PRP and HSR deliver every frame twice, so a broken LAN, a cross-patched cable or a node that is not removing duplicates stays invisible to the applications — until the second path fails too. RSTP rings hide their problems in the same way until a root change or a flapping port floods the network. A capture shows these hidden faults directly.
Standards followed
IEC 62439-3 (PRP and HSR: redundancy control trailer, HSR tag, supervision frames, LifeCheckInterval)IEEE 802.1Q (RSTP: BPDU format, root bridge, topology change)IEC 61850-90-4 (network engineering guidelines for substations)
What it can do
- PRP: frames per node with their LAN identifier; frames tagged LAN A and LAN B at one capture point (LANs joined); a node's frames all on the wrong LAN (A/B swapped); gaps in each node's sequence numbers on this LAN; LSDU size that does not match the frame; supervision frames, their interval, RedBoxes and the single-attached nodes they represent; nodes without supervision.
- HSR: frames per node, the same sequence number passing the port twice (circulating frame), sequence gaps, LSDU size.
- RSTP: root bridge and its changes, topology-change flags, legacy STP, devices that send BPDUs claiming to be root (an IED with bridging enabled), BPDUs from devices not in your switch list.
Example problem
The sample capture on PRP LAN A of the S13 station bus, with RSTP between the switches.
- GW1's frames all carry LAN B: its LAN B port is patched into the LAN A switch.
- P87T_T1 loses 5 frames on LAN A; BCU_J01 sends one frame with a wrong LSDU size and no supervision frames.
- The root moves from SW1 to SW2 (priority set to 4096 by mistake), and BCU_J01's port sends its own BPDUs.
Five hidden faults that PRP and RSTP were masking, each with the device to look at.
Capture on one LAN (A or B) at a mirror port of a switch; to compare A and B, capture both. A capture point sees only the traffic forwarded to it — mirror an uplink to see more nodes.