Configuring Link Aggregation Groups (LAGs)

Creating LAG Interfaces

Syntax:

set link-aggregation interface <name> <attribute> <value>

Attribute Description

<name>

Specifies the name of the LAG interface. The supported LAG interface names: 'lag-1' to 'lag-99'.

<description>

Link aggregation interface description

mode <mode>

Specifies the LAG mode. The default mode is LACP. The possible modes are:

  • lacp - In this mode, the member links processes LACP packets received. When you create a LAG interface in LACP mode, the LACP PDUs are sent and received through member interfaces.

  • static: In this mode, the member links do not initiate or process any of the LACP packets received.

<minimum-link-count>

Specify the minimum number of active member links required for the link aggregation interface.

<member-interface>

Specify name of the member interface.

redundancy-session-id

Specify the value for the redundancy group session ID. Range from 1 to 65535 is allowed.

system-id

Specify the MAC address (as system ID) of the device for the link-aggregation interface.
redundancy-session-id, and system-id attributes can only be used when you deploy RBFS in redundancy mode. For information about LAG configuration when deploying RBFS in redundancy mode, see the RBFS Redundancy Solution Guide.

Example: LAG Interfaces Configuration

supervisor@rtbrick: cfg> show config link-aggregation
{
    "rtbrick-config:link-aggregation": {
      "interface": [
        {
          "interface-name": "lag-3",
          "mode": "lacp",
          "minimum-link-count": 2,
          "member-interface": [
            {
              "member-interface-name": "ifp-0/0/1",
              "lacp-mode": "active"
            },
            {
              "member-interface-name": "ifp-0/0/5",
              "lacp-mode": "active"
            }
            {

Example: LAG Interfaces Configuration

supervisor@rtbrick: cfg> show config link-aggregation
{
    "rtbrick-config:link-aggregation": {
      "interface": [
        {
          "interface-name": "lag-3",
          "mode": "lacp",
          "minimum-link-count": 2,
          "member-interface": [
            {
              "member-interface-name": "ifp-0/0/1",
              "lacp-mode": "active"
            },
            {
              "member-interface-name": "ifp-0/0/5",
              "lacp-mode": "active"
            }
            {
              "member-interface-name": "ifp-0/0/5",
              "lacp-mode": "passive"
            }
          ]
        }
      ]
    }
  }

Configuring LAG Member Interfaces

You can add member ports to the LAG interface. The command below allows you to bundle multiple physical interfaces with similar properties like speed, MTU, MRU.

Syntax:

set link-aggregation interface <name> member-interface <name> <attribute> <value>

Attribute Description

lacp-mode <mode>

Specifies the LACP mode. The default lacp-mode is active.
active: LACP packets are generated on each of the member links on the trad, the receive side.
passive: LACP packets are generated on the member link in response to the LACP packet received at one side of the LAG should be configured as active to bring the LAG interface.

lacp-timeout <timeout-value>

Specifies the timeout for the LACP session. A long timeout is 90 seconds, while a short is 3 seconds (default is short). Setting the timeout value will instruct the partner at which interval it should send the updates (30 seconds for long timeout, 1 second for short timeout).

Having mismatching timeouts will not break the operation, even though it is not desirable design-wise. This is because in LACP both Actor and Partner negotiate the transmission rate, so the transmitter sends at the receiver’s expected interval.

Configuring QoS on LAG Interface

RBFS supports QoS at physical interface level for LAG. Users can apply QoS profile at physical interface level through which one common QoS classification can be applied for all traffic on that port, irrespective of the destination logical interface.

The following features are supported:

  • Classification (IEEE-802.1)

  • Remarking (IEEE-802.1)

  • Ingress Policing

  • Egress Policing

For information about configuring the above features, refer the HQoS Configuration Guide.

  • You cannot apply QoS class of service on LAG logical interface

  • Currently, queuing and scheduling are not supported

Syntax:

set interface <physical interface> class-of-service <class-of-service>

Attribute Description

<interface>

Name of the interface

<class-of-service>

Specifies the class of service

Example:

supervisor@rtbrick: cfg> set interface lag-11 class-of-service Retail_profile
supervisor@rtbrick: cfg> commit
supervisor@rtbrick: cfg> show config int lag-11
{
    "rtbrick-config:interface": [
      {
        "name": "lag-11",
        "class-of-service": "Retail_profile"
      }
    ]
  }

Configuring L2X on LAG Interface

All forms of L2X that are supported on the regular physical interfaces are supported on LAG. The incoming packet is be matched to a specific L2X profile based on the Cross Connect configuration on the specified LAG interface.

The following match conditions are supported on the LAG interface:

  • Incoming LAG interface without any VLAN

  • Incoming LAG interface with a single VLAN

  • Incoming LAG interface with inner and Outer VLAN

  • Incoming LAG interface with any single VLAN

  • Incoming LAG interface with inner VLAN and any outer VLAN

For information about configuring L2X, see the L2X Configuration Guide.

The following table provides the L2X match action attributes which are supported on LAG interface.

Attribute Description

nexthop6 <nexthop>

Next-Hop address

match-type <match-type>

Match types with which traffic can be matched.

service-label <service_label>

Service label value. NOTE: Supported MPLS label values are 0 - 1048575. The reserved MPLS label range is 0 - 15. In RBFS, BGP uses the label range 20000 - 100000. It is recommended to assign label values outside of these reserved ranges to avoid conflicts.

ingress-vlan-operation <ingress-vlan-action>

VLAN operation on ingress side outer VLAN

ingress-outer-vlan <vlan-id>

Outer VLAN at ingress side

outgoing_ifp

Outgoing interface

vlan_operation

VLAN operation

outgoing_outer_vlan1

Outgoing outer VLAN