The MPLS WG Archive

[Eric Gray <ewgray@mindspring.com>]

Bill,

    Yangguang is exactly correct.  What the L bit
means is that the next listed hop in the ER does not
have to be an adjacent peer.  It still must be in the
eventual ER.  Some people have been known to
refer to an "abstract node" as a "loose node", and
- in that sense - the L bit indicates a "loose hop"
rather than a "loose node".

--
Eric Gray

Yangguang Xu wrote:

> > > The "L" bit for a node doesn't mean the node is loose.
> >
> > That is exactly what it means.
> >
> > > Indeed, the node is
> > > strict but the path from preceding node to this node is
> > > loose. Am I right?
> > >
>
> > If you have a "loose prior abstract node," the address *may* be the previous
> > hop in the ERO list.  If you have a "strict prior abstract node," the
> > address *must* be the previous hop in the ERO list.
> >
> > Bill
>
> Bill,
>
> Because node marked as "loose" can also be interpreted as that the node itself
> is not necessary in the ERO. What I am saying is that no matter the abstract
> node is marked as loose or strict, it has to be in the ERO, however somewhere
> depending on its being loose or strict.
>
> Thanks,
>
> Yangguang