[an error occurred while processing this directive][an error occurred while processing this directive]

Configuring Routers to Support an ETH-DM Session

Configuring MEP Interfaces

Before you can start an Ethernet frame delay measurement session across an Ethernet service, you must configure two MX Series routers to support ETH-DM.

To configure an Ethernet interface on a DPC in MX Series router to support ETH-DM:

  1. On each router, configure two physical or logical Ethernet interfaces connected by a VLAN. The following configuration is typical for single-tagged logical interfaces:

    [edit interfaces]interface {ethernet-interface-name {vlan-tagging;unit logical-unit-number {vlan-id vlan-id; # Both interfaces on this VLAN}}}
    Both interfaces will use the same VLAN ID.
  2. On each router, attach peer MEPs to the two interfaces. The following configuration is typical:

    [edit protocols]oam {ethernet {connectivity-fault-management {maintenance-domain md-name { # On both routerslevel number;maintenance-association ma-name { # On both routers continuity-check {interval 100ms;hold-interval 1;}mep mep-id { # Attach to VLAN interfaceauto-discovery;direction (up | down);interface interface-name;priority number;}}}}}}

Ensuring that Distributed ppm Is Not Disabled

By default, the router’s period packet management process (ppm) runs sessions distributed to the Packet Forwarding Engine in addition to the Routing Engine. This process is responsible for periodic transmission of packets on behalf of its various client processes, such as Bidirectional Forwarding Detection (BFD), and it also receives packets on behalf of client processes.

In addition, ppm handles time-sensitive periodic processing and performs such processes as sending process-specific packets and gathering statistics. With ppm processes running distributed on both the Routing Engine and the Packet Forwarding Engine, you can run such processes as BFD on the Packet Forwarding Engine.

Distributed ppm Required for ETH-DM

Ethernet frame delay measurement requires that ppm remains distributed to the Packet Forwarding Engine. If ppm is not distributed to the Packet Forwarding Engines of both routers, ETH-DM PDU frame timestamps and ETH-DM statistics are not valid.

Before you start ETH-DM, you must verify that the following configuration statement is NOT present:

[edit]routing-options { ppm { no-delegate-processing; }}

If distributed ppm processing is disabled (as shown in the stanza above) on either router, you must re-enable it in order to use the ETH-DM feature.

Procedure to Ensure that Distributed ppm is Not Disabled

To ensure that distributed ppm is not disabled on a router:

  1. Display the packet processing management (PPM) configuration to determine whether distributed ppm has been disabled.

    • In the following example, distributed ppm is enabled on the router. In this case, you do not need to modify the router configuration:
      [edit]user@host# show routing-optionsppm;
    • In the following example, distributed ppm is disabled on the router. In this case, you must proceed to Step 2 to modify the router configuration:
      [edit]user@host show routing-optionsppm {no-delegate-processing;}
  2. Modify the router configuration to re-enable distributed ppm and restart the Ethernet OAM Connectivity Fault Management process ONLY IF distributed ppm is disabled (as determined in the previous step).

    1. Before continuing, make any necessary preparations for the possible loss of connectivity on the router.

      Restarting the ethernet-connectivity-fault-management process has the following effect on your network:

      • All connectivity fault management (CFM) sessions re-establish.
      • All ETH-DM requests on the router terminate.
      • All ETH-DM statistics and frame counts reset to 0.
    2. Modify the router configuration to re-enable distributed ppm. For example:
      [edit]user@host# delete routing-options ppm no-delegate-processing
    3. Commit the updated router configuration. For example:
      [edit]user@host# commit and-quitcommit completeexiting configuration mode
    4. To restart the Ethernet OAM Connectivity-Fault-Management process, enter the restart ethernet-connectivity-fault-management <gracefully | immediately | soft> operational mode command. For example:
      user@host> restart ethernet-connectivity-fault-managementConnectivity fault management process started, pid 9893

Enabling the Hardware-Assisted Timestamping Option

By default, Ethernet frame delay measurement uses software for timestamping transmitted and received ETH-DM frames. For Ethernet interfaces on Enhanced Dense Port Concentrators (DPCs) and Enhanced Queuing DPCs only, you can optionally use hardware timing to assist in the timestamping of received ETH-DM frames to increase the accuracy of delay measurements.

Enabling hardware-assisted timestamping of received frames can increase the accuracy of ETH-DM calculations when the DPC is loaded with heavy traffic in the receive direction.

To enable Ethernet frame delay measurement hardware assistance on the reception path, include the hardware-assisted-timestamping statement at the [edit protocols oam ethernet connectivity-fault-management performance-monitoring] hierarchy level:

Published: 2010-04-20

[an error occurred while processing this directive]