What is it?
Media redundancy is primarily used to avoid single points of failure in industrial communication networks. If a failure occurs on a redundant structure, the network falls back to a secondary state in which communication is still viable, and repair can be made to restore the system to the previous fault-free state. Ethernet technology does not allow physical loops, as they cause packets to circulate endlessly and overload the network. This means providing media redundancy within an Ethernet network requires the use of a protocol that is able to monitor and resolve the physical loops introduced by redundant pathways. This protocol must ensure that, even with multiple physical pathways to any device, only one is activated at any one time and the remaining are in standby mode. This is achieved by: monitoring links, detecting interruptions, and switching to an alternative path in the event of failure as soon as possible. Media Redundancy Protocol (MRP) is a protocol for providing redundancy in Ethernet networks via a ring. MRP is specified for ring networks with up to 50 devices. It guarantees fully predictable switchover behavior.
Important concepts
- MRP node: A switch contained within a ring network monitored by MRP.
- Redundancy Domain: A ring controlled by MRP.
- Media Redundancy Manager (MRM): A single MRP node that monitors and controls the ring network and reacts to any failure.
- Media Redundancy Client (MRC): Any node within the ring that is not an MRM.
- Ring port: The two designated ports on each MRP node that are physically connected to the ring topology.
How it works
Ring ports can be in one of three states:
- Disabled: all packets received by the port are dropped.
- Blocking: all packets received by the port are dropped, with the exception of MRP protocol packets.
- Forwarding: all packets received by the port are forwarded.
The ring itself can be in one of three states:
- Closed: both ports on all nodes are linked up, one port on the MRM is blocking, all others are forwarding.
- Open: both ports on the MRM are forwarding, either because one of the MRCs has signalled a port down, or the MRM failed to receive its own test frames.
- Undefined: this is set before the actual state of the ring is known.
During normal operation, one of the MRM’s ring ports is blocking, while the other is forwarding. This puts the network in a closed state, where the physical ring topology is in a ring structure, and at the logical level is in a linear structure. The MRM sends MRP_Test frames (at a predetermined interval) out both its ring ports. These will be forwarded around the ring by the MRCs until they again reach the MRM.
In the case of a transmission failure (for example, a ring port becomes disabled), the MRM will fail to receive its test frames. After a predefined number of test packets are lost, it will set its blocking ring port to forwarding. This enables the secondary network path, and puts the network in an open state where both MRM ring ports are forwarding.
About ATTEST-CTS MRP
Overview
Veryx ATTEST™-CTS Media Redundancy Protocol (MRP) automated test suite provides Equipment Manufacturers, an easy and efficient solution for the verification of MRP in Industrial Ethernet Switches. ATTEST enables significant speeding up of testing cycles and reduces the “time-to-market”.
Veryx has devised about 111 test cases that comprehensively test for MRP compliance. These test cases have been grouped into 8 convenient test groups based on the IEC specifications for each category of functions.
ATTEST MRP test cases verify the MRP support for:
- Manager : MRP switch connected with other MRP clients in a ring
- Client: MRP switch connected with other MRP manager and other MRP clients in a ring.
A typical industrial Ethernet network showing a ring with MRP Manager, Client IEC 62439: Industrial communication networks: high availability automation networks. – 65c, 62439/Ed 1.0, 2007-04-06
Functionalities covered:
Applicant state machine
MRP Manager - Ring Port states
- MRP Manager Protocol State Machine
- MRP Manager with multiple clients
- MRP Clients - Ring port states
- MRP Client Protocol State Machine
- MRP Clients with MRM and other MRCs
- Multiple MRP managers in a Ring
- MRP Frame Format verification