Siemens SCALANCE MRP Ring Redundancy Switch Selection

Overview of Industrial Ethernet Ring Redundancy

Ring topology is widely adopted in industrial automation because it provides a single deterministic reconfiguration path when a link or switch fails. Unlike star topologies where a single cable break isolates downstream devices, a ring interconnects every switch such that traffic can travel in two directions. A redundancy protocol detects a break and instructs the ring to "close" by activating a previously blocked port, restoring connectivity within a bounded time. The general concept is described in the Ring network entry and in vendor guides such as the Schneider redundancy backbone article.

Industrial Ethernet protocols — PROFINET, EtherNet/IP, and Modbus TCP — are tolerant of cable breaks, but only when the recovery time is shorter than the application's watchdog. For PROFINET, the typical update time of 1–10 ms requires a reconfiguration time well under 200 ms. This requirement is the primary reason unmanaged switches are excluded from redundant rings: they cannot signal the break and cannot close the loop, so a single break partitions the ring into isolated segments. The recovery-time discipline is also discussed in HMS Networks' Ethernet Ring Redundancy technical note.

Ring Redundancy Protocols: MRP, HRP, HSR, and PRP

Four redundancy mechanisms dominate industrial networking. Each has a defined reconfiguration time and topology.

For most plant-floor installations, MRP remains the baseline. It specifies a single Redundancy Manager (RM) per ring; all other switches operate as MRP Clients that forward test frames on both ring ports but do not block traffic. MRP test frames use EtherType 0x88E3 with default destination MAC 01:15:4E:00:00:01; the RM sends these on its primary ring port and expects to see them reappear on its secondary ring port within the configured test interval (default 20 ms).

HRP is Siemens' legacy ring protocol. It is functionally similar to MRP but supports Standby Manager coupling between redundant rings. New deployments should use MRP for interoperability with non-Siemens devices and for compliance with IEC 62439-2.

HSR and PRP are zero-loss protocols used in substation automation (IEC 61850) and motion-safety applications. They require every node to be doubly attached, which doubles port count and cabling. They are outside the scope of typical PROFINET cells but are available on the SCALANCE XR-500 family and the IEC 62439-3 standard.

SCALANCE X Switch Family and Ring Capability Matrix

Siemens categorizes SCALANCE switches by ring capability. The matrix below summarizes the most common families; firmware versions and exact article numbers are documented in the SCALANCE X-200 Operating Instructions and the Siemens IA&DT catalog IK PI.

When planning a new ring or refurbishing an existing one, choose the lowest-cost managed switch that supports the required PROFINET conformance class. The X-200 family is the workhorse for cabinets and field segments. The XF-200 and X204-2 IRT PRO target cabinet-less installations with IP65 sealing.

SCALANCE X005 Limitations in Redundant Rings

The SCALANCE X005 and X005TS are unmanaged, plug-and-play 5-port switches with no configuration interface. The Operating Instructions explicitly state, in section 2.1, that the X005 cannot be used in a redundant ring because it does not support the redundancy function. Siemens publishes this restriction across all firmware variants and has not relaxed it across the product life cycle.

The technical reason is straightforward: an MRP Client must pass MRP test frames (EtherType 0x88E3, destination MAC 01:15:4E:00:00:01) through both ring ports, detect their absence on a link failure, and report link-down/up events. The X005 has no firmware hook to do this — it forwards frames by MAC table only and ignores frame EtherType filtering rules. Inserting an X005 between two managed switches breaks the test-frame loop and prevents the RM from ever detecting a fault, leaving the ring permanently in a partitioned state when a single break occurs.

Field reports sometimes describe X005 rings that "appear to work" because they are effectively passive bus segments (a ring of stars), but the moment a single link breaks, isolated devices go offline. This is not a redundant ring — it is a degraded star with shared cabling risk.

SCALANCE X204-2 IRT PRO vs X208 as Redundancy Manager

Both switches can serve as Ring Manager. The choice depends on the application, not on MRP capability, because both fully implement IEC 62439-2 RM functionality.

