Ethernet Ring Protection

Overview

Ethernet Ring Protection (ERP) is a protection mechanism that maintains high availability of Ethernet services for ring topology networks. ERP is based on the ITU-T G.8032 specification.

An ERP ring consists of at least two units: one unit is configured as the owner and a second adjacent unit as the neighbor. Similar to Spanning Tree Protocol (xSTP), ERP implements a mechanism to prevent loops. Thus, although the physical topology is ring, traffic on one of the links is blocked and therefore the actual logical topology is linear. In ERP, the blocked link is the one between the owner and neighbor and is referred to as the Ring Protection Link (RPL). During normal operation, customer traffic does not travel on the RPL. During a failure, the RPL is unblocked and traffic can move between the owner and neighbor; the failed link thus becomes the new RPL.

Switchover to protection is triggered by link failure, which can be based on link status or Connectivity Fault Management (802.1ag CFM). ERP protection allows switchover within 50 milliseconds.

Supported Hardware

Configuration Examples

The examples in this section explain how to configure an ERP ring for the following situations. The configuration of the ERP is the same on all units (GT, LT, NE, etc.). The only difference is the LAG port and physical ports that will be used.

• Example: Setting Up an ERP Ring
• Example: Adding a Unit to an ERP Ring
• Example: Removing a Unit from an ERP Ring

Example: Setting Up an ERP Ring

The figure below shows a typical configuration for this scenario.
The first diagram is a typical example of how the ring is used in real world examples; the second diagram will be used for our lab setup.

Fault detection, you have two options:

1. Link status to trigger a protection switch (Port down will trigger the switch)

2. Y.1731 CCM messages to detect a link failure and trigger a protection switch. (missing CCM messages will trigger the switch)

Options 1 and 2 are both valid, the reason you would use one or the other may vary but here is the basic scenario.

If you have dark fiber between two sites then you may simply want to switch on a port down (Link status) failure. If you are traversing a network with other network elements between the two elements then you will want to use Y.1731 CCM as a trigger method, using a simple Port status will not work in this scenario. (The local port may be up but the path to the remote is down)

Some ERP basic steps:
Configure one or more ERP instances. ERP v1 supports only one ERP instance. ERP v2 supports multiple ERP instances. You should determine which version of ERP you need when configuring the LAG ports. The LAG (Link Aggregation Group) ports are used for associating ports in ring mode. An ERP instance also determines which VLANs can circulate on the ring. Each ERP instance has its own signaling channel, also referred to as APS. In this example, we create one ERP instance that uses VIDs 100 to 110 inclusively, and for
which APS is on VLAN 2.

When configuring an ERP ring, it is preferable to maintain a consistent pattern for the ports. For example, you can always use port SFP-1-1 heading clockwise and port SFP-2-1 heading counter clockwise. This will make all configurations easier to remember.

If you are mixing different models its best practice to prepare a diagram you can refer to while configuring the units.

The basic steps required to set up Ethernet Ring Protection are:

• "Physically connect the units in a ring"
• "Create the ERP instances"
• "Put the LAG port into ERP mode"
• "Create the CFM instances "
• "Associate the CFM instances to the ERP instances"
• "Set up in-band management"
• "Set up an EVC drop"

Each of these steps is described in detail below.

Physically connect the units in a ring
Connect the units in a ring using ports SFP-1-1 and SFP-2-1 for each unit, as shown in the figure above.

Create the ERP instances
Create the ERP instances on all units using the following configuration summary table as a guideline.

Configuration Summary of the ERP Instances (Using Y.1731 CCM)

Configuration Summary of the ERP Instances (Using Link Status) MEP0-idx, MEP1–idx: are set to NONE for Port status configuration

For Unit 1 (ERP Owner):

1. Go to Protection ▶ ERP ▶ Configuration and click Add.

2. Set up this ERP instance as required using the figure below.
   a. For all units, set the Name to RING.
   b. For Unit 1, select RPL-role to Owner and RPL-port to 0 (SFP-1-1).
   c. For Unit 1, enter the appropriate APS-VID, VID-list and LAG port. Example: APSVID is set to 2, VID-list to 100-110, LAG port to LAG-1-1.

3. Click Apply.

4. For Unit 2 (Neighbor), create an ERP instance by repeating the steps above, but set the RPL-roleto Neighbor and the RPL-port to 1 (SFP-2-1).

