What are the Different Types of Network Redundancy?

There are four main kinds of network redundancies, known as Multiple Spanning Trees (MST), Ring Networks, Diverse Trunking, and Multi-Protocol Label Switching (MPLS). 

‍

Below, I’ll dive into more detail on how each of these work, and what makes them distinct from each other. 

‍

1. Multiple Spanning Trees (MST)

‍

This is essentially a network protocol that creates multiple spanning tree instances for different VLANs (Virtual Local Area Networks) within a network. Since it’s a tree, even if a branch fails, there’s always another branch that can take over, and continue the network without disrupting the traffic. 

‍

MSTP is derived from STP (Spanning Tree Protocol), and the core part of understanding here is that each group of VLANs gets its own tree, which reduces the chance of looping, and adds a multitude of paths in case one fails, giving more leeway in mission-critical components. 

‍

Here are the components of a Multiple Spanning Trees (MST) system:

‍

‍

2. Ring Networks

‍

A ring network is a type of computer network configuration where each node (computer, server, or network device) is connected to exactly two other nodes, forming a single continuous pathway for signals through each node - a ring. 

‍

Data travels in one direction around the ring, from one node to the next, until it reaches its destination. Ring networks were more popular in the past, particularly with the IBM Token Ring technology. 

‍

However, with advancements in networking technologies, other configurations like star and mesh networks have become more prevalent due to their flexibility, scalability, and fault tolerance.

‍

3. Diverse Trunking

‍

Diverse trunking, refers to the practice of using multiple physical connections (trunks) between network switches, routers, or between a network and its service provider, that take different physical paths through the network. 

‍

Redundant content delivery in trunk links is achieved by setting up multiple physical or logical paths for data to travel between these devices. This setup ensures that if one path fails (due to hardware failure, cable issues, etc.), another path can take over, maintaining network connectivity and services without interruption.

‍

By ensuring that the trunks do not share the same physical route, diverse trunking protects against failures that could occur due to issues like cable cuts, hardware failures, or other disruptions that might affect a single path.

‍

To break this concept down in more detail:

‍

Imagine two switches, Switch A and Switch B, connected by a single trunk link. If that link fails, all traffic between the switches is halted, potentially disrupting network operations. To create redundancy, you would add a second trunk link between Switch A and Switch B, possibly over a different physical route. 

‍

This setup is managed by protocols like LACP (Link Aggregation Control Protocol) that can automatically detect and switch to the backup link if the primary link fails, ensuring continuous network operation. This concept is similar to having multiple roads between two cities; if one road is blocked, traffic is rerouted to another road, preventing a complete halt in movement.

‍

4. Multi-Protocol Label Switching (MPLS)

‍

Multi-Protocol Label Switching (MPLS) is a data-carrying technique for high-performance telecommunications networks. MPLS directs data from one network node to the next based on short path labels rather than long network addresses, avoiding complex lookups in a routing table and speeding up traffic flow.

‍

In practice, an MPLS network assigns labels to packets at the entry point, forwards them based on the labels through the network, and then removes the labels at the exit point.

‍

Here is an example MPLS sequence:

‍

• Label Assignment (Ingress Router in City A): When data enters the MPLS network in City A, the ingress router examines the data packet's destination (City D) and assigns a label based on the preconfigured Label Switched Path (LSP) that leads to City D. This label effectively says, "This is the path to take to reach City D efficiently."
• Label Switching (Intermediate Routers in Cities B and C): As the packet arrives at each router (first in City B, then in City C), the routers look at the label to determine how to forward the packet. They don't need to perform complex routing table lookups based on the destination IP address. Instead, they use the label to make a quick decision on where to send the packet next. Each router might swap the incoming label with a new label before forwarding the packet, guiding it along the predetermined LSP.
• Label Removal (Egress Router in City D): Once the packet reaches the final router in the MPLS network (the egress router in City D), the label is removed. The packet is then forwarded to its final destination within City D based on its IP address, as it would be in a traditional IP network.

‍

For more insight, I’d recommend checking out how Dynamic Traffic Redundancy works. 

‍