Hmm, I am not shure I understand your reasoning, I always tought * If the cueue increases there are to few hostmasters, or the hostmasters spend to long time on each request. * If the queue is long but not increasing, increasing the number of hostmasters will bring down the cueue to zero, but Possible fixes: - increase the number of hostmasters - lower the ammount of time spent on each request the first is simple, the second is more complex: - make the hostmasters work faster (better training, faster computers, other incentives) - simplify the procedures to make the work faster - simplify the policy so that the procedures could be simpler - clearify the policy/procedures so that the requests are resolved at first iteration rather than requiring a number of interactions to solve a problem - make the LIR staff sumitt better requests (better training, faster computers, cleverer people)
From looking at this from time to time, I don't think one single factor can solve the problem or bring us out of the situation we are in.
-hph