5. For Units 3 and 4, create an ERP instance by repeating the steps above, but set the RPL-role to None.

Put the LAG port into ERP mode
Put the LAG-1-1 port into ERP mode on all units using the following configuration summary table as a guideline.

Configuration Summary of the LAG-1-1 Ports

For each unit:

1. Log in to the unit.

2. Access the page Port ▶ Configuration.

3. Click the LAG Port name (e.g. LAG-1-1) to edit its settings.
   

   ---

   Note: LAG-1-1 protects ports SFP-1-1 and SFP-2-1; traffic is mapped to Traffic-1-1. LAG-1-2 protects ports SFP-1-2 and SFP-2-2; traffic is mapped to Traffic-1-2.

   ---

4. Define the LAG port protection parameters as required by your setup using the figure below.
   a. Select the Enable and Protection Enable options.
   b. Select the Protection Mode of ERP Status.
   c. Select the Version number 2 (recommended).

5. Click Apply.

Create the CFM instances

Set up an 802.1ag CFM (Connectivity Fault Management) maintenance association (MA) on each link of the ring in order to detect a node or link failure. The failure of a CFM MA will trigger the ERP switchover.

The following tables show a configuration summary for MEG1 and MEG2, and MEP1 and MEP2 units.
For MEG1 and MEG2, the MEG name represents the different links. MEG 1 of each unit is also associated with MEG 2 of its next neighbor, going clockwise around the ring. These MEG names must match exactly to establish a CFM connection between units.

Refer to Application Note AN138 to learn more about configuring Service OAM and CFM MA.

Configuration Summary of the MEG1

Configuration Summary of the MEG2

Configuration Summary of the MEP2

Associate the CFM instances to the ERP instances
Return to the ERP configuration settings to select the appropriate MEP index for each port.

CAUTION: You must complete this step to finish configuring the protection switch based on ERP status using CFM instances.

For each unit:

1. Access the page Protection ▶ ERP ▶ Configuration and click the Index of the RING ERP instance to edit its settings.

2. Select MEP0-idx to 1 and MEP1-idx to 2.

Set up in-band management
In-band management enables the use of a dedicated management VLAN to control the nodes in the ring by connecting the station to the out-of-band management port (ETH-1) of any given ring node. Although in-band management is not mandatory, its use is highly recommended to prevent having to manually set each unit each time a configuration change is required.

Example: we will use Unit 1 for connecting to the management station.

1. On each unit, create a VLAN interface on port LAG-1-1.

2. On Units 2, 3 and 4, assign an IP address on the VLAN interface.

3. On Unit 1 only, create a bridge between LAG-1-1 and ETH-1 and assign an IP address to the bridge.

Set up an EVC drop
You must set up a policy to ensure that customer traffic flows from a LAG port in the ERP ring to a customer port. In ring networking terminology, this is known as a drop. The ERP implementation of the 10GE Performance Element allows an EVC to exist only between a pair of nodes (E-LINE or EV-LINE). Example: to create an EVC on VLAN 105 between Unit 1 and Unit 3, we must set up traffic policies on Units 1 and 3 to define what traffic to forward on the EVC. To do this, set up Traffic-1-1 as an outbound policy, i.e. traffic direction is from ring (LAG-1-1) to customer (SFP-1-2), and Traffic-1-2 as an inbound policy (from customer to ring).

To define what traffic to allow and what to block, we must first create a traffic filter based on a VID number on Unit 1.

1. Access the page Traffic ▶ VLAN ▶ VID Sets and click Add.

The screen below is displayed.

2. Setup the VID set as shown in the figure above, then click Apply.
   

   ---

   Note: Although the VLAN IDs can be used in multiple policies, the name of the VID set must be unique for both Traffic-1-1 and Traffic-1-2.

   ---

3. Access the page Traffic ▶ VLAN ▶ VID Sets and click Add to create a new VID Set for the opposite direction.

4. Setup the VID set as shown in the figure above, but you must enter a different Name
   (e.g. VID 105 (in)) and a different Policy list (e.g. Traffic-1-2), then click Apply.

5. Access the page Traffic ▶ Policies.

6. Under the Policy lists, select Traffic-1-1 to display the Traffic-1-1 policies.

7. Select a policy index (1-1 to 1-123). This displays the screen below.
   

8. Set up the policy set as shown in the figure above.
   a. Set Outgoing port to SFP-1-2.
   b. Set Filter Type to VID set.
   c. Select the Filter created previously. e.g. VID 105 (out).
   d. Set Action to Permit Traffic.

