What Is MRP?

Why Do We Need MRP?

Traditional Ethernet loop protection protocols, such as the Spanning Tree Protocol (STP), suffer from relatively slow convergence, often taking several seconds. As the network diameter increases, convergence time also rises. That makes these protocols unsuitable for large-scale industrial Ethernet networks, which require reliability and real-time responsiveness.

To address these limitations, MRP is introduced. MRP eliminates loops and mitigates broadcast storms in ring networks, while also providing node and link redundancy. In the event of a device or link failure, MRP can restore network connectivity quickly, meeting the stringent requirements of industrial applications. This not only ensures high availability but also enhances operational maintainability.

Typical Application Scenarios of MRP

As shown in the figure below, MRP can be deployed in industrial environments for higher reliability and easier maintenance. It provides node and link redundancy in a ring topology. When a single device or link failure occurs, the protocol can rapidly restore network connectivity.

In this example, DeviceA to DeviceD form MRP redundancy domain 1. DeviceA serves as the MRM for MRP redundancy domain 1, detecting loops and managing the domain.

MRP application scenario

How Does MRP Work?

MRP is deployed in ring networks composed of multiple devices, with each device running the MRP protocol.

MRP redundancy domain

MRP redundancy domain

On an MRP network, each device connects to exactly two neighboring devices, forming a closed-loop ring topology known as an MRP redundancy ring. Each MRP redundancy ring constitutes a redundancy domain.

Devices may be connected through multiple physical links, creating multiple ring topologies and consequently multiple redundancy domains. For example, connections established through Interface1 on each device form MRP redundancy domain 1, while connections through Interface2 form MRP redundancy domain 2. In such cases, each MRP redundancy ring is uniquely identified by an MRP redundancy domain UUID.

An MRP redundancy domain is identified by the redundancy domain name and UUID. The redundancy domain name takes effect locally, and the redundancy domain UUID is a unique identifier across the network.

MRP interconnection domain

MRP interconnection domain

On an MRP network, dual links can be deployed between two MRP redundancy rings to provide interconnection redundancy. However, as shown in the preceding figure, connecting MRP redundancy rings 1 and 2 with two links creates a new loop. To manage this loop, MRP introduces the concept of an interconnection domain, which consists of all devices forming the loop between the two MRP redundancy rings (DeviceC, DeviceD, DeviceE, and DeviceF in this example).

In an interconnection topology, devices can establish multiple MRP interconnection rings using different physical ports. For example, DeviceC, DeviceD, DeviceE, and DeviceF may constitute interconnection ring 1, while DeviceC, DeviceD, DeviceI, and DeviceJ form interconnection ring 2.

An MRP interconnection domain is identified by an interconnection domain name and ID. The interconnection domain name takes effect locally, and the interconnection domain ID is a unique identifier across the network.

Roles in an MRP redundancy domain

Each device participating in an MRP-enabled network must be assigned a specific role. Depending on the configuration, a single device may assume different roles across multiple MRP domains.

In an MRP redundancy domain, each device assumes one of the following roles:

• Media Redundancy Manager (MRM): This device monitors the redundancy ring topology and manages links. It blocks one of its ring ports to break a network loop when the ring is complete, and unblocks the previously disabled port to restore connectivity if a link failure occurs in the redundancy ring.
• Media Redundancy Client (MRC): All MRP-enabled devices in an MRP redundancy ring, other than the MRM, function as MRCs. They monitor their own ring ports and report link changes to the MRM, which then takes appropriate actions to maintain network connectivity.
• Media Redundancy Automanager (MRA): Any MRP-capable device can be configured to operate as either an MRM or MRC. However, only one device can function as the active MRM within an MRP redundancy ring at any given time. If this MRM fails, the ring no longer has centralized management, compromising the reliability of the MRP protocol. To address this, MRP introduces the MRA role. Devices configured as MRAs participate in an automatic election process to provide MRM redundancy. Upon system startup, all MRAs within the same MRP redundancy domain automatically initiate an election to select a single active MRM. The remaining MRAs then operate as MRCs. If the elected MRM fails, the remaining MRAs automatically re-elect a new MRM, enhancing protocol reliability.

  In an MRP redundancy domain, network administrators can configure an MRP-capable device as an MRM, MRA, or MRC. However, the MRP protocol ensures that only one MRM is active at any time.

Roles in an MRP interconnection domain

In an MRP interconnection domain, each device assumes one of the following roles:

• Media Redundancy Interconnection Manager (MIM): This device monitors the interconnection ring topology and manages links. It blocks one of its interconnection ports to break a network loop when the ring is complete, and unblocks the previously disabled port to restore connectivity if a link failure occurs in the interconnection ring.
• Media Redundancy Interconnection Client (MIC): All MRP-enabled devices in an MRP interconnection ring, other than the MIM, function as MICs. They monitor their own interconnection ports and report link changes to the MIM, which then takes appropriate actions to maintain network connectivity.

  In an MRP interconnection domain, network administrators can configure an MRP-capable device as an MIM or MIC. However, the MRP protocol ensures that only one MIM is active at any time.

  Roles assigned in an MRP interconnection domain are independent of those in an MRP redundancy domain. A single device may simultaneously serve as an MRC or MRM in a redundancy domain and as an MIC or MIM in an interconnection domain.

MRP ring port

Ports that connect MRP-enabled devices to their local MRP redundancy rings are referred to as MRP ring ports. Each device can have exactly two ring ports in an MRP redundancy domain. All other ports on the device are excluded from MRP operations and are used solely for connecting user terminals.

MRP ring ports support the following two states:

• Blocked: Ring ports in this state discard all the received frames except MRP frames and frames compliant with IEEE 802.1D.
• Forwarding: Ring ports in this state forward all the received frames.

  On an MRM, the ring port whose physical link state comes up first is designated as the primary port, while the port that comes up later functions as the secondary port.

MRP interconnection port

Ports that connect MRP-enabled devices to adjacent MRP redundancy rings are MRP interconnection ports. Each device can have only one interconnection port in an MRP interconnection domain.

MRP interconnection ports support the following two states:

• Blocked: Interconnection ports in this state discard all the received frames except MRP frames, frames compliant with IEEE 802.1D, and link detection frames compliant with IEEE 802.1Q.
• Forwarding: Interconnection ports in this state forward all the received frames.

MRP VLAN

In both MRP redundancy domains and MRP interconnection domains, all MRP frames are transmitted in designated VLANs. To ensure that MRP frames can be forwarded between redundancy and interconnection domains that are connected, the same VLAN must be used for MRP frames across these domains. Different MRP redundancy or interconnection domains may use different VLANs to carry MRP frames. Devices can also assume different roles in different domains, enabling flexible deployment and planning of MRP networks.