A signaled LSP has a primary path, which is either user-defined or computed by the ingress LER.
Note
This section describes the behavior of redundant paths. However, the user can exercise further control over the path selection process by specifying the path selection mode and preferred path using the select-path command.Optionally, the user can configure one or more redundant paths to serve as a backup. When the primary path fails, traffic for the LSP can be forwarded over the redundant path. When no redundant path is configured for the LSP, and when the primary path fails, the ingress LER automatically attempts to compute a new path to the egress LER, establish the new path, and then redirect traffic from the failed path to the new path.
Configuring a redundant path allows the user to exercise greater control over the rerouting process than when the ingress LER simply calculated a new path to the egress LER. When a redundant path is configured, when the primary path fails, the ingress LER attempts to establish the redundant path. As with the primary path, a redundant path follows an explicit route of loose or strict hops.
By default, the redundant path is established only when the primary path fails. The user can optionally configure a redundant path to operate in hot-standby mode. A hot-standby path is established at the same time the primary path in the LSP is established. Resources are allocated to the hot-standby path, although no packets for the LSP are sent over the hot-standby path until the primary path fails. When the primary path fails, the already-established hot-standby path immediately takes over from the primary path. Since the hot-standby path is already active, service outages that can arise from the process of signaling and establishing a new path are eliminated.
After the redundant path has been activated, the ingress LER continues to try to connect to the egress LER over the primary path, either indefinitely or up to the configured retry limit. When a connection over the primary path can be established, the redundant path is deactivated, and traffic for the LSP is again sent over the primary path. Once the primary LSP becomes available again, the redundant path is torn down; when the path is a hot-standby path, it reverts to its backup status.
The user can configure multiple redundant paths. When the primary path fails, the ingress LER attempts to establish a connection to the egress LER using the first redundant path configured for the LSP. When a connection cannot be established using the first redundant path, the second redundant path is tried, and so on. When a connection cannot be established after trying each redundant path in the configuration, the first redundant path is tried again, and the process repeats. This behavior can be further modified using the select-path command.
To configure a secondary path, first create a path. After the user creates the path, the user can specify that it is to be used as a redundant path. For example, complete the following steps to cause a path called alt_sf_to_sj to be used when the primary path in LSP tunnel1 fails.
The following example enables a path called alt_sf_to_sj to be used when the primary path in LSP tunnel1 fails. Once the LSP is enabled, both the primary and hot-standby paths are activated, although packets are directed over only the primary path.
device# configure device(config)# router mpls device(config-router-mpls)# lsp tunnel1 device(config-router-mpls-lsp-tunnel1)# secondary-path alt_sf_to_sj device(config-router-mpls-lsp-sec-path)# standby
Note
At the secondary path level, the user can configure separate values for the following parameters: Class of Service (CoS), setup and hold priority, bandwidth allocations, and inclusion or exclusion of interfaces in administrative groups. When the user does not configure these parameters at the secondary path level, the secondary path uses the default values for these parameters.