Controlling CRLSP path selection
Before performing the configuration tasks in this section, be aware of each configuration objective and its impact on your device.
MPLS TE uses CSPF to calculate a path according to the TEDB and constraints and sets up the CRLSP through RSVP-TE. MPLS TE provides measures that affect the CSPF calculation. You can use these measures to tune the path selection for CRLSP.
Configuring the metric type for path selection
Each MPLS TE link has two metrics: IGP metric and TE metric. By planning the two metrics, you can select different tunnels for different classes of traffic. For example, use the IGP metric to represent a link delay (a smaller IGP metric value indicates a lower link delay), and use the TE metric to represent a link bandwidth value (a smaller TE metric value indicates a bigger link bandwidth value).
You can establish two MPLS TE tunnels: Tunnel 1 for voice traffic and Tunnel 2 for video traffic. Configure Tunnel 1 to use IGP metrics for path selection, and configure Tunnel 2 to use TE metrics for path selection. As a result, the video service (with larger traffic) travels through the path that has larger bandwidth, and the voice traffic travels through the path that has lower delay.
To configure the metric type for tunnel path selection:
Step | Command | Remarks |
---|---|---|
1. Enter system view. | system-view | N/A |
2. Enter MPLS TE view. | mpls te | N/A |
3. Specify the metric type to use when no metric type is explicitly configured for a tunnel. | path-metric-type { igp | te } | By default, a tunnel uses the TE metric for path selection. Execute this command on the ingress node of an MPLS TE tunnel. |
4. Return to system view. | quit | N/A |
5. Enter MPLS TE tunnel interface view. | interface tunnel tunnel-number [ mode mpls-te ] | N/A |
6. Specify the metric type for path selection. | mpls te path-metric-type { igp | te } | By default, no link metric type is specified and the one specified in MPLS TE view is used. Execute this command on the ingress node of an MPLS TE tunnel. |
7. Return to system view. | quit | N/A |
8. Enter interface view. | interface interface-type interface-number | N/A |
9. Assign a TE metric to the link. | mpls te metric value | By default, the link uses its IGP metric as the TE metric. This command is available on every interface that the MPLS TE tunnel traverses. |
Configuring route pinning
When route pinning is enabled, MPLS TE tunnel reoptimization is not available.
Perform this task on the ingress node of an MPLS TE tunnel.
To configure route pinning:
Step | Command | Remarks |
---|---|---|
1. Enter system view. | system-view | N/A |
2. Enter MPLS TE tunnel interface view. | interface tunnel tunnel-number [ mode mpls-te ] | N/A |
3. Enable route pinning. | mpls te route-pinning | By default, route pinning is disabled. |
Configuring tunnel reoptimization
Tunnel reoptimization allows you to manually or dynamically trigger the ingress node to recalculate a path. If the ingress node recalculates a better path, it creates a new CRLSP, switches the traffic from the old CRLSP to the new CRLSP, and then deletes the old CRLSP.
Perform this task on the ingress node of an MPLS TE tunnel.
To configure tunnel reoptimization:
Step | Command | Remarks |
---|---|---|
1. Enter system view. | system-view | N/A |
2. Enter MPLS TE tunnel interface view. | interface tunnel tunnel-number [ mode mpls-te ] | N/A |
3. Enable tunnel reoptimization. | mpls te reoptimization [ frequency seconds ] | By default, tunnel reoptimization is disabled. |
4. Return to user view. | return | N/A |
5. (Optional.) Immediately reoptimize all MPLS TE tunnels that are enabled with the tunnel reoptimization feature. | mpls te reoptimization | N/A |
Setting TE flooding thresholds and interval
When the bandwidth of an MPLS TE link changes, IGP floods the new bandwidth information, so the ingress node can use CSPF to recalculate the path.
To prevent such recalculations from consuming too many resources, you can configure IGP to flood only significant bandwidth changes by setting the following flooding thresholds:
Up threshold—When the percentage of the reservable-bandwidth increase to the maximum reservable bandwidth reaches the threshold, IGP floods the TE information.
Down threshold—When the percentage of the reservable-bandwidth decrease to the maximum reservable bandwidth reaches the threshold, IGP floods the TE information.
You can also set the flooding interval at which bandwidth changes that cannot trigger immediate flooding are flooded.
This task can be performed on all nodes that the MPLS TE tunnel traverses.
To set TE flooding thresholds and the flooding interval:
Step | Command | Remarks |
---|---|---|
1. Enter system view. | system-view | N/A |
2. Enter interface view. | interface interface-type interface-number | N/A |
3. Set the up/down threshold. | mpls te bandwidth change thresholds { down | up } percent | By default, the up/down threshold is 10% of the link reservable bandwidth. |
4. Return to system view. | quit | N/A |
5. Enter MPLS TE view. | mpls te | N/A |
6. Set the flooding interval. | link-management periodic-flooding timer interval | By default, the flooding interval is 180 seconds. |