[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [manet] DYMO routing information's usefulness
I do have some problems with the 'superior check' aswell.
Superior routes:
The first part checks if the new information has a higher sequence number. If so, it is superior.
In the second part the hop count is checked the same way for RREQs and RREPs, though within the inferior check they are handled different. Why they do differ in the Route.Broken flag I do not understand.
The third part is wrong. The sequence number is not checked, so it's possible an information with lower sequence number is marked superior.
If you put another pair of brackets to check if the sequence number are equal, it's still wrong. If the route is not broken, and the new route has the same hop count, it is inferior in case of a RREQ!
Maybe in the third part it should be Route.Broken == true.
For now I will use the negated form of stale, loop-possible and inferior information to determine if it's superior. It's longer, but I think it's correct.
regards,
Henner Jakob
On Mon, Feb 26, 2007 at 08:09:09AM -0800, Ian Chakeres wrote:
> Excellent catch. I'll correct the pseudo code to match the text.
>
> Ian Chakeres
>
> On 2/26/07, Romain Thouvenin <romain.thouvenin at gmail.com> wrote:
> >Hi,
> >
> >You'll probably show me that I am wrong, but it seems that there is a
> >mistake in DYMO specs, sections 5.2.1.3 and 5.2.1.4.
> >
> >In the text of 5.2.1.3, the information is supposed to be inferior if
> >Node.HopCnt >= Route.HopCnt (for a non-broken route during a RREQ),
> >but according to the pseudo-code test, it is when Node.HopCnt >
> >Route.HopCnt.
> >
> >The completeness test in 5.2.1.4 agrees with the test of 5.2.1.3, and
> >thus disagrees with the text. But I would have thought that equivalent
> >information must be discarded during a RREQ because even if it
> >loop-free, the information will be forwarded twice by each node on the
> >rest of the path toward the target node. This seems to be confirmed on
> >a small example. Did I miss something ?
> >
> >Thanks for your clarifications,
> >Romain
> >
> >_______________________________________________
> >manet mailing list
> >manet at ietf.org
> >https://www1.ietf.org/mailman/listinfo/manet
> >
>
> _______________________________________________
> manet mailing list
> manet at ietf.org
> https://www1.ietf.org/mailman/listinfo/manet
--
_______________________________________________
manet mailing list
manet at ietf.org
https://www1.ietf.org/mailman/listinfo/manet