If the ring carries PROFINET IRT traffic (e.g., SINAMICS drives with isochronous mode, SIMOTION, or ET 200SP motor starters requiring deterministic update), the X204-2 IRT PRO is justified because IRT frames must be cut-through forwarded. If the ring is purely standard PROFINET RT or TCP/IP, the X208 provides the same MRP functionality at lower cost and with more ports.

Ring Topology Design Rules and Constraints

The following rules apply to any MRP ring on SCALANCE hardware. They are derived from the SCALANCE Operating Instructions and the IEC 62439-2 conformance test plan.

1. Exactly one Redundancy Manager (RM). Two RMs in the same ring will generate contradictory test-frame results and the ring will thrash between "closed" and "open" states every test interval.
2. All other switches in the ring must be MRP Clients. They pass MRP test frames transparently and forward configuration changes; they do not make blocking decisions.
3. Maximum of 50 switches per ring (Siemens specification; smaller rings reconfigure faster because fewer switches have to forward the topology-change propagation).
4. Maximum 50 devices in standby coupling across two rings.
5. Ring ports must be physically identified in each switch's configuration. By default, the two lowest-numbered gigabit ports are used; this can be overridden in WBM under Layer 2 > Ring Redundancy > Port assignment.
6. Mixing HRP and MRP in the same ring is forbidden. Both protocols use the same EtherType but have incompatible state machines.
7. Ring ports must match speed and duplex. A 100 Mbit/s port connected to a 1 Gbit/s port will fall back to 100 Mbit/s but must be configured for full duplex to avoid late collisions. Disable autonegotiation on ring ports when both ends are known.
8. Do not enable Spanning Tree (STP/RSTP) on ring ports. STP and MRP conflict because both protocols manipulate the forwarding state of the same port. Disable RSTP on the switch if the only role for STP would be on ring ports.
9. Avoid daisy-chained hubs or media converters inside the ring. They extend the link-down detection window and break deterministic reconfiguration.
10. Verify maximum cable length. Copper (10/100): 100 m segment. Fiber: 2 km (multi-mode 100 Mbit/s SC) up to 70 km (single-mode 1 Gbit/s LX). Longer runs increase link-down detection time.

Multi-Ring Topologies: Standby Coupling and Dual Homing

Larger plants use multiple rings coupled together. Siemens supports two patterns:

• Standby Coupling (HRP): two rings share a redundant link through two switches acting as Standby Master and Standby Slave. The standby link is normally blocked; one of the two rings can fail and the standby opens. Reconfiguration time is comparable to MRP (≤ 300 ms). Configure under Layer 2 > Standby Coupling on X-300 / X-400 switches.
• Dual Homing (MRP): a device with two ports attaches to two different rings (typically a controller or a redundant gateway). This provides path redundancy beyond a single ring. The controller's PROFINET stack must support AR (Application Relationship) to both rings.

When designing multi-ring networks, keep the diameter (number of switches between any two devices) under 50 switches and avoid crossing the same physical cable tray through both rings — a single physical incident (fire, fork-lift) could sever both rings.

PROFINET Stack Classes and Reconfiguration

PROFINET devices are conformance-tested against classes that determine how they behave during network changes:

With MRPD, the PROFINET controller re-routes the IRT frames around the broken segment using the recovered ring path. The IO device never sees an AR abort; the visible effect is a single dropped sample. This is why modern SIMATIC installations pair an X-200 IRT ring with an S7-1500 CPU and ET 200SP HF IO.

Configuring MRP in Web Based Management (WBM)

The WBM is accessible at the switch's IP address on port 80/443 (HTTPS recommended). Default credentials are admin / admin on legacy firmware or admin / (initial password printed on device label) on firmware 4.x and later.

Step 1: Assign IP and PROFINET name

Navigate to System > General > Device and configure:

IP address:    192.168.0.100
Subnet mask:   255.255.255.0
PROFINET name: rm-switch-1

The PROFINET name must be unique within the cell. Use a naming convention such as <cell>-rm-<number> for the manager and <cell>-sw-<number> for clients.

