Help us improve your experience.

Let us know what you think.

Do you have time for a two-minute survey?

Configuring LMP Peers

 

You need to configure network peers for GMPLS. A peer is a network device that your router communicates with when setting up the control and data channels. The peer is often an optical cross-connect (OXC).

To configure an LMP peer name, include the peer statement at the [edit protocols link-management] hierarchy level:

The following sections describe how to configure an LMP peer:

Configuring the ID for LMP Peers

To configure the LMP peer ID, include the address statement at the [edit protocols link-management peer peer-name] hierarchy level. The default value for the LMP peer ID is the loopback address.

Configuring the Interface for Control Channels Between LMP Peers

You must configure one or more control channels between the LMP peers. The control channels must travel across either a point-to-point link or a tunnel.

To configure the interface for the control channel, include the control-channel statement at the [edit protocols link-management peer peer-name] hierarchy level:

You can configure a generic routing encapsulation (GRE) interface (gre-x/y/z) for the control channel. This type of interface does not require a Tunnel PIC.

Note

You can configure GRE interfaces (gre-x/y/z) only for GMPLS control channels. GRE interfaces are not supported or configurable for other applications. For more information, see the Junos OS Network Interfaces Library for Routing Devices.

Configuring the LMP Control Channel Interface for the Peer

In an environment that uses LMP to establish and maintain an LMP control channel between peers, you can configure a number of attributes associated with LMP. To configure the interface to be associated with the LMP control channel for the peer, include the lmp-control-channel statement:

You can configure this statement at the following hierarchy levels:

  • [edit protocols link-management peer peer-name]

  • [edit logical-systems logical-system-name protocols link-management peer peer-name]

You can configure a GRE interface for the LMP control channel. This type of interface does not require a Tunnel PIC.

Note

You can configure GRE interfaces only for GMPLS control channels. GRE interfaces are not supported or configurable for other applications. For more information, see the Junos OS Network Interfaces Library for Routing Devices.

When this LMP control channel interface comes up, the peers use LMP to negotiate channel parameters and configure the control channel.

The local peer repeatedly sends a Config message to the remote peer. The Config message contains the local control channel ID, the local peer’s node ID, a message ID, and a CONFIG object that includes hello message attributes (the hello interval and the hello dead interval).

The channel is activated when the remote peer responds with a ConfigAck message. The remote peer does so only when its own configured hello interval and hello dead interval match the values in the received Config message or the default values. If these values do not match, the remote peer responds with a ConfigNack message. The local peer logs this event and resends the Config message until the message retry limit is reached. When the message retry limit is reached, the local peer logs that event and restarts the configuration process.

Configuring the Remote IP Address for LMP Control Channels

You need to specify the address of the remote end of the LMP control channel.

To configure the remote IP address for the LMP control channel, include the remote-address statement:

You can configure this statement at the following hierarchy levels:

Configuring Hello Message Intervals for LMP Control Channels

Hello messages are exchanged between LMP peers to maintain the control channel after LMP has activated the control channel. The LMP control channel is considered to be up only when the hello negotiation is successful. Successful negotiation consists of the local peer sending a hello message to the remote peer and receiving a hello message in response.

The LMP peers continue to exchange hello messages after the LMP control channel is up in order to maintain the channel.

The hello interval specifies the interval between periodic hello messages. The hello dead interval specifies how long the local peer waits for a hello response before it declares the LMP control channel to be down. When the channel goes down, the local peer restarts the LMP control channel negotiation and configuration process.

You can specify a hello interval from 150 through 300,000 milliseconds. The default hello interval is 150 milliseconds.

You can specify a hello dead interval from 500 through 300,000 milliseconds. The default hello dead interval is 500 milliseconds.

To configure the attributes for hello messages exchanged between LMP peers, include the hello-interval and hello-dead-interval statements:

You can configure these statements at the following hierarchy levels:

When an LMP control channel comes up after a successful exchange of hello messages between LMP peers, LMP uses link property correlation to verify the traffic engineering and data link information on both sides of a link. To do so, the local peer sends a LinkSummary message for each traffic engineering link governed by the LMP control channel. The LinkSummary message contains information that characterizes the traffic engineering link and each data link in the traffic engineering link.

The local peer continues sending a LinkSummary message for each link until the remote peer responds with a LinkSummaryAck message or until the message retry limit is reached. When the message retry limit is reached, the local peer logs that event and restarts the link property correlation process.

When the remote peer receives a LinkSummary message, it examines its own link information. If this information agrees with that in the LinkSummary message, the remote peer responds with a LinkSummaryAck message. If the information is different, the remote peer responds with a LinkSummaryNack message.

Controlling Message Exchange for LMP Control Channels

You can configure message attributes that control the exchange of LMP Config and LinkSummary messages. The retransmission interval specifies the interval between resubmitted LMP messages. The retry limit specifies how many times LMP sends a message before restarting the process.

You can specify a retransmission interval from 500 through 300,000 milliseconds. The default retransmission interval is 500 milliseconds.

You can specify a retry limit from 3 through 1000 attempts. The default number of retry attempts is three.

To configure attributes governing the exchange of LMP messages between peers, include the retransmission-interval and retry-limit statements:

You can configure these statements at the following hierarchy levels:

Preventing the Local Peer from Initiating LMP Negotiation

You can specify that the local peer does not initiate LMP negotiation. Instead, the local peer waits for the remote peer to configure the LMP control channel.

To configure the local peer to wait for the remote peer to configure the LMP control channel, include the passive statement:

You can configure this statement at the following hierarchy levels:

Associating Traffic Engineering Links with LMP Peers

To specify the name of a traffic engineering link to be associated with this peer, include the te-link statement at the [edit protocols link-management peer peer-name] hierarchy level:

For information about how to configure a traffic engineering link, see Configuring LMP Traffic Engineering Links.

Disabling the Traffic Engineering Link for LMP Peers

To disable a specific traffic engineering link, include the disable statement:

You can configure this statement at the following hierarchy levels:

  • [edit protocols link-management te-link te-link-name]

  • [edit logical-systems logical-system-name protocols link-management te-link te-link-name]