The MPLS WG Archive

["Kullberg, Alan" <akullber@netplane.com>]

Ed,

I see the problem now.  The figure doesn't match the TERO.

The original TERO should be:
   {A(0,1),B(1,1),E(2,1),F(3,1),G(3,1),H(3,1),C(1,1),D(1,1)}
for tree
        A
        |
     +--+--+
     |  |  |
     B  C  D
     |
     E
     |
  +--+--+
  |  |  |
  F  G  H

as you stated in your original email and the modified TERO
used for pruning should be:
   {A(0,2),B(1,2),E(2,2),H(3,1),C(1,1),D(1,1)}
for (eventual) tree
        A
        |
     +--+--+
     |  |  |
     B  C  D
     |
     E
     |
     +--+
        |
        H


Thanks for pointing this out.

Alan

> -----Original Message-----
> From: Edward Harrison [mailto:eph@dataconnection.com]
> Sent: Tuesday, April 08, 2003 1:46 PM
> To: 'Kullberg, Alan'; Nippon - Seisho Yasukawa
> Cc: mpls@UU.NET
> Subject: RE: Clarification on draft-yasukawa-mpls-rsvp-p2mp-01.txt
> 
> 
> Thanks Alan,
> 
> I'm still a bit confused, though.
> 
> The message flow in figure 8 seems to show that node E is 
> subordinate to
> node B (as the signaling for the pruning flows from B to E 
> and not D to E).
> Moreover, I thought that as the nodes A, B and D all have the 
> subtree-id of
> 2 in the TERO, they were part of the same sub-tree.  If E is 
> subordinate to
> D, shouldn't D have a subtree-ID of 2 and B have a subtree-id of 1?
> 
> Answering my confusion another way, maybe you could clarify 
> what the tree
> looks like for the modified TERO in example 6.1
> {A(0,2),B(1,2),C(1,1),D(1,1),E(2,2),F(3,2),G(3,2)}?
> 
> Does the new
> 
>      E
>      |
>    +---+
>    |   |
>    F   G
> 
> subtree hang from node D (as this is where it appears in the 
> TERO) or node B
> (as this is the node with the same subtree ID and is where 
> the signaling
> flow seems to go in figure 6)?
> 
> Thanks,
> 
> Ed
> 
> -----Original Message-----
> From: Kullberg, Alan [mailto:akullber@netplane.com]
> Sent: 08 April 2003 18:34
> To: Edward Harrison; Nippon - Seisho Yasukawa
> Cc: mpls@UU.NET
> Subject: RE: Clarification on draft-yasukawa-mpls-rsvp-p2mp-01.txt
> 
> 
> Ed,
> 
> I believe that the TEROs as shown in the document are correct.
> In the modified TERO {A(0,2),B(1,2),C(1,1),D(1,1),E(2,2),H(3,1)},
> notice that E is at level 2 and is subordinate to D at level 1
> since D is the most recent node at level 1 encountered when
> parsing the TERO from left to right.
> 
> Hope this helps.
> 
> Alan
> 
> > -----Original Message-----
> > From: Edward Harrison [mailto:eph@dataconnection.com]
> > Sent: Tuesday, April 08, 2003 11:20 AM
> > To: Nippon - Seisho Yasukawa
> > Cc: mpls@UU.NET
> > Subject: Clarification on draft-yasukawa-mpls-rsvp-p2mp-01.txt
> > 
> > 
> > Hi,
> > 
> > I have one small question on draft-yasukawa-mpls-rsvp-p2mp-01.txt.
> > 
> > Can you confirm whether the Original and Modified TEROs in 
> section 6.3
> > (sender initiated pruning) are correct?
> > 
> > From section 4.6.1.1, I would expect the original TERO
> > {A(0,1),B(1,1),C(1,1),D(1,1),E(2,1),F(3,1),G(3,1),H(3,1)} to 
> > correspond to
> > the following tree:
> > 
> >        A
> >        |
> >     +--+--+
> >     |  |  |
> >     B  C  D
> >           |
> >           E
> >           |
> >        +--+--+
> >        |  |  |
> >        F  G  H
> > 
> > However, from the example in 6.1, I would expect the modified TERO
> > {A(0,2),B(1,2),C(1,1),D(1,1),E(2,2),H(3,1)} to correspond to:
> > 
> >        A
> >        |
> >     +--+--+
> >     |  |  |
> >     B  C  D
> >     |
> >     E
> >     |
> >     H
> > 
> > Have I missed something, or should the original TERO for 
> this example
> > actually be 
> {A(0,1),B(1,1),E(2,1),F(3,1),G(3,1),H(3,1),C(1,1),D(1,1)},
> > corresponding to:
> > 
> >        A
> >        |
> >     +--+--+
> >     |  |  |
> >     B  C  D
> >     |
> >     E
> >     |
> >  +--+--+
> >  |  |  |
> >  F  G  H
> > 
> > Regards,
> > 
> > Ed
> > 
>