Step 2: Enable MRP and designate as Manager

Navigate to Layer 2 > Ring Redundancy. Set the role of the switch to Manager. Choose the ring ports (default Port 1 + Port 2). For SCALANCE X204-2 IRT PRO, ring ports are typically Port 1 and Port 2 (RJ45) — the fiber ports can serve as uplink to a control room. For X208, use Port 7 and Port 8 to free up Port 1–6 for IO devices and HMIs.

Step 3: Configure clients

On every other switch in the ring, navigate to Layer 2 > Ring Redundancy and set role to Client. Assign the same two physical ports as ring ports. Confirm that MRP is selected (not HRP).

Step 4: Verify topology

From the RM, open Information > Ring Redundancy. The current ring state should report Ring closed with one of the two ring ports in Blocking state. If both ports show Forwarding, the RM has not received its own test frames — check that the ring is physically a closed loop and that no client is incorrectly set as RM.

Configuring MRP via TIA Portal / STEP 7

TIA Portal and STEP 7 V5.5+ provide a graphical ring editor that pushes the configuration to every device via PROFINET DCP. This is the preferred path for new installations because it bundles ring, IO, and HMI configuration into one project.

1. Open the project, switch to Devices & Networks.
2. Drag SCALANCE switches from the catalog into the network view. Assign PROFINET names and IP addresses.
3. Right-click each switch > Properties > PROFINET interface > Ring redundancy.
4. Select Manager on one switch only. For all others, select Client.
5. Assign ring ports under Port assignment. TIA validates that both ring ports are members of the same MRP domain.
6. Compile and download to all devices.
7. Observe online diagnostics: Online & diagnostics > Ring redundancy > Status.

For older STEP 7 installations without TIA, configure via the WBM or use the SIMATIC Automation Tool for batch deployment across hundreds of switches.

Integrating CP443-1 as MRP Client

The S7-400 communications processor CP443-1 (6GK7443-1EX11-0XE0 with firmware V3.x or newer) supports MRP Client mode. When the CP is in MRP Client mode, two of its PROFINET ports are assigned as ring ports and the S7-400 participates directly in the ring rather than sitting behind a single switch. This reduces switch count and provides a second path between the PLC and IO devices.

Configuration steps:

1. In HW Config (STEP 7), open CP443-1 properties > PROFINET interface > Operating mode.
2. Set MRP Client and assign ring ports (typically Port 1 and Port 2).
3. Ensure the CP's MAC addresses are unique; the CP does not perform blocking but must transparently forward MRP test frames.
4. Verify in the S7-400 diagnostic buffer: the event Ring redundancy state change appears with new state "Ring closed" after each link event.
5. If the CP443-1 is replaced with a CP443-1 Advanced (6GK7443-1GX30-0XE0), the same procedure applies but with a higher port count (four instead of two) and additional security features.

Benefits: reduces the number of managed switches in the ring (cost saving) and centralizes the PLC on the same physical ring as the IO devices. Caveats: the CP cannot be the Ring Manager, so a separate managed switch must serve as RM. The CP443-1 also reserves one port for PROFINET IO, leaving only one spare for non-ring use; plan accordingly.

Verification and Diagnostics

After configuration, perform the following checks. All are non-disruptive except the final cable-pull, which should be coordinated with operations.

To measure reconfiguration time precisely, use the PROFINET controller's diagnostic buffer. With an S7-1500, the event Network fault (IO device) and the corresponding Network fault resolved timestamps are logged with millisecond resolution. Subtract the two; the difference is the ring reconfiguration time. Validate against the IEC 62439-2 test report values shipped with the switch.

Reconfiguration Time Calculation

MRP recovery time is the sum of three components. Engineers sizing watchdog timers should add a 50% margin for stack variance and broadcast handling.

