MDRR Insertion and Drop Policy

MDRR Insertion and Drop Policy

• MDRR uses a traditional tail-drop scheme if a queue is congested.
• MDRR can also use weighted random early detection (WRED) to prevent congestion.

• Virtual Output Queue

• Tail Drop
• or
• WRED

DRR Scheduling

• Service policy for one queue in one round:
• Add MTU+(Weight-1)*512 tokens to the token bucket.
• Transmit packets until tokens are used up or the queue is empty.
• Reset the token bucket to 0 if the queue is empty. Otherwise remember the deficit (how much more tokens were used than available).
• Start serving the next queue.

• VOQ 0

• VOQ 1

• VOQ 7

• Each queue can transmit a configured amount of bytes in one round:
• MTU + (weight-1)*512

• Round
• Robin
• Scheduler

MDRR Scheduling with Strict Priority Queue

• Service policy for MDRR with strict priority:
• Transmit packets from the strict priority low-latency queue until the queue is empty.
• Serve the next-in-line round-robin queue.
• Start serving the low-latency queue again.

• VOQ 0

• VOQ 1

• VOQ 7

• The strict priority, low-latency queue is not limited by the token bucket mechanism.

• Round-Robin Scheduler

• LL Queue

• Strict Priority
• Queuing

MDRR Scheduling with Alternate Priority Queue

• Service policy for MDRR with alternate priority:
• Transmit packets from the alternate priority, low-latency queue until the tokens are used up or the queue is empty.
• Serve the next-in-line round-robin queue
• Start serving the low-latency queue again.

• VOQ 0

• VOQ 1

• VOQ 7

• The alternate priority queue is using the token bucket to limit the amount of bytes it can transmit in one round.

• Round-Robin Scheduler

• LL Queue

• Alternate
• Priority
• Queuing