WRR

weighted round robin

This queuing technique creates a number of queues and allows a user to assign incoming traffic to each queue by some distinguishing factor. This could be service class, address, protocols, or any other number of factors. To ensure each queue is serviced fairly, the user defines a weighting for each queue. Like round robin queuing, the scheduler visits each queue in turn. However, the weighting impacts the number of packets released from each queue when it is visited.

The primary problem with WRR is that it operates at the packet level. This means that if the queues contain packets of differing average lengths, the packet percentages won’t be realized as bandwidth percentages.