QoS strict priority egress traffic scheduling

Strict priority scheduling (SP) scheduling is used to facilitate support for latency sensitive traffic. A strict priority scheduler drains all frames queued in the highest-priority queue before continuing on to service lower-priority traffic classes.

The following figure displays the frame scheduling order for an SP scheduler servicing two SP queues. The higher-numbered queue, SP2, has a higher priority.

Click to expand in new window
Strict priority schedule — two queues

The disadvantage of strict priority-based scheduling is that lower-priority traffic can be starved of any access.

The devices classify packets into one of eight internal priorities. For each egress port, there are 8 Virtual output queues (VOQ) allocated on each ingress TM core to support 8 priorities. SP queue input values map to traffic classes and range from 0 through 7. These are:

  • 0 - No strict priority queue.
  • 1 - Traffic Class 7 strict priority queue.
  • 2 - Traffic Class 6 through 7 strict priority queues.
  • 3 - Traffic Class 5 through 7 strict priority queues.
  • 4 - Traffic Class 4 through 7 strict priority queues.
  • 5 - Traffic Class 3 through 7 strict priority queues.
  • 6 - Traffic Class 2 through 7 strict priority queues.
  • 7 - Traffic Class 1 through 7 strict priority queues.

When configuring egress traffic scheduling you use credit request and grant mechanisms to perform QoS. The credit size is 1024B.

9038049-00 Rev AA