Load sharing using directly connected loopback peering does not apply to Carrier Supported Carrier (CSC) networks that use the Label Distribution Protocol (LDP) and an Interior Gateway Protocol (IGP) to distribute routes and Multiprotocol Label Switching (MPLS) labels.
The software does not support load balancing in interautonomous system (Inter-AS) and CSC when there are multiple links between provider edge (PE) or Autonomous System Boundary Router (ASBR) devices.
When you configure static routes in an MPLS or MPLS virtual private network (VPN) environment, some variations of the ip route and ip route vrf commands are not supported. These variations of the commands are not supported in Cisco software releases that support the Tag Forwarding Information Base (TFIB. The TFIB cannot resolve prefixes when the recursive route over which the prefixes travel disappears and then reappears. However, the command variations are supported in Cisco software releases that support the MPLS Forwarding Infrastructure (MFI). Use the following guidelines when configuring static routes.
Supported Static Routes in an MPLS Environment
The following ip route command is supported when you configure static routes in an MPLS environment:
-
ip route destination-prefix mask interface next-hop-address
The following ip route commands are supported when you configure static routes in an MPLS environment and configure load sharing with static nonrecursive routes and a specific outbound interface:
-
ip route destination-prefix mask interface1 next-hop1
-
ip route destination-prefix mask interface2 next-hop2
Unsupported Static Routes in an MPLS Environment That Uses the TFIB
The following ip route command is not supported when you configure static routes in an MPLS environment:
-
ip route destination-prefix mask next-hop-address
The following ip route command is not supported when you configure static routes in an MPLS VPN environment and enable load sharing where the next hop can be reached through two paths:
-
ip route destination-prefix mask next-hop-address
The following ip route command is not supported when you configure static routes in an MPLS VPN environment and enable load sharing where the destination can be reached through two next hops:
-
ip route destination-prefix mask next-hop1
-
ip route destination-prefix mask next-hop2
Use the interface and next-hop arguments when specifying static routes.
Supported Static Routes in an MPLS VPN Environment
The following ip route vrf commands are supported when you configure static routes in an MPLS VPN environment, and the next hop and interface are associated with the same virtual routing and forwarding (VRF) instance:
-
ip route vrf vrf-name destination-prefix mask next-hop-address
-
ip route vrf vrf-name destination-prefix mask interface next-hop-address
-
ip route vrf vrf-name destination-prefix mask interface1 next-hop1
-
ip route vrf vrf-name destination-prefix mask interface2 next-hop2
The following ip route vrf commands are supported when you configure static routes in an MPLS VPN environment, and the next hop is in the global table in the MPLS cloud in the global routing table. For example, these commands are supported when the next hop is pointing to the internet gateway.
-
ip route vrf vrf-name destination-prefix mask next-hop-address global
-
ip route vrf vrf-name destination-prefix mask interface next-hop-address (This command is supported when the next hop and the interface are in the core.)
The following ip route commands are supported when you configure static routes in an MPLS VPN environment and enable load sharing with static nonrecursive routes and a specific outbound interfaces:
-
ip route destination-prefix mask interface1 next-hop1
-
ip route destination-prefix mask interface2 next-hop2
Unsupported Static Routes in an MPLS VPN Environment That Uses the TFIB
The following ip route command is not supported when you configure static routes in an MPLS VPN environment, the next hop is in the global table in the MPLS cloud within the core, and you enable load sharing where the next hop can be reached through two paths:
-
ip route vrf destination-prefix mask next-hop-address global
The following ip route commands are not supported when you configure static routes in an MPLS VPN environment, the next hop is in the global table in the MPLS cloud within the core, and you enable load sharing where the destination can be reached through two next hops:
-
ip route vrf destination-prefix mask next-hop1 global
-
ip route vrf destination-prefix mask next-hop2 global
The following ip route vrf commands are not supported when you configure static routes in an MPLS VPN environment, and the next hop and interface are in the same VRF:
-
ip route vrf vrf-name destination-prefix mask next-hop1
-
ip route vrf vrf-name destination-prefix mask next-hop2
Supported Static Routes in an MPLS VPN Environment Where the Next Hop Resides in the Global Table on the CE Device
The following ip route vrf command is supported when you configure static routes in an MPLS VPN environment, and the next hop is in the global table on the customer edge (CE) side. For example, the following command is supported when the destination-prefix is the CE device’s loopback address, as in external Border Gateway Protocol (eBGP) multihop cases.
-
ip route vrf vrf-name destination-prefix mask interface next-hop-address
The following ip route commands are supported when you configure static routes in an MPLS VPN environment, the next hop is in the global table on the CE side, and you enable load sharing with static nonrecursive routes and a specific outbound interfaces:
-
ip route destination-prefix mask interface1 nexthop1
-
ip route destination-prefix mask interface2 nexthop2