[c-nsp] QoS Bandwidth percent vs bandwidth remaining percent

[Kenny Sallee]

So - I've research the difference between the 'bandwidth percent' and
'bandwidth remaining percent' commands with regards to configuring a
policy-map on a Cisco router.  There are some good links to folks who have
the theory behind each command:

Cisco:
http://www.cisco.com/en/US/tech/tk543/tk757/technologies_tech_note09186a0080103eae.shtml
Other Web stuff:
http://ardenpackeer.com/qos-voip/tutorial-what-is-the-difference-between-bandwidth-percent-and-bandwidth-remaining-percent/


And a couple others.  What I gather as far as difference is that with
'bandwidth percent' you allocate % of bandwidth based on the entire real
(physical interface) or configured (logical interface) bandwidth.  So the
configured bandwidth %'s must all equal 100%.  With 'bandwidth remaining
percent' you allocate your % for each class map based on what's left over
after a class is configured with the 'priority <bw> or <%>' command.  So the
'priority' command could reserve 90% of an interface - then the remaining BW
you can split that to equal 100% of the remaining bandwidth between all the
other classes you have configured.  I also get that max-reserved-bandwidth
can be manipulated to change what the real % you can allocate is...

So from this - the one advantage I can see to 'bandwidth remaining percent'
is that you can get more granular with how you allocate traffic to each
class.  Outside of that - I'm not really sure / clear on the operational
advantage of one over the other?  Any thoughts or opinions?

Thanks,
Kenny