The packet shaping algorithm dequeues packets in the following order:
- Classes under guaranteed SLA marked as priority
- Classes under guaranteed SLA
- Classes over guaranteed marked as priority
- Classes over guaranteed
The guaranteed SLA corresponds to the Minimum of the SLA shaping setting (see image below).
Also in the image one of the user-defined SLA shaping setting’s Priority flag is set to true. All the classes with this SLA shaping setting will have its network packets dequeued according to positions 1 and 3 in the packet shaping algorithm shown above.
Marking as priority is implemented to reduce the latency of packets when they are in shaping queues. Note that the SLA shaping setting priority nor minimum values have no effect unless the classes’ packets are in the shaping queue. For a classes’ packets to be in the shaping queue the total of all the network traffic would be exceeding the WAN Link Outbound Maximum or that particular class is exceeding it’s own SLA shaping setting’s Maximum. See the article ‘Creating a Shaping Policy‘ for more explanation.
So the Priority flag only helps when there is congestion and shaping occurring. The latency of network packets would be much less if they did not accumulate on the shaping queues.