9. Click Apply.

10. Set up another policy for Traffic-1-2 that uses an outgoing port of SFP-1-1. To do so, repeat the three steps above but with the following differences:
    a. On the Policies tab, select Traffic-1-2.
    b. Choose a Policy index from 2-1 to 2-123.
    c. Click the Enable policy option.
    d. Select Outgoing port of SFP-1-1.
    e. Set Filter Type to VID set.
    f. Select the Filter created previously, e.g. VID 105 (in).
    g. Set Action to Permit Traffic.

11. Click Apply.

12. Set up two VID sets and two traffic policies on Unit 3 by repeating all steps above.
    This creates the EVC drop on unit 3.

You have now created an EVC between unit 1 and unit 3 for VLAN105.

Example: Adding a Unit to an ERP Ring

This example explains how to add a 10GE Performance Element to an ERP ring without affecting the traffic on the ring.

The figure below shows a typical network for this scenario. The new unit will be added between unit 2 and unit 3.

The basic steps required to add a unit to an ERP ring are:

• "Isolate the link that needs to be broken to insert the new node"
• "Delete the MA/MEG and MEP on the link to be broken (not required for Link status
• configuration)"
• "Physically connect the new unit"
• "Create an ERP instance on the new unit"
• "Put the LAG port into ERP mode on the new unit"
• "Create CFM instances on the new unit"
• "Reconfigure MA/MEG and MEP on adjacent units"
• "Verify connectivity"
• "Clear the forced switch"

Note: Before adding a unit to an ERP ring, you must ensure that there is no failure in the ring.
Isolate the link that needs to be broken to insert the new node
Before adding a new unit to an ERP ring, you must first force a switch to the standby port on each adjacent unit. Example: we want to insert the new unit between Unit 2 and Unit 3.

For Unit 2:

1. Access the page Protection ▶ ERP ▶ Status.

2. Click the ERP Index of the ring you want to switch traffic from.

3. Click the Port 0 Force Switch button to force a switch from this port (in this example SFP-1-1) to the standby port.

For Unit 3:

1. Access the page Protection ▶ ERP ▶ Status.

2. Click the ERP Index of the ring you want to switch traffic from.

3. Click the Port 1 Force Switch button to force a switch from this port (in this example SFP-2-1) to the standby port.

Delete the MA/MEG and MEP on the link to be broken (not required for Link status configuration)
Delete the MA/MEG and MEP associated with this ring configuration on each adjacent unit.

Refer to Application Note AN138 to learn more about configuring Service OAM and CFM MA/MEG and MEP.

Physically connect the new unit
Power on the new unit and connect its SFP-1-1 and SFP-2-1 ports to Unit 2 and Unit 3.

Create an ERP instance on the new unit
Create an ERP instance on the new unit following the procedure described in "Example: Setting Up an ERP Ring" . The following table shows a configuration summary of the ERP instances of all units.

Configuration Summary of the ERP Instances (Y.1731 CCM protection)

Configuration Summary of the ERP Instances (Link Protection)

Put the LAG port into ERP mode on the new unit

Put the LAG-1-1 port into ERP mode on the new unit following the procedure described in
"Example: Setting Up an ERP Ring" . The following table shows a configuration summary of the LAG ports of all units.

Configuration Summary of the LAG-1-1 Ports

Create CFM instances on the new unit

Create the CFM instances on the new unit following the procedure described in "Example: Setting Up an ERP Ring".

The following tables show a configuration summary of MEG1 and MEG2, and MEP1 and MEP2 of all units.

Refer to Application Note AN138 to learn more about configuring Service OAM and CFM MA.

Configuration Summary of the MEG1

Configuration Summary of the MEG2

Configuration Summary of the MEP2

Configuration Summary of the MEP2

Reconfigure MA/MEG and MEP on adjacent units
Re-establish an 802.1ag CFM (Connectivity Fault Management) MA/MEG and MEP on both adjacent units.

Refer to Application Note AN138 to learn more about configuring Service OAM and CFM MA/MEG and MEP.

Verify connectivity
Verify Ethernet connectivity on the new unit. Verify also that all MEPs are active on the new unit and on both adjacent units (Units 2 and 3).

