From: Thomas Perrier (thomas@perrier.name)
Date: Fri Oct 12 2007 - 08:06:26 ART
> What I could also see is some sort of QoS mechanism where somebody who just
> keeps booking the exam over and over again would get lower and lower
> priority on the queue in order to free up "bandwidth" for others. In this
> case, the 'bandwidth' would be seats. We have to have a system that
> discourages users from hogging bandwidth (seats). For example, somebody
> who is booking their 21st attempt might have to be made to give up their
> seats to those who are booking their very first attempt. Again, the idea is
> that if you've already had 20 shots at it, you should be made to give other
> people a chance. You can't just keeping taking shots over and over again
> and thereby hindering others from having their chance.
Maybe a nice idea, but as already hinted by Scott Morris IIRC, where are
your statistics showing that this is a real problem? Begin your analysis by
proving that the backlog is really caused by people trying the lab 10 or 20
times. Then you can discuss how to prevent it. Otherwise, it's just talking
in the wind. If only 5 % of seats are effectively occupied by
serial-test-takers, that's negligible and not the real issue.
-- Thomas Perrier CCIE #18731 (R&S)
This archive was generated by hypermail 2.1.4 : Fri Nov 16 2007 - 13:11:14 ART