In our home network, we have a WIFI router (which acts as a gateway router). What if it incurs hardware failure or some service interruption? We can no longer access the internet. Right?
Similarly, in an organization with a network design, there is just one gateway router helping every host in a LAN to connect to the outside world, i.e., the internet. Unluckily that gateway router fails for some reason. How can the host communicate with the outside world, remote sites, and the main branch office? Replacing a router or reconfiguring it may require time, and it is not an effective way to deal with such outages.
Why is redundancy needed in a network?
Redundancy in a network is needed to reduce the problem of outages. The network can eventually face a problem, be it a fault in the router supply, a cable might break, or some error with the configuration may occur. An effective network design has the least or no single point of failure.
In the case of WAN, routing protocols help find the next best path possible if a router fails.
In the case of LAN with redundant links and switches, protocols like Spanning Tree Protocol and Rapid Spanning Tree Protocol help to find the alternate path. But how can we get redundancy in the case of gateway routers?
First Hop Redundancy protocols help to achieve this goal.
In the given figure, LAN has redundancy, but WAN does not. It has a single point of failure has shown. If this point fails, traffic cannot flow from one side of the network to another.
The figure shows the same network with the reduced point of failure with the redundant router at the main site. In this design also, the risk of an outage is not eliminated.
The best design of our network would be a design with redundant links at the main site and at the remote sites as shown in the given figure.
Why do we need First Hop Redundancy Protocol?
Since we have already discussed the design of a network and got the optimized design with zero point of failure, why do we need a protocol since we already have redundant links?
To understand this, let us first understand how to use these redundant routers-
- All host of a LAN uses R1 as a default gateway, and when R1 goes down, the host is statically configured to use R2 as the default gateway.
- All host of a LAN uses R2 as a default gateway, and when R2 goes down, the host is statically configured to use R1 as the default gateway.
- Some hosts are configured to use R1 as the default gateway, while some are using R2 as the default gateway, and when either router fails, the remaining hosts are statically configured to the available router.
These options do not look convenient to use. The user has to take action and reconfigure the default gateway settings and revert the configuration when the primary router comes up.
First Hop Redundancy Protocol can help with this problem.
When this protocol gets configured, the two routers look like a single gateway router to the hosts, and if any router goes down, the host needs to do nothing to change its default gateway. Virtual IP Address is shared between the routers and the host use that virtual IP Address as their gateway router IP Address.
Since the virtual IP Address is shared between 2 or more gateway routers (even if one gateway goes down), the host does not need to change the gateway IP address. Gateway configured with this protocol shares messages to stay updated and to choose which gateway takes responsibility when the primary gateway goes down.
What are the types of First-Hop Redundancy protocols?
There are three types of First-Hop Redundancy protocols, namely-
- Hot Standby Router Protocol
- Virtual Router Redundancy Protocol
What is Hot Standby Router Protocol?
HSRP is a Cisco proprietary protocol that allows two or more routers to work together, which looks like a single virtual router to the hosts connected in a LAN. The purpose of HSRP is to eliminate the single points of failure and maintain uninterrupted services.
Common HRSP messages: –
It is sent by the active router to the standby router every 3 seconds to indicate its operational state. If the standby router does not receive a hello message for 10 seconds, it assumes that the active router is down and the standby router becomes active.
It is sent by a standby router to take the role of an active router even if the active router is not down. When the standby router wants to become an active router, it sends out coup messages with a higher priority value than the current active router.
It is sent by the active router when it wants to give up its role. This occurs when the router is going through maintenance or configuration change.
It is sent by the active router to inform the host connected to LAN about the virtual IP Address. This message is sent in response to ARP requests.
What are the HSRP states?
- Initial: – It Is the state when the router initially comes up and HSRP is configured to it.
- Learn: – In this state, the router is trying to learn the virtual IP Address.
- Listen State: – Router starts to listen to the hello messages from other routers.
- Speak: – In this state, the router sends a hello message and actively participates in an election to become an active router.
- Active: – When a router becomes active, it starts to act as the primary gateway and forwards network traffic.
- Standby: – Routers other than active router comes to the standby state. When the active router goes down, the standby router takes its place.
How does HSRP work?
HSRP works on an Active-Standby model where one gateway is in an Active state while others remain in the standby state. Once the Active router goes down, the standby router takes its place. HSRP operates by assigning a Virtual IP Address and a Virtual MAC Address.
Active router is selected based on priority. The router with higher priority becomes the active router. In case, 2 or more routers have the same priority, Router with a higher interface IP Address becomes active. If needed, priority can also be changed manually.
An active router periodically sends HSRP hello messages to the standby routers, indicating their availability. If the standby router does not receive a hello packet in a particular time frame (usually 10 seconds), it assumes that the active router is down, and hence standby router starts performing as a standby router.
NOTE: HSRP Version 1 uses 22.214.171.124 as the broadcast address, while HSRP version 2 uses 126.96.36.199 as the broadcast address.
Since the standby router sits idle until the active router goes down, there is a waste of bandwidth. In this case, the concept of load balancing can be used. Different routers can be configured as active routers for different subnets. For example: – R1 can be configured as an active router for VLAN 1, and R2 can be configured as active for VLAN 2.
Since both routers are connected to all the VLANs, even if an active router for a particular VLAN goes down, it can shift its traffic to a standby router.
What is Virtual Router Redundancy Protocol?
Virtual Router redundancy protocol is similar to HSRP. The working mechanism of both protocols is the same. VRRP also works on the Active-Standby model. VRRP can also use real IP Addresses as virtual IPs.
VRRP is an open-source IEEE standard protocol. VRRP uses 188.8.131.52 as its multicast address.
Hello timer for VRRP is 1 second while its hold time is 3 seconds.
What is Gateway Load Balancing Protocol?
Gateway Load Balancing Protocol is a Cisco proprietary protocol. The working mechanism of GLBP is similar to HSRP, but along with providing redundancy, it also provides load balancing for outgoing traffic. In the case of HSRP, each subnet needs to be configured individually to implement load balancing, which requires extra administrative work.
Gateway Load Balancing Protocol can do the load balancing over multiple gateways by providing a single virtual IP Address and multiple virtual Mac addresses. Each host is configured with the same virtual IP Address, while all the gateways in the virtual group can forward packets.
GLBP routers communicate with each other using 184.108.40.206 as a broadcast address. Hello timer for GLBP is 3 seconds, while the hold time is 10 seconds. GLBP only supports IPv4.