm@ietf.org
https://www.ietf.org/mailman/listinfo/dmm

Reply via email to