29.2.4 Switching communication paths
In order to switch to an alternate path when a ring fault is detected, a master node changes the status of the secondary port from Blocking to Forwarding. Likewise, in order to perform path switch-back after recovery from the ring fault is detected, the master node changes the secondary port from Forwarding to Blocking. Therefore, in order to promptly restore communication, the MAC address table entries are cleared for all nodes in the ring. If MAC address table entries are not cleared, because data frames are sent according to the information before switching (or switch-back), data might not be received properly. Therefore, to restore communication, the MAC address table entries for all nodes in a ring are cleared. The entries in MAC address table to be cleared follow the settings in the configuration command mac-clear-mode.
The following figure shows the switching operations for both master nodes and transit nodes.
|
|
![[Figure Data]](./GRAPHICS/ZU117260.GIF)
- <Structure of this section>
(1) Switching routes of the master node
When a ring fault is detected on the master node, Blocking is removed for the secondary port, It also clears MAC address table entries. Because of this, flooding occurs until MAC address learning is performed. MAC address learning is performed by sending and receiving frames over the secondary port, and switching is completed to a new path.
(2) Transit Node Route Switching
When a ring fault is detected on the master node, a control frame called a flush control frame is sent to other transit nodes in the ring of the same control VLAN, to request that MAC address table entries be cleared. When this flush control frame is received, the transit node clears MAC address table entry. Because of this, flooding occurs until MAC address learning is performed. MAC address learning is performed by sending and receiving frames on the new path, and communication path switching is completed.