WAN Failover

WAN failover is the process in which redundant WAN connections are configured to increase WAN uptime. To establish WAN failover, two separate WAN connections from different ISPs are required which work as primary and secondary connections. It is important that both WAN connections are from different ISPs so that if any problem arises in the WAN link from one ISP, link from second ISP can be used to establish a WAN connection. On the other hand, if both connections are from same ISP and the ISP’s link goes down, both primary and secondary connections fail at the same time, hence making WAN links unavailable even if WAN failover has been configured.

When WAN failover is configured, the two WAN links are configured as active/passive connections. This means that under normal conditions (when no WAN downtime occurs), primary WAN connection from one ISP is used to route the entire network traffic and it works as active connection. In this case secondary connection from the other ISP remains in inactive/idle state and stays in passive mode. When primary WAN connection fails, the secondary connection automatically comes out of the passive mode and becomes active. This process is technically known as failover and is completely transparent to end-users. Once automatic failover takes place, network traffic is automatically routed through the secondary WAN connection. Secondary WAN connection remains active till the primary connection is up again to route the network traffic. As soon as the primary WAN connection becomes active, the secondary connection automatically becomes idle and enters into passive mode. This process is technically known as WAN failback.

WAN Failover Components

Typically WAN failover is configured using two ports. Utilization of each port is discussed below:

• Primary Port – As discussed above, primary port is used to connect primary WAN link from one, more preferred ISP. This port handles the enter network traffic under normal conditions and works as active WAN connection.
• Secondary Port – The second port is configured to be used as passive WAN connection port, and the secondary WAN link from the less preferred ISP is connected to it. This (passive) port becomes active only when the primary port faces downtime and fails to handle the network traffic. During the unavailability of primary WAN connection, secondary port plays the role of primary port.

Before configuring WAN failover, administrators must verify if the device supports failover configuration. In case failover is not supported by the device, two devices must be used to configure active/passive WAN failover.

WAN Load Balancing

WAN load balancing is a process in which at least two registered IP addresses are connected to a local area network to distribute network load among registered IP addresses equally. WAN load balancing ensures that a single WAN link does not lose its efficiency due to heavy network load. Unlike WAN failover where the participating registered IP addresses were required to be from different ISPs, in WAN load balancing, registered IP address from the same ISP can be used.

Since WAN load balancing is a kind of active/active configuration, WAN traffic is equally distributed among two participating WAN links using round-robin method. Round-robin ensures that none of the participating WAN link gets overloaded due to excess of internet traffic.

Before configuring WAN load balancing on a device, administrators must ensure that the device supports the load balancing feature. In case WAN load balancing is not supported by the device, two devices must be used to accomplish the task.

While configuring WAN failover or load balancing, administrators must keep in mind that:

• Device Support – Not all devices support WAN failover and load balancing configuration.
• WAN Failover or Load Balancing – Any one of the above discussed configurations can be done on a device at a time. This means that the device can be configured either for WAN failover or for WAN load balancing at a single given time.
• Traffic Type – WAN failover and load balancing is only applied on outgoing WAN traffic.