[email protected]
[Top] [All Lists]

Re: composite link - candidate for respin, maybe

Subject: Re: composite link - candidate for respin, maybe
From: Greg Mirsky
Date: Mon, 29 Mar 2010 11:55:31 -0700
Dear Tony,
a link might have B/W per TC. Thus encoding
  • composite links
bandwidth 10, 40, 100
will become
  • composite links
    • P0 a, b, c
    • P1 ...
That will same some encoding but I wonder if such saving is real motivation for this work.

Regards,
Greg

On Mon, Mar 29, 2010 at 10:59 AM, Tony Li <tony.li@tony.li> wrote:

> Then why bother with link bundling?  This level of control has been part of TE
> since the beginning.  Link bundling was introduced to trade information hiding
> for scalability.


Because we can gain a great deal of scalability by advertising vectors of
metrics rather than explicit metrics per link.  So, rather than:

   link 1
       bandwidth 10gbps
   link 2
       bandwidth 40gbps
   link 3
       bandwidth 100gbps

We have:

   composite links
       bandwidth 10, 40, 100

This saves an enormous amount of overhead in TLV encoding.  If we can
further only vectorize a subset of the metrics for the composite link (e.g.,
IGP metric and link coloring might be common), we get further scaling
advantages.  And on top of that, little or no loss of useful information.
Perfect, thanks.


Tony


_______________________________________________
rtgwg mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/rtgwg

_______________________________________________
rtgwg mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/rtgwg
<Prev in Thread] Current Thread [Next in Thread>