Default convergence: < 200 ms
Hirschmann switches converge in 30 to 80 ms with default parameters.
11.3 MRP Timing and Convergence
The previous chapter covered MRP frame types and Python tools for monitoring ring health. Knowing what frames look like on the wire is essential for diagnosis. Knowing how timing parameters interact is essential for optimization.
Why Timing Matters
Section titled “Why Timing Matters”MRP convergence time is the interval between a failure occurring and traffic being restored. A convergence time of 200 ms keeps most PLCs running. A convergence time of 30 ms keeps PROFINET IRT applications running. The difference comes down to three configurable parameters.
Timing Parameters
Section titled “Timing Parameters”| Parameter | Default | Description |
|---|---|---|
| Test interval | 20 ms | Interval between MRP_Test frames |
| Test retries | 3 | Missed tests before declaring failure |
| Detection time | 60 ms | Test interval x test retries |
| Total convergence | ~200 ms | Detection + MAC flush + forwarding |
Convergence Timeline
Section titled “Convergence Timeline”Hirschmann switches typically converge in 30 to 80 ms with default parameters.
Fast Timing Configuration
Section titled “Fast Timing Configuration”For applications requiring faster recovery, reduce the test interval:
| Configuration | Test Interval | Retries | Detection Time | Total Convergence |
|---|---|---|---|---|
| Default | 20 ms | 3 | 60 ms | ~200 ms |
| Fast | 10 ms | 3 | 30 ms | ~100 ms |
| Very fast | 10 ms | 1 | 10 ms | ~30 ms |
MRP_LinkDown Acceleration
Section titled “MRP_LinkDown Acceleration”When a ring port link goes down, the MRC sends an MRP_LinkDown frame immediately, without waiting for the test timeout. The MRM reacts in less than 10 ms.
Without MRP_LinkDown, the MRM waits for the test timeout: 20 ms x 3 = 60 ms. MRP_LinkDown is enabled by default on Hirschmann switches.
PROFINET and MRP Timing
Section titled “PROFINET and MRP Timing”| PROFINET Class | MRP Convergence Requirement |
|---|---|
| Class 1 (RT) | < 200 ms |
| Class 3 (IRT) | < 80 ms |
For PROFINET IRT applications, configure MRP with fast timing: 10 ms test interval, 3 retries.
Key Takeaways
Section titled “Key Takeaways”MRP_LinkDown gives < 10 ms reaction
MRP_LinkDown is enabled by default on Hirschmann switches. It bypasses the test timeout entirely.
What Comes Next
Section titled “What Comes Next”Timing parameters determine how fast MRP recovers. When recovery does not happen as expected, systematic troubleshooting is needed. The next chapter provides a diagnostic reference for the most common MRP problems: open rings, dual managers, VLAN mismatches, and flaky rings.
References
Section titled “References”- IEC 62439-2:2016 — Media Redundancy Protocol, Section 9 (Timing)
- Hirschmann. (2023). Application Note: MRP Timing Optimization. Belden/Hirschmann.
- PROFIBUS & PROFINET International. (2022). PROFINET System Description, Section 6 (Network Redundancy).