Re: [Roll] "Link quality" as a metric for MRHIF

On Jun 2, 2012, at 4:51 AM, Omprakash Gnawali <gnawali@cs.uh.edu> wrote:

> On Fri, Jun 1, 2012 at 3:14 PM, Ralph Droms <rdroms@cisco.com> wrote:
>> How is the "link quality" metric used as a selected metric in MRHOF?  As I read RFC 6551, the link quality container carries a list of individual link qualities.  To use the link quality metric in MRHOF, does the computing node add up all the link qualities from the container option to compute the parent path cost?  What about link quality sub-objects with link quality value "0"?
> 
> Yes, adding up all the link quality level to compute the path cost.
> Adding up all the link quality level, including a 0, will result in
> some value which is converted to rank but it is not clear if MRHOF
> would be the best objective function to use if you want to use link
> quality level metric.

Ok, then the MRHOF spec should be revised to drop link quality level as one of the candidate selected metrics or to specify in detail how to use link quality level as the selected metric.

- Ralph

> 
> - om_p
> _______________________________________________
> Roll mailing list
> Roll@ietf.org
> https://www.ietf.org/mailman/listinfo/roll