Talk:Weighted round robin

Unnecessary transformation
there is unnecessary transformation going on. mathematically this for each flow f   f.normalized_weight = f.weight / f.mean_packet_size min = findSmallestNormalizedWeight for each flow f   f.packets_to_be_served = f.normalized_weight / min is equal to min=find smallest waight for each flow f (f.weight/f.mean_packet_size)/ (f.weight(0)/f.mean_packet_size) = f.weight/f.weight(0) is there any reason to to do n times more divisions ? if there are some rounding problems it should be mentionned on main page. thanks. — Preceding unsigned comment added by Xchmelmilos (talk • contribs) 10:30, 4 May 2008‎ (UTC)

The mean packet size can differ from one flow to another.

f.packets_to_be_served = f.normalized_weight / min is equal to f.packets_to_be_served = (f.weight / f.mean_packet_size) / (min_f.weight / min_f.mean_packet_size)

In fact the weakness section of this article mentions that having to know the mean packet size is a problem with this algorithm - having the mean packet size cancel out doesn't make sense. --195.37.186.61 (talk) 23:13, 22 August 2012 (UTC)

Related links
The following links might be helpful for improving the article but are inappropriate to be included directly: --EnOreg (talk) 14:09, 23 December 2011 (UTC)
 * A simple explanation of weight round-robin distribution
 * one more explanation