Re: I-D Action: draft-voyer-6man-extension-header-insertion-07.txt

[Gyan Mishra <[e-mail]>]

In-line 

Sent from my iPhone

> On Oct 17, 2019, at 12:19 PM, Tom Herbert <[e-mail]> wrote:
> 
> 
> 
>> On Thu, Oct 17, 2019, 6:20 AM Alexandre Petrescu <[e-mail]> wrote:
>> 
>> 
>> Le 15/10/2019 à 00:12, Fernando Gont a écrit :
>> > On 12/10/19 16:57, Brian E Carpenter wrote:
>> >> Hi,
>> >> 
>> >> I'd like to comment on this version. It is in fact a complete
>> >> rewrite compared to its predecessors and I thank the authors for
>> >> that. The tone is now purely technical, and that's a great
>> >> improvement.
>> > 
>> > It is somewhat frustrating that the draft still fails to argue why
>> > EH insertion instead of encapsulation.
>> 
>> As usual, I think one of the reasons is a difficulty in qualifying what
>> it means 'encapsulation'.
>> 
>> There is IP-in-IP encapsulation.
>> 
>> But there is also encapsulation like in transporting, or carrying, by
>> means of other intermediary headers, layer2, MPLS, security headers and
>> future internet shims and GRE and routing headers.
>> 
>> IP-in-IP encapsulation is clearly an alternative to EH insertion.
>> 
>> But all the other encapsulations are so messy that one may legitimately
>> think that a new EH insert/delete standardized according to good WG
>> principles would be proper, universal, and solve all  problems of GRE
>> for example.
> 
> 
> Alex,
> 
> What are the problems of GRE to which you referring?
> 
> Tom
> 
> -[Gyan] Good old protocol 47.  GRE is the Swiss Army knife of every network designers toolbox that I have used countless times to solve routing issues where traditional native routing was not possible and GRE tunneling was the savior.  Hats off to GRE for being with us all these years.  The main caveat In my experience have seen with GRE is accounting for the GRE/IP 20-IP 4-GRE = 24 overhead bytes and having to adjust the MTU to accommodate bumping the MTU along the path or lowering the GRE tunnel MTU to prevent fragmentation which most vendor platforms punt to software path to the RP impacting CPU utilization.

> GRE can be very handy from a routing perspective to solve routing issues and can also be used in transition or migration strategy or to fix an issue or allow routing islands to communicate such as with transition technology tunnels. The big down side with GRE  example is using IPv6 transition technology 6to4 tunnels to provide routing reachability over between IPv6 islands over an IPv4 backbone which works for small use cases but can become unmanageable as the tunnel grow to any to any for very large use cases not practical.


> There are a number of protocol specification which adopted GRE as the transport mechanism such as Net to Net IPSEC VPN DMVPN MGRE VTI tunnel mode is handing for running a  routing protocol over the VTI.  Another good protocol specification using GRE is MVPN Rosen MT tunnel P-instance default and Data MDT PIM enabled mpls core for inclusive default MDT      I-PMSI or Selective Data MDT S-PMSI. GRE has also been used for vxlan evpn NVGRE tunnel to anycast VTEP for VNIs.  


> The main reason for using GRE encapsulation which requires termination endpoint is if you are going to run a routing instance over that tunnel.

> If not requiring a routing instance a similar analogy is IPEC VPN transport mode is used if you don’t need a routing instance and if you do then you use tunnel mode and have a VTI for the routing instance.  


> With SRv6 just as with mpls an encapsulation is what we are adding similar to IPSEC transport mode since we don’t need a routing instance there is not need to have to terminate the A and Z end and you can then add as many encapsulations as you want similar to an mpls label stack.  I think it does add complexity since you have to terminate the endpoints but also now to support Ti-LFA backup paths additional encapsulation for the new EH insertion for the SRH header routing type 4 I think it’s better to just do the 6in6 encapsulation and not add GRE 4 byte header protocol 47.  


> I read the link but I did not see where it stated you can stack GRE tunnels. 

https://tools.ietf.org/html/draft-ietf-intarea-tunnels-10
>> 
>> Alex
>> 
>> > 
>> > 
>> 
>> --------------------------------------------------------------------
>> IETF IPv6 working group mailing list
>> [e-mail]
>> Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
>> --------------------------------------------------------------------
> --------------------------------------------------------------------
> IETF IPv6 working group mailing list
> [e-mail]
> Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
> --------------------------------------------------------------------