4.2 Ring Topologies

The previous section showed that RSTP converges in 1 to 2 seconds on a ring topology. For most industrial applications, that is too slow. A PLC with a 4 ms cycle time will fault during a 1-second outage. Ring-specific protocols like MRP were designed for exactly this topology and converge in under 200 ms. This section explains why rings are the dominant topology in industrial automation.

Why Rings Instead of Mesh

A mesh topology connects every switch to multiple other switches. It provides redundancy but requires complex cabling and a loop-prevention protocol (STP or RSTP) that takes 1 to 2 seconds to converge after a failure.

A ring topology connects switches in a closed loop. Each switch connects to exactly two neighbors. The cabling is simple: one cable to the left neighbor, one cable to the right neighbor. Ring-specific protocols (MRP, ERPS) converge in under 200 ms because they are optimized for this specific topology.

Industrial networks favor rings for three reasons:

1. Simple cabling — each switch needs exactly two ring ports
2. Fast convergence — ring protocols converge in under 200 ms; RSTP on a ring takes 1 to 2 seconds
3. Deterministic behavior — the failure and recovery behavior is predictable and testable

The ring operates differently under normal conditions versus during a failure.

Ring Operation

Under normal operation, one link in the ring is logically blocked to prevent loops. Traffic flows in one direction around the ring.

When a link fails, the Ring Manager detects the failure and unblocks its secondary port. Traffic now flows the other way around the ring.

Several ring protocols exist, each with different convergence times and use cases.

Ring Protocols

Protocol	Standard	Convergence	Primary Use
MRP	IEC 62439-2	under 200 ms	Factory automation (Hirschmann, Siemens)
ERPS	ITU-T G.8032	under 50 ms	Carrier-grade, multi-ring

MRP is the standard in Hirschmann/Belden networks. See Chapter 11 for a full treatment. Before enabling MRP, you need to verify that the ring is physically wired correctly.

Verifying Ring Topology via LLDP

After installing a new ring, verify that the switches are connected in the correct order and that no switch is missing from the ring. The following script reads LLDP neighbor information from each switch and shows the ring topology. Run it before enabling MRP:

# pip install netmiko
from netmiko import ConnectHandler
import re

SWITCHES = [
    {"host": "192.168.1.100", "name": "SW-Cell1-MRM"},
    {"host": "192.168.1.101", "name": "SW-Cell1-MRC1"},
    {"host": "192.168.1.102", "name": "SW-Cell1-MRC2"},
]

def get_ring_neighbors(host: str) -> list[dict]:
    conn = ConnectHandler(device_type="hirschmann_ssh", host=host,
                          username="admin", password="private")
    output = conn.send_command("show lldp neighbors")
    conn.disconnect()
    neighbors = []
    for line in output.splitlines():
        m = re.match(r"(\S+)\s+(\S+)\s+(\S+)", line)
        if m and "/" in m.group(1) and m.group(1) in ("1/1", "1/2"):
            neighbors.append({"local_port": m.group(1),
                               "neighbor":   m.group(2),
                               "remote_port": m.group(3)})
    return neighbors

print("Ring topology (LLDP neighbors on ring ports):")
for sw in SWITCHES:
    neighbors = get_ring_neighbors(sw["host"])
    print(f"\n{sw['name']}:")
    for n in neighbors:
        print(f"  {n['local_port']} -> {n['neighbor']} port {n['remote_port']}")

If the reconstructed topology does not match the expected ring order, a cable is connected to the wrong port or a switch is missing. Fix the cabling before enabling MRP.

Key Takeaways

Rings are the OT standard

Ring topology with MRP is the standard for industrial automation. Simple cabling, fast convergence, deterministic behavior.

Verify topology before enabling MRP

Use LLDP neighbor data to confirm the ring is wired correctly before enabling MRP. A miscabled ring will not converge correctly.

What Comes Next

Ring topologies provide redundancy with simple cabling. The next section covers LACP — link aggregation for uplinks that need more bandwidth than a single cable provides. After Part 1, Part 2 covers the network services (DHCP, DNS, NTP, NAT) that make IP networks self-configuring.

References

• IEC 62439-2:2016 — Industrial communication networks — High availability automation networks — Part 2: MRP
• ITU-T G.8032 — Ethernet Ring Protection Switching (2015)
• Hirschmann. (2023). User Manual — HiOS: MRP and Ring Redundancy. Belden/Hirschmann.