Allied Telesis AT-8000S User Manual

Page 31

Most difficulties incurred in Ring topologies occur when a device in the ring becomes non-functional, or a link is 
severed. In a stack, the system automatically switches to a Stacking Failover topology without any system 
downtime. An SNMP message is automatically generated, but no stack management action is required. However, 
the stacking link or stacking member must be repaired to ensure the stacking integrity.

After the stacking issues are resolved, the device can be reconnected to the stack without interruption, and the 
Ring topology is restored.

If a failure occurs in the stacking topology, the stack reverts to Chain Stacking Topology. In the Chain topology, 
devices operate in a chain formation. The Stacking Master determines where the packets are sent. Each unit is 
connected to two neighboring devices, except for the top and bottom units.

Stacking Unit IDs are essential to the stacking configuration. The stacking operation is determined during the boot 
process. The Operation Mode is determined by the Unit ID selected during the initialization process. For example, 
if the user selected stand-alone mode, the device boots as a stand-alone device.

The device units are shipped with the default Unit ID of the stand-alone unit. If the device is operating as a 
standalone unit, all stacking LEDs are off. Once the user selects a different Unit ID, the default Unit ID not erased, 
and remains valid, even if the unit is reset.

Unit ID 1 and Unit ID 2 are reserved for Master enabled units. Unit IDs 3 to 6 can be defined for stack members.

When the Master unit boots or when inserting or removing a stack member, the Master unit initiates a stacking 
discovering process.

Note

If two members are discovered with the same Unit ID the stack continues to function, however only the 
unit with the older join time joins the stack. A message is sent to the user, notifying that a unit failed to join 
the stack.