28.2.1 Network configuration

(1) Single ring configuration

The following figure shows a single ring configuration.

Figure 28-4: Single ring configuration

A single ring configuration consisting of one master node and multiple transit nodes is called a single ring configuration. The nodes in the ring are connected as ring ports by physical ports or link aggregations. Note that the same VLAN must be used as the control VLAN for all nodes in the ring, and a common VLAN must be used for data frame transfer. Control frames sent from the master node are circulated within the control VLAN. The VLANs used to send and receive data frames are aggregated into a single logical group called a VLAN group. VLAN groups can group multiple VLANs, and set a maximum of two groups for clockwise and counter-clockwise circulation in a single ring from the master node.

(2) Multi-ring configuration

Of the possible multi-ring configurations, the following figure shows one in which a single node is the contact point for the neighboring ring.

Figure 28-5: Multi-ring configuration

Each node in the ring runs as a single independent ring. Therefore, ring fault detection and recovery detection are performed independently by each ring.

(3) Multi-ring configuration with shared link

Of the possible multi-ring configurations, the following figure shows one in which multiple nodes are the contact points for the neighboring ring.

Figure 28-6: Multi-ring configuration with shared links

When multiple single rings are connected by multiple nodes, links are shared by multiple rings. These links are called shared links, and multi-ring configuration with these links is called a multi-ring configuration with shared links. On the other hand, when, as in (2), multiple single rings are connected by a single node, because no shared links exist, this is called a multi-ring configuration without shared links.

In a multi-ring configuration with shared links, when a common VLAN on a neighboring ring is used as a VLAN group for data transfer and a fault occurs for a shared link, the neighboring ring detects that a fault has occurred on each master node, and a loop spanning multiple rings (known as a super loop) occurs. Therefore, unlike a single ring configuration, this configuration requires that fault detection and switching operations be performed.

With the Ring Protocol, of the multiple rings for which shared links are a part of the ring, one ring is monitored for shared link faults and restoration (shared link monitoring ring), and the other rings are not monitored for shared link faults or restoration (shared link non-monitoring rings). Also, the nodes placed at both ends of a shared link are called terminal nodes (or shared nodes) in shared link non-monitoring rings. Here, because the monitored rings are unique within the master node for each ring, loops caused by faults between shared links can be prevented.

(4) Ring configuration with stacked nodes

The following figure shows a ring configuration that includes stacked nodes.

Figure 28-7: Ring configuration including stacked nodes

In a stack configuration, two member switches operate as one node. Stacked nodes connect as ring ports on the master switch and the backup switch, and configure the ring through the stack link. Note that stacked nodes do not support shared nodes in a multi-ring configuration with shared links.

Stacked nodes use stack links to transfer control and data frames. Therefore, set up two or more stack links to ensure sufficient bandwidth for ring traffic. For details about the maximum number of stacking ports, see 2.2 Number of lines that can be handled.