WAN Load Balancing

Topic

This article describes the WAN Load Balancing feature of the Datto Networking Appliance (DNA).

Environment

  • Datto Networking Appliance (DNA)

Description

The WAN Load Balancing feature allows an operator to configure the DNA to use multiple WAN connections for increased throughput, per-WAN load percentages to allow optimal performance in deployments with asymmetric link capacity, and local subnets to prefer either the ISP 1 or ISP 2 interface.

To configure load balancing on your DNA, log into the DNA web interface, and click Networks, as shown in Figure 1.

mceclip0.png
Figure 1: Networks

Procedure

1. Click the WAN Load Balancing option in the Networks pane.

Figure 2: WAN Load Balancing

2. You will see the configuration card shown in Figure 3.


Figure 3: Load balancing configuration card

Select from the following options to configure disable load balancing on your device:

A. Enabled: Select Yes to enable load balancing or select No to disable it.
B. Load Percentage Balancing: Allows you to specify the total amount of traffic load to pass over each interface. Modifying a single value will cause the second value to adjust automatically.
C. LAN Load Balancing Policy: Allows you to set preferred interfaces for each configured network on the DNA to use. Selected LANs will favor the designated WAN.

Technical Notes

  • LAN traffic will be simultaneously routed through both WAN interfaces when both are configured and active.
  • When either WAN interface is unavailable, the LAN will be routed through the other, non-preferred or secondary, active WAN interface.
  • LTE failover will only work when both WAN interfaces are down or unavailable.
  • If both WAN interfaces are active, you will see both active gateway IP addresses in the WAN Details card.
  • Traffic shaping and UDS will work in the same manner regardless of which WAN interface the traffic is passed through.
  • The maximum throughput for any single traffic flow can not exceed the link capacity of the uplink. Only the aggregate throughput is increased. For example, a single file download won’t go faster, but two separate downloads could because they are separate flows.