Fred,
In Section 3.2, we put implementation details of the probing mechanism out of
scope.
Regardless of MTU and fragmentation issues, a robust GRE implementation
requires a probing mechanism. Without such a probing mechanism, a GRE ingress
node might activate a GRE tunnel when the egress was not reachable at all. This
would result in black-holing.
Since RFC 2784 doesn't specify a probing mechanism (or even require one) most
vendors have developed proprietary probing mechanisms. These proprietary
mechanism must deal with the problems of LAG and ECMP. However, because the
probing mechanisms aren't IPv6 specific, and because they are proprietary, we
put them out of scope.
Ron
> ------------------------------
>
> Message: 2
> Date: Tue, 14 Apr 2015 17:12:53 +0000
> From: "Templin, Fred L" <[email protected]>
> To: "[email protected]" <[email protected]>
> Subject: Re: [Int-area] I-D Action: draft-ietf-intarea-gre-ipv6-07.txt
> Message-ID:
> <2134F8430051B64F815C691A62D9831832E495AD@XCH-BLV-
> 504.nw.nos.boeing.com>
>
> Content-Type: text/plain; charset="us-ascii"
>
> I will say also that the act of probing in general is a bit troublesome. In
> particular, when there is Equal Cost MultiPath (ECMP) how can you tell that
> the probe packets will follow the same path as the data packets?
>
> Thanks - Fred
> [email protected]
>
[RPB]
***
_______________________________________________
Int-area mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/int-area
