load-balance hash-fields ip l3

Function

The load-balance hash-fields ip l3 command configures IP-address-based load balancing in the following scenarios: On PEs over a backbone network in an L3VPN over L2VPN, L2VPN over L2VPN, or an IP over IP scenario.On Ps in common scenarios. L2VPN supports VPLS and VLL that use Ethernet interfaces for access and homogeneous media without control word.

The undo load-balance hash-fields ip l3 command restores the default configuration.

IP packets are balanced in different ways depending on different scenarios by default:

In an L3VPN over L2VPN, L2VPN over L2VPN, IPv6 Over L2VPN scenario, traffic is balanced based on MPLS labels carried in IP packets.
In an IP over IP scenario, traffic is balanced based on an inner IP header if it can be obtained or an outer IP header if the inner IP header cannot be obtained.
If a device functions as a P in a common scenario, traffic is balanced based on MPLS labels. Specifically, traffic is balanced based on the third outer MPLS label in case of more than five but less than thirteen incoming MPLS labels.

Format

load-balance hash-fields ip l3

undo load-balance hash-fields ip l3

Parameters

None

Views

100GE interface view, 400GE interface view, 40GE interface view, GE optical interface view, GE interface view

Default Level

2: Configuration level

Task Name and Operations

Task Name	Operations
forwarding	write

Usage Guidelines

Usage Scenario

In L2/L3VPN over L2VPN scenarios, L2-P1 resides on the backbone network.By default, L2-P1 balances traffic among outbound Eth-Trunk member interfaces based on the MPLS labels of the packets. Because packets transmitted over the same VPLS tunnel carry the same MPLS labels, traffic from the same site but different L3VPNs cannot be load-balanced among the Eth-Trunk member interfaces. To resolve this problem, run the load-balance hash-fields ip l3 command on the inbound interface of L2-P1 to allow L2-P1 to load-balance traffic from the same site but different L3VPNs among links between L2-P1 and L2-PE2 in the L3VPN over VPLS scenario.

For information about fundamentals, see Configuration Guide > IP Service > Load Balancing Configuration > Configuring Hash Factors (L2/L3VPN Over L2VPN Scenarios).

Configuration Impact

After this command is configured:

The Layer 2 ACL becomes unavailable in the inbound direction. This change affects only MPLS L2VPN and L3VPN packets transmitted over the VPLS tunnel.
Port-based MAC address learning restriction becomes unavailable in the inbound direction. This change affects only MPLS L2VPN and L3VPN packets transmitted over the VPLS tunnel.
The performance of forwarding L3VPN over L2VPN and L2VPN over L2VPN traffic decreases on a board.
PWE3 services of TDM, ATM and SDH cannot be deployed.
The P devices cannot function as egress PEs.

Precautions

This command cannot be run on an Eth-Trunk. It must be run on a trunk's member interface. After this command is run on a trunk's member interface and the member interface is removed from the trunk, you must delete the command and reconfigure it for the configuration to take effect.

Example

# Configure IP-address-based load balancing on GE 0/1/0 in the L3VPN over VPLS scenarios. GE 0/1/0 is the inbound interface on a NetEngine 8000 F.

<HUAWEI> system-view
[~HUAWEI] interface GigabitEthernet 0/1/0
[*HUAWEI-GigabitEthernet 0/1/0] load-balance hash-fields ip l3