Clear the forced switch
Finally, clear the forced switch on Units 2 and 3 to re-enable ERP protection in the complete ring.
For each unit (1 and 3):

1. Access the page Protection ▶ ERP ▶ Status.

2. Click the ERP Index of the ring into which the new unit was inserted.

3. Click the Clear button to clear the forced switch.

Example: Removing a Unit from an ERP Ring

This example explains how to remove a 10GE Performance Element from an ERP ring without affecting the traffic on the ring.

The figure in "Example: Setting Up an ERP Ring" displays the standard configuration for this scenario. Unit 3 will be removed.

The basic steps required to remove a unit from an ERP ring are:

Ensure that there is no failure on the ring

• "Isolate the unit to be removed by forcing a port switch on the adjacent units"
• "Delete the MA/MEG and MEP on the adjacent units"
• "Physically disconnect the unit being removed"
• "Reconfigure MA/MEG and MEP on adjacent units"
• "Verify connectivity"
• "Clear the forced switch"

Isolate the unit to be removed by forcing a port switch on the adjacent units

After ensuring there is no failure on the ring, you can then force a port switch on both adjacent units. Example: to remove Unit 3, we must switch ports on Unit 2 and Unit 4.

For Unit 2:

1. Access the page Protection ▶ ERP ▶ Status.

2. Click the ERP Index of the ring.

3. Click the Port 0 Force Switch button to force a switch from this port (in this example SFP-1-1) to the standby port.

For Unit 4:

1. Access the page Protection ▶ ERP ▶ Status.

2. Click the ERP Index of the ring.

3. Click the Port 1 Force Switch button to force a switch from this port (in this example SFP-2-1) to the standby port.

Delete the MA/MEG and MEP on the adjacent units
Delete the MA/MEG and MEP associated with this ring configuration on the two units adjacent to the unit being removed.

Refer to Application Note AN138 to learn more about configuring Service OAM and CFM MA/MEG and MEP.

Physically disconnect the unit being removed Power off the unit being removed. Disconnect the cables and reconnect them between Unit 2 and Unit 4.

Reconfigure MA/MEG and MEP on adjacent units
Re-establish an 802.1ag CFM (Connectivity Fault Management) MA/MEG and MEP on both adjacent units.
The following tables show a configuration summary for all MEG1 and MEG2, and MEP1 and MEP2 units.

Configuration Summary of the MEG1

Configuration Summary of the MEG2

Configuration Summary of the MEP1

Configuration Summary of the MEP2
| Parameter | Unit 1 | Unit 2 | Unit 4 |
| --- | --- | --- | --- | --- |
| MA/MEG | Link 4 to 1| Link 1 to 2 | Link 2 to 4 |
| MEPID | 2 | 2 | 2 |
| Port | SFP-2-1 | SFP-2-1 | SFP-2-1 |
|MEP Name | Link 4 to 1 | Link 1 to 2 | Link 2 to 4 |
|Active | Enable | Enable | Enable |
| CCI enable | Enable | Enable | Enable |
| Primary VID| 0 | 0 | 0|

Verify connectivity
Verify Ethernet connectivity on the new unit. Verify also that all MEPs are active on the new unit and on both adjacent units (Units 2 and 4).

Clear the forced switch
Finally, clear the forced switch on both Units 2 and Unit 4 to re-enable ERP protection in the complete ring.

For each unit (2 and 4):

1. Access the page Protection ▶ ERP ▶ Status.

2. Click the ERP Index of the ring.

3. Click the Clear button to clear the forced switch.

Additional Considerations

After you have successfully configured and deployed your Ethernet ring, you may wish to perform CFM between sites. For this purpose, you will need to add new CFM instances and associate them to the LAG ports.

However, you must be aware of the following constrains: in order to allow VLAN tag CFM instances on the LAG ports, you must not use VLAN tag in your Port CFM instances. In this application note, the VLAN tags were not used for the CFM instances in order to provide you with the choice of VLAN tag, or not, for the CFM associated with the LAG ports. In other words, it is recommended that the ERP be build using untagged CFM instances, as presented here.

Also, when adding the “site to site” CFM instances, you need to ensure that you use a higher Y.1731 maintenance domain level for the MEGS meant for the LAG ports. In this application note, all the ERP CFM instances were all created with a domain level of 0 in order to allow maximum flexibility into the choices of level domains for the site to site
CFM.