Load Balancer

Azion Load Balancer is an Edge Application module designed to balance traffic across your origins, data centers, or cloud providers, mitigating network congestion and server overload. To achieve high levels of fault tolerance and performance in case there are any issues with your origin servers, your content and applications must be available from multiple sources.

With Load Balancer, you can:

• Distribute traffic across your origin servers.
• Ensure the availability of your content and applications even in case of incidents.
• Choose between load balancing algorithms.
• Create routing rules to balance HTTP and HTTPS requests.

For more options on how to direct traffic to Load Balancer based on other criteria, see the Rules Engine for Edge Application page.

Implementation

Balancing method

The balancing method defines how the load will be distributed among your origins.

Round-Robin

The Round-Robin balancing algorithm ensures an even distribution of traffic among origins in rotation. Instead of considering the response time of each origin, it focuses on the volume of requests.

Each origin receives a load in proportion to its assigned weight in the round-robin. While slower origins may accumulate more connections in parallel, the algorithm guarantees equitable load balancing across all origins.

Least Connections

The Least Connections balancing algorithm monitors the active connections with each origin and consistently directs the next request to the origin with the fewest active connections. This approach optimizes load distribution by sending fewer requests to slower origins, while allowing faster origins to handle more requests in succession.

IP Hash

The IP Hash balancing algorithm tracks user IP addresses and associates the same origin with each IP. By creating a consistent mapping between users and origins, this approach ensures that each user’s requests are directed to the same origin server.

Address

You may add more than one Address for your origin. Addresses in Load Balancer follow the same logic as a Single Origin address.

Weight

You may define a Weight for each address by assigning them a number from 1 to 10.

The weight sets the proportion of load that the origin will receive. The larger the weight, the more load the origin will handle. For example, if you set the weight of an origin to 3, it will receive 3 times more load than an origin with a weight set to 1.

Server Role

If you use the Round-Robin or Least Connections methods, you can define a Server Role for each origin:

• Primary: are the go-to origin for requests and will be prioritized over Backup origins.
• Backup: act as standby origins and only receive the load if all Primary origins fail.

Active

In cases of maintenance or origin outages, you can disable an address temporarily from the balance. Turn off the Active switch and your server will become inactive, ensuring your infrastructure will smoothly undergo maintenance procedures without disrupting the overall performance.

Limits

Addresses

With Load Balancer active, you may create up to 10 origin addresses for each origin.