Spanning Tree and Alpha Ring Protocols
As computer networks have become mission-critical assets for most, if not all, businesses, keeping the network up and running has assumed a crucial importance. Just as there are different types of traffic that run over a computer network, there are different solutions to keeping that traffic flowing,each with its own pros and cons.
Spanning Tree Protocol (STP)
The Spanning Tree Protocol (standardized as IEEE 802.1d) specifies a network design with redundant links to provide automatic backup paths if an active link fails. STP also avoids the creation of bridge loops that cause broadcast storms. Without STP, Ethernet switches with redundant links have no standardized way to keep from looping data over and over again to the other switches in the network, eventually disabling the network’s ability to pass data.
The idea behind a Spanning Tree topology is to enable switches to automatically discover a subset of the network topology that is loop-free, i.e., a tree. With STP turned on, the switches will perform the spanning tree algorithm when they are first connected, as well as any time there is a topology change, and automatically communicate with each other in a loop-free mode. Then, should a failure of one of the active links occur, STP unblocks the redundant links to enable the network to continue transmitting traffic.
The Alpha-Ring Protocol
The Alpha-Ring protocol is a proprietary protocol designed to provide a faster network recovery time after a failure than standard STP. As the name suggests, Alpha-Ring enables the switches to be organized in a ring arrangement. During normal operation, the backup path for the Alpha-Ring is blocked, and data follows the other links around the ring.
If, however, one of the active links fails, the Alpha-Ring protocol unblocks the backup path to enable data to keep flowing. Typical failover for Alpha-Ring protocol is less than 30 milliseconds.
In addition, unlike STP, Alpha-Ring does not operate using any bandwidth-consuming packets to check the ring status. The ring port connections are monitored by each switch individually without the need for test packets to be generated and transmitted around the ring.
Ethernet Ring Protocols
Although Ethernet is usually thought of as having a star or bus topology, it’s also possible to build an Ethernet network as a ring. This configuration has the advantage of providing a redundant pathway if a link goes down. A ring topology is often used in applications such as traffic signals and surveillance where long distances may make it difficult to run links in a star formation from a central switch and where downtime must be limited.
Generally speaking, ring architectures have these advantages:
1. They have fast failover times, typically sub-50ms.
2. They require a decreased number of ports. Fewer ports are needed to provide the same amount of resiliency as centralized switched networks with redundant paths. This results in decreased initial investment and lower ongoing maintenance costs.
3. They are scalable and enable a step-by-step network rollout. More switches can be added to the ring incrementally. The full traffic does not need to traverse a main/distribution switch.
4. They use bandwidth efficiently; dedicated paths are not required.
5. They simplify configuration. Predefined paths between the switches that are connected to the ring are not needed.