What is a redundant switch and how we can configure redundant topology with links?
A redundant switch is a fail-over switch that keeps the network up and running in the case of a network outage due to hardware or software failure.
Redundancy is important for every network however building a complete redundancy can be expensive so we have to configure the redundancy depending upon our requirements and the size of our network.
If we have a small company that can handle the downtime without much impact then we can create partial redundancy in our network however for the big network where downtime can cause some serious impact on revenue, we must configure full redundancy which will keep the network up and running without a noticeable downtime.
How to make switches redundant
To make the switches redundant, we have to use at least two switches at every layer on our network and interconnect those switches so that in case of failure, the network will use the redundant switch to transfer traffic.
Redundant switch topology
Below is an example of a redundant switch topology.
Above image shows an example of a redundant small office network in which switches are interconnected so if one of the switches fails then traffic will be routed via another switch.
You can see that one of the ports in one switch is blocked by the spanning tree protocol; you can download the lab and try disabling random switch to see how the blocked port will be used whenever there is a failure in the network.
Above image shows an example of a big enterprise where redundancy is built from the core layer to the access layer.
You can download this lab to enable and disable the switches to see how switches change paths while transferring traffic.
However, you can see that in this network design, there is a critical server that is attached to access layer switch 5 so if that switch comes down then it will be a downtime for the critical traffic and it will require a manual intervention to bring the server up.
In the above design, failure of one switch can cause a high impact even though we have proper redundant switches at each layer.
There is no perfect way to build a full redundancy as it is highly customization, we can design a redundant network as per our requirements and depending upon the type of traffic.
The below image shows how we can eliminate a single point of failure for the critical endpoint server, we have connected the server with two switches using two different network cards.
If one of the switches fails then the server will still be accessible in the network.
Redundant switch power supply
The power supply of the redundant switch depends upon the size and criticality of the network, using a single power supply can be an issue in case of power supply failure however that can be quickly changed with manual intervention.
If the network is handling critical traffic where even downtime for a short time can cause issues then we have to use a separate power supply for each switch as this will eliminate a single point of failure.
Redundant links between switches
To implement redundancy between switches, we have to connect each switch with the switch on the other layer. Daisy chaining the switches will have a single point to failure so to prevent the downtime; we must have redundant links between switches.
Cisco switch redundancy configuration
In Cisco, we have system already in place to enable the redundancy in switches, wherever switches are interconnected for redundancy, one prevalent issue is the loop of traffic between them however Cisco has a spanning tree protocol enabled by default on every switch to prevent loop so we can safely interconnect switches in Cisco for redundancy.
The default configuration of the spanning tree protocol may not work properly by default so we have to tweak the configuration to make it suitable for the network and to fully utilize the hardware capabilities.
Packet tracer switch redundancy
We have configured the redundant switch network in packet tracer, you can download the following lap and try to disable the different switches in the network to see how the network changes the path and uses the redundant switch for all traffic.
Usually, a redundant switch also shares the traffic load and it’s not idle however when one of the switches goes down the traffic is handled by the one switch only.
Redundancy in the core switch
We have to set up two core switches instead of one to enable redundancy in the core of the network. Its shown in the above image.
HSRP between switches for redundancy
HSRP is a router technology so to use this feature in switches we have to use layer 3 switches only as these switches have routing capabilities as well.