Talk:Processor sharing

Potential contradiction
The section labeled "Queuing Theory" appears to contradict the previous section.

"Queueing theory

A single server queue operating subject to Poisson arrivals (such as an M/M/1 queue or M/G/1 queue) with a processor sharing discipline has a geometric stationary distribution.[1]"

The previous section says that there is no queuing when a processor sharing discipline is used.


 * There is no waiting time before service, so in that sense there is no queueing, but the number of jobs at the service node has a geometric stationary distribution. Different authors in different situations will consider the queue to be either just customers waiting or the sum of customers waiting and those in service. You are right that under regular processor sharing there will never be any customers waiting for service, so the size of the queue is made up entirely of jobs currently in service. Gareth Jones (talk) 12:25, 21 March 2015 (UTC)