Zyxel USG FLEX H Series - WAN Trunk Load balancing for USG FLEX H series

This article explores WAN Trunk Load Balancing for the Zyxel USG FLEX H Series, covering key concepts, configuration steps, and practical use cases. It explains different balancing methods, their real-world applications, and how to optimize performance on Zyxel Firewall H Series devices.

Note: The default algorithm, Weighted Round Robin, distributes traffic equally across multiple WAN interfaces in a round-robin fashion. Each session is alternately assigned to the WAN interfaces in sequence, ensuring a balanced distribution.

About WAN Load Balancing

• Directs traffic to the WAN interface with the least current load.
• Balances traffic based on utilization(percentage), ensuring efficient usage of available bandwidth.
• Useful for maximizing the performance of multiple WAN connections with varying bandwidths.

• Primarily uses the main WAN interface until a specified traffic threshold is reached.
• After reaching the threshold, additional traffic "spills over" to the secondary WAN interface.
• Ideal for scenarios where the primary WAN interface has unlimited bandwidth and the secondary interface is metered.

Load Balancing Index - The load balancing index determines the criteria for measuring bandwidth load usage. It offers three options:

• Inbound: Measures incoming traffic.
• Outbound: Measures outgoing traffic.
• Sum: Measures the total of both inbound and outbound traffic.

For example, for  a site with a light load, the best load-balancing strategy depends on your goals, but here’s how they compare:

Weighted Round Robin is best if you have servers of different capacities and want to distribute requests proportionally. It works well for predictable, consistent traffic.
Least Load First is best if you want to optimize server utilization dynamically, ensuring requests go to the least busy server. It is also good for sites with fluctuating traffic.
Spillover is useful when you want to keep traffic on primary servers and only use backups when they reach capacity. It is more relevant for handling occasional spikes rather than light loads.

Least Load First is often the best for a light load because it keeps servers as evenly utilized as possible while avoiding unnecessary complexity. Weighted Round Robin is also a simple and effective choice if all servers are similar.

Practical Scenarios for WAN Load Balancing

Spillover Scenario

• Primary WAN (WAN1): Unlimited bandwidth, preferred for regular traffic.
• Secondary WAN (WAN2): Metered connection, used only when primary exceeds 800 Mbps.
• Outcome: Regular traffic utilizes WAN1, and only excess traffic spills over to WAN2, saving costs on metered usage.

Least Load First Scenario

• Multiple WAN Connections: Different bandwidth capacities.
• Outcome: Traffic is dynamically routed to the interface with the least load, optimizing the usage of available bandwidth and improving overall network performance.

Configuring WAN Load Balancing

• Navigate to Network Interface:
  • Go to the Network section in your firewall’s configuration interface.
  • Select Interface and then Trunk.

• Create User-Defined Trunk:
  • Click on the Add button to create a new trunk.
  • Choose the load-balancing algorithm. 
• Set Load Balancing Index:
  • Select the appropriate load balancing index (Inbound, Outbound, or Sum).
• Add Interfaces:
  • Add the WAN interfaces you want to include in the trunk.
  • Specify the bandwidth limits for each interface

Set Default Trunk Routing:

• After creating a user-defined trunk, go to the Network section and select Interface.
• Click on Default Trunk and choose the user-defined trunk profile to set it as the default WAN load balancing policy.