Media Redundancy Protocol (MRP), defined in International Electrotechnical Commission (IEC) standard 62439-2, provides fast convergence in a ring network topology for Industrial Automation networks. MRP Media Redundancy Manager (MRM) defines its maximum recovery times for a ring in the following range: 200 ms and 500 ms.

Note	The default maximum recovery time on the Cisco IE switch is 200 ms for a ring composed of up to 50 nodes. You can configure the switch to use the 500 ms recovery time profile as described in Configure MRP Auto-Manager. The 10 ms and 30 ms recovery time profiles are not supported.

MRP operates at the MAC layer and is commonly used in conjunction with the PROFINET standard for industrial networking in manufacturing.

MRP is supported on the switches.

MRP CLI mode is managed by the Cisco IOS XE CLI and WebUI, a web-based user interface (UI).

Note	When managing the switch in MRP CLI mode, you cannot download the MRP configuration from Siemens STEP7/TIA.

In an MRP ring, the MRM serves as the ring manager, while the Media Redundancy Clients (MRCs) act as member nodes of the ring. Each node (MRM or MRC) has a pair of ports to participate in the ring. The MRM initiates and controls the ring topology to react to network faults by sending control frames on one ring port over the ring and receiving them from the ring over its other ring port, and conversely in the other direction. An MRC reacts to received reconfiguration frames from the MRM and can detect and signal link changes on its ring ports.

On the switch, certain nodes or all nodes in the ring can also be configured to start as a Media Redundancy Automanager (MRA). MRAs select one MRM among each other by using a voting protocol and a configured priority value. The remaining MRAs transition to the MRC role.

All MRM and MRC ring ports support the following states:

• Disabled: Ring ports drop all received frames.

• Blocked: Ring ports drop all received frames except MRP control frames and some standard frames, for example, LLDP.

• Forwarding: Ring ports forward all received frames.

• Not Connected: The link is physically down or disconnected. (This state differs from the Disabled state, in which the MRP Port is manually disabled through software.)

During normal operation, the network operates in the Ring-Closed state (see figure below). To prevent a loop, one of the MRM ring ports is blocked, while the other port is forwarding. Most of the time, both ring ports of all MRCs are in the forwarding state. With this loop avoidance, the physical ring topology becomes a logical stub topology.

In the figure, note the following details about the two rings, left and right:

• Left Ring: The connection (small blue square, top) on the MRM is in a blocked state (as shown by the two parallel lines) because no ports are disconnected.

• Right Ring: Two MRC connections (left and center small white squares) are in the disabled state because the link between them is broken, as marked by a red “x”.

If a network failure occurs:

• The network shifts into the Ring-Open state.

• In the case of failure of a link connecting two MRCs, both ring ports of the MRM change to the forwarding state, the MRCs adjacent to the failure have a disabled and a forwarding ring port, and the other MRCs have both ring ports forwarding.

In the Ring-Open state, the network logical topology becomes a stub.

Layer 2 Ethernet frames will be lost during the time required for the transition between these two ring states. The MRP protocol defines the procedures to automatically manage the switchover to minimize the switchover time. A recovery time profile, composed of various parameters, drives the MRP topology convergence performance. The 200 ms profile supports a maximum recovery time of 200 ms.

MRP uses three types of control frames:

• To monitor the ring status, MRM regularly sends test frames on both ring ports.

• When MRM detects failure or recovery, it sends TopoChange frames on both ring ports.

• When MRC detects failure or recovery on a local port, it sends LinkChange subtype frames, Linkdown and Linkup, to the MRM.

If configured to start as a Media Redundancy Automanager (MRA), the node or nodes select an MRM using a voting protocol and configured priority value. The remaining MRAs transition to the MRC role. All nodes must be configured as MRA. A manually configured MRM and MRA in the same ring is not supported.

Note	You can activate MRA through the CLI. See the section Configuring MRP CLI Mode in this guide. Although MRAs transition to the MRC role after an MRM is selected, you cannot explicitly configure an MRC.

The MRA role is not an operational MRP role like MRM or MRC. It is only an administrative, temporary role at device startup, and a node must transition to the MRM role or the MRC role after startup and the MRM is selected though the manager voting process.

MRA functions as follows:

1. At power on, all MRAs begin the manager voting process. Each MRA begins to send MRP_Test frames on both ring ports. The MRP_Test frame contains the MRA's priority value. The remote manager's priority value contained in the received MRP_Test frames are compared with the MRA's own priority. If its own priority is higher than the received priority, the MRA sends a negative test manager acknowledgment (MRP_TestMgrNAck) frame, along with the remote manager's MAC address.

