QoS - LLQ priority commands
When implementing QoS LLQ, there are various commands and syntaxes in the configuration of the priority. This can be applied using the
police command or using the
police-rate-in-kbps command. Here’s an overview of the differences involved:
Policing with the
- This method sets up an LLQ strict priority queue.
- The priority command itself enables LLQ for the class of traffic to which it’s applied, ensuring that traffic is given preferential treatment.
- When you combine the
prioritycommand with the
policecommand, you are explicitly setting a rate limit for the priority traffic. This is recommended to prevent the priority queue from consuming all available bandwidth and starving other queues.
- With this configuration, the policing is always in effect, regardless of whether the network is experiencing congestion or not. This means that the priority traffic will always be limited to the configured rate.
Policing with the
- This configuration also sets up LLQ for priority traffic.
- However, the policing in this case is conditional—meaning it’s only in effect during times of congestion.
- When congestion occurs, traffic exceeding the configured rate in kbps will be policed down to the configured rate.
- If there is no congestion, the policing doesn’t take effect, and the priority traffic isn’t rate-limited.
The main difference is that with the
priority command combined with the
police command, the priority traffic is always subjected to the configured rate limit. In contrast, with the
priority police-rate-in-kbps command, the rate limit only applies during periods of congestion. This means that in a congestion-free scenario, priority traffic could potentially use more bandwidth than the rate specified in the priority police-rate-in-kbps command.