28.6.4 Operation when multi-faults occur

The following explains the operation when multi-faults occur due to shared link faults and other link faults in a shared link monitoring ring.

<Structure of this section>

(1) Operation in case of shared link failure

The following figure shows the operation when a fault occurs for shared links in a shared link monitoring ring.

Figure 28-33: Operation when a shared link fails

(a) Operation of each node of the shared link monitoring ring

A ring fault is detected by lack of HC(M) reception.: The master node can no longer receive HC(M)s from both directions, and detects a ring fault. The operation of the master node and the transit node when a ring failure is detected is the same as that of the multi-ring operation. For details, see section 28.4.2, Operation at Shared Link Failure and Recovery (1), Operation at Fault Detection.
Multi-fault monitoring frames cannot be received between shared links.: Shared nodes can no longer receive multi-fault monitoring frames between shared links, but because reception is still possible on the other ring port, multi-fault monitoring continues.

(b) Behavior of each node in the backup ring

HC(M)s sent by the master node can no longer be received on a backup ring, but because HC(S)s sent by shared nodes can be received, no fault operation is performed.

(2) Operation in case of multiple failures

The following figure shows operation when multi-faults occur due to shared link faults or other link faults within a shared link monitoring ring.

Figure 28-34: Operation when multiple errors occur

(a) Operation of each node of the shared link monitoring ring

A multi-fault is detected for the shared link monitoring ring.: Shared nodes can no longer receive multi-fault monitoring frames for both ring ports, and a multi-fault is detected.

(b) Behavior of each node in the backup ring

HC(S) sending is stopped.: The shared node detecting the multi-fault stops sending backup ring HC(S)s.

(3) Switching to Backup Ring

The following figure shows operation for switching to the backup ring due to multi-fault detection.

Figure 28-35: Switching to the backup ring

(a) Behavior of each node in the backup ring

A ring fault is detected due to no reception of HC(S)s.: The master node receives neither HC(M)s sent by itself from both directions nor HC(S)s sent by shared nodes, and detects a ring fault. The operation of the master node and transit node when a ring failure is detected is the same as that of the multi-ring operation. For details, refer to section 28.4.3, Operation of Non-Shared Link Monitoring Ring for Failures and Recovery (1) Operation when Failures are Detected.

(b) Operation of each node of the shared link monitoring ring

Flush control frames are sent from shared nodes.: When shared nodes receive a flush control frame from the master node of the backup ring, they send to the shared link monitoring ring only flush control frames that clear the MAC address table.
The MAC address table is cleared.: Transit nodes receive flush control frames sent from shared nodes, and clear the MAC address table.