2. If the receiving MRA receives an MRP_TestMgrNAck with its own MAC address, the receiving MRA initiates the transition into the client (MRC) role.

3. The MRP_TestPropagate frame informs other MRA devices in the client role about the role change and the new higher priority manager. The clients receiving this frame update their higher priority manager information accordingly. This ensures that clients remain in the client role if the monitored higher priority manager role changes.

You do not need a feature license to use MRP with Switches. MRP works with either base license—Network Essentials or Network Advantage.

In an Industrial Ethernet network, an MRP ring in a cell/area is a sub-ring of the access layer. You can connect multiple MRP rings, which you can then aggregate into the distribution layer.

You can configure up to 3 rings. The MRP switch can be configured only as an auto-manager.

MRP works with Spanning Tree Protocol (STP) to prevent unwanted broadcast loops in the event that a user accidentally connects a device that does not participate in the MRP ring. In a network operating with MRP and STP, spanning tree bridge protocol data units (BPDUs) are not sent on MRP-enabled ports. If ports are unconfigured from an MRP ring, then the ports are added to the spanning tree.

MRP-STP interoperability is supported in MRP CLI mode, and functions without additional CLI configuration.

• Because MRP is deployed in a physical Ring topology, before configuring or unconfiguring the MRP feature, it is advised to leave one physical connection between two nodes in each ring open by either issuing a shut command on the connecting interfaces or physically removing the cable to avoid any network storms. After you have properly configured all MRMs, issue a no shut command on the port or re-connect the cable between the nodes.

• MRP is disabled by default; MRP CLI is the default mode when MRP is enabled.

• The default VLAN is 1.

  Note	Create the non-default VLAN before you assign it to MRP ring 1.

To configure MRP, configure the node as MRA and specify the two MRP ports. You can configure up to 3 rings on the device (the device can be manager or client) with a manager instance for each ring and one manager per device.

The following MRP configuration parameters are optional:

• domain-id: A unique ID that represents the MRP ring.

• domain-name: Logical name of the configured MRP domain-ID.

• profile: 200 ms (the default)

• vlan-id: VLAN for sending MRP frames.

The following example shows the MRP switch configured as automanager:

Switch#configure terminal
Switch# no profinet mrp
Enter configuration commands, one per line. End with CNTL/Z.
Switch(config)#mrp ring 1
Switch(config-mrp)#mode auto-manager
Switch(config-mrp-auto-manager)#priority 36864
Switch(config-mrp-auto-manager)#end
Switch#configure terminal
Enter configuration commands, one per line. End with CNTL/Z.
Switch(config)#interface gi1/1
Switch(config-if)#switchport mode trunk
Switch(config-if)#mrp ring 1
WARNING% Enabling MRP automatically set STP FORWARDING. It is recommended to shutdown all interfaces which are not currently in use to prevent potential bridging loops.
Switch(config-if)#exit
Switch(config)#interface gi1/1
Switch(config-if)#switchport mode trunk
Switch(config-if)#mrp ring 1
WARNING% Enabling MRP automatically set STP FORWARDING. It is recommended to shutdown all interfaces which are not currently in use to prevent potential bridging loops.
Switch(config-if)#end
 
Switch#show mrp ring
MRP ring 1

Profile     : 200 ms
Mode        : Auto-Manager
Priority    : 36864
Operational Mode: Manager
From        : CLI
License     : Active
Best Manager MAC Address    :84:B8:02:ED:E8:01     priority 36864     

Network Topology: Ring
Network Status  : OPEN
Port1:                                     Port2:
 MAC Address    :84:B8:02:ED:E8:02          MAC Address    :84:B8:02:ED:E8:01
 Interface      :GigabitEthernet1/1      Interface      :GigabitEthernet1/1
 Status         :Forwarding                 Status         :Forwarding

VLAN ID     : 1
Domain Name : Cisco MRP Ring 1
Domain ID   : FFFFFFFF-FFFF-FFFF-FFFF-FFFFFFFFFFFF

Topology Change Request Interval        : 10ms
Topology Change Repeat Count            : 3
Short Test Frame Interval               : 10ms
Default Test Frame Interval             : 20ms
Test Monitoring Interval Count          : 3
Test Monitoring Extended Interval Count : N/A

          
Topology Change Request Interval        : 10ms
Topology Change Repeat Count            : 3
Short Test Frame Interval               : 10ms
Default Test Frame Interval             : 20ms
Test Monitoring Interval Count          : 3
Test Monitoring Extended Interval Count : N/A