ZyXEL Communications ZyWall ATP series User Manual page 289

Since WAN 2 has a smaller load balancing index (meaning that it is less utilized than WAN 1), the Zyxel
Device will send the subsequent new session traffic through WAN 2.
Table 113 Least Load First Example
OUTBOUND
INTERFACE
AVAILABLE (A)
WAN 1 512 K
WAN 2 256 K
Weighted Round Robin
Round Robin scheduling services queues on a rotating basis and is activated only when an interface has
more traffic than it can handle. A queue is given an amount of bandwidth irrespective of the incoming
traffic on that interface. This queue then moves to the back of the list. The next queue is given an equal
amount of bandwidth, and then moves to the end of the list; and so on, depending on the number of
queues being used. This works in a looping fashion until a queue is empty.
The Weighted Round Robin (WRR) algorithm is best suited for situations when the bandwidths set for the
two WAN interfaces are different. Similar to the Round Robin (RR) algorithm, the Weighted Round Robin
(WRR) algorithm sets the Zyxel Device to send traffic through each WAN interface in turn. In addition, the
WAN interfaces are assigned weights. An interface with a larger weight gets more chances to transmit
traffic than an interface with a smaller weight.
For example, in the figure below, the configured available bandwidth of WAN1 is 1M and WAN2 is 512K.
You can set the Zyxel Device to distribute the network traffic between the two interfaces by setting the
weight of wan1 and wan2 to 2 and 1 respectively. The Zyxel Device assigns the traffic of two sessions to
wan1 and one session's traffic to wan2 in each round of 3 new sessions.
Figure 198 Weighted Round Robin Algorithm Example
Spillover
The spillover load balancing algorithm sends network traffic to the first interface in the trunk member list
until the interface's maximum allowable load is reached, then sends the excess network traffic of new
sessions to the next interface in the trunk member list. This continues as long as there are more member
interfaces and traffic to be sent through them.
Suppose the first trunk member interface uses an unlimited access Internet connection and the second
is billed by usage. Spillover load balancing only uses the second interface when the traffic load exceeds
the threshold on the first interface. This fully utilizes the bandwidth of the first interface to reduce Internet
usage fees and avoid overloading the interface.
Chapter 9 Interfaces
MEASURED (M)
412 K
198 K
ZyWALL ATP Series User's Guide
289
LOAD BALANCING INDEX
(M/A)
0.8
0.77