• Link-down detection time: < 10 ms for fiber (Loss of Light, no autonegotiation), < 50 ms for copper (PHY autonegotiation pause and link pulse).
• RM reaction time: RM detects missing test frames and unblocks its secondary port. Typical 10–25 ms. Configurable via WBM under Layer 2 > Ring Redundancy > Test interval.
• PROFINET stack restart: On Class C controllers (S7-1500, ET 200SP HF) using MRPD, this is zero. On Class A/B devices, the controller re-issues AR (Application Relationship) which can add 50–100 ms.

Total worst-case MRP reconfiguration: ≈ 200 ms. Best case with fiber and IRT-aware stack: ≈ 25 ms. These figures are bounded by IEC 62439-2 and verified by Siemens test reports.

Compare to RSTP (Rapid Spanning Tree, IEEE 802.1w), which typically converges in 1–3 seconds because each switch recomputes its own root port. RSTP is acceptable for office networks and non-deterministic HMI traffic; MRP is required when PROFINET IO update times drop below 250 ms or when process safety (PROFIsafe) requires deterministic fault response.

SNMP Monitoring and Fault Traps

SCALANCE switches expose MRP status through standard MIB-II objects and Siemens-specific enterprise MIBs. Configure SNMPv3 traps to forward ring state changes to a central NMS (e.g., SINEC NMS, PRTG, or Nagios).

Engine ID:       0x800072810500abcd
Trap destination: 192.168.0.250:162
Trap OID:         1.3.6.1.4.1.4329.6.3.1.1.3 (rmStateChange)
Trap OID:         1.3.6.1.4.1.4329.6.3.1.1.4 (linkDown)
Trap OID:         1.3.6.1.4.1.4329.6.3.1.1.5 (linkUp)

Trap reception at the NMS should be acknowledged within 30 seconds; a missed trap indicates either NMS downtime or a network segment that has dropped from the ring path. For safety-critical installations, integrate ring traps into a SCADA alarm channel so that operators see "Ring topology degraded" as a banner event.

Migration from HRP to MRP

Plants installed before 2010 often run HRP. Migrating to MRP requires careful staging because HRP and MRP cannot coexist in the same ring.

1. Document the current HRP ring topology, including standby coupling and HRP manager switch.
2. Identify the new MRP ring manager. The same physical switch can be reused if it supports both protocols.
3. Schedule a maintenance window. Both HRP and MRP require a single-ring structure; transitioning one switch at a time will cause the ring to thrash.
4. During the window, change all switches to MRP Client first, then change the designated switch to MRP Manager last.
5. Verify ring closure with WBM and PROFINET diagnostics before releasing the system to operations.
6. If standby coupling exists, retain HRP for the coupled rings until the controller rings are validated, then migrate standby coupling to MRP (Siemens added MRP-based standby on firmware 4.x and later for X-300/X-400).

Cable, Media, and Grounding Considerations

Ring topology is sensitive to physical layer faults because every link is critical. Apply the following cabling rules:

• Use shielded twisted pair (SF/UTP or S/FTP) for copper ring segments. Industrial Cat 6A or Cat 7 cable rated for the installation environment (PVC, PUR, FRNC jacket for drag chains).
• Ground both ends of the shield. In Siemens practice, ring cables use equipotential bonding to a ground bar at each cabinet; the shield is bonded via the M12 or RJ45 connector's drain wire.
• Avoid routing ring segments through the same cable tray. A tray fire or fork-lift strike could sever the entire ring. Place redundant ring cables in physically separated paths.
• Fiber segments must use matched SFP pairs. Mismatched wavelength (e.g., 1310 nm on one end and 1550 nm on the other) will appear as link-down with no diagnostic value. Siemens SFPs (6GK5 991-1Ax00-xAAx) are vendor-locked by firmware on X-300 and XR-500; use Siemens-branded SFPs on those families.
• Bend radius. Maintain ≥ 4× cable diameter for copper and ≥ 10× for fiber. Tight bends degrade bit error rate long before they cause link-down.
• Segment length. Copper ≤ 100 m. Fiber ≤ 2 km (multi-mode 100 Mbit/s SC), ≤ 70 km (single-mode 1 Gbit/s LX). Longer runs degrade link-down detection time.

Troubleshooting Matrix

Frequently Asked Questions