The MPLS WG Archive

["chetan kumar s" <chetan.kumar@wipro.com>]

>
> hi!
>      IMHO this is not correct as the LSR from which the label mapping

Hi,
	If look at the Notes:7 given in A.1.2, things are quite clear. The
reason for including the message source is to handle the unsolicited
ordered control mode. If the label is not sent to the next hop(upstream)
for the FEC, the next hop may not have the label to use. Later if the next
hop becomes the downstream for the FEC in question and since we are in
unsoliciated ordered, the new next hop would have no way to request for
the label. Of course the received may wish to release the same based on
his retention mode, under which condition the next hop would be in
request-on-request so that it can request for label.

Thanks
Chetan S

> has beeen learnt can never be the upstream LSR for the same mapping -
> liberal or conservative mode notwithstanding.
>
> regards
> sumit
>
>
>
>
> "zze-PELLE Thierry StagiaireFTRD/DAC/ISS" <thierry.pelle@rd.francetelecom.fr> on
> 06/12/2001 12:02:39 PM
>
> To:   "'Manoj Dutta'" <manoj@ipinfusion.com>, mpls@UU.NET
> cc:    (bcc: Sumit Chauhan/HSS)
>
> Subject:  RE: Label Mapping !!
>
>
>
>
> Hi,
> > I had some doubts in RFC 3036. In section A.1.2, LMp.17 and A.1.6
> > FEC.1,  an LSR seems to send mappings for a given FEC to the next hop
> > for the same FEC.  If so, why ?
>
> As I understand it, this is due to the existance of liberal mode : in case
> of LDP reconstuction, the former Next Hop (downstrem located) may become a
> LSR upstream located.
>
> Rigth?
>
> Thierry
>
>
>
>