Load Balancer

On Fransys, the entire infrastructure relies on load balancers. It's not an option to enable — it's the default behavior of every cluster.

Built-in load balancing

Every Fransys cluster is provisioned with a load balancer that automatically distributes incoming traffic to available nodes. This mechanism is transparent: you have nothing to configure, nothing to maintain.

In practice, this means:

• Traffic is distributed across the different instances of your applications
• If a node fails, traffic is automatically redirected to healthy nodes
• During scaling (adding nodes or replicas), new instances immediately receive traffic

Handling high traffic

Fransys's load-balanced architecture is designed to absorb traffic spikes. Combined with horizontal scaling and AutoPilot, your infrastructure adapts automatically:

1. The load balancer detects the load on each instance
2. AutoPilot scales the number of nodes or replicas if needed
3. The load balancer redistributes traffic to the new instances

This cycle is fully automatic — you don't need to intervene during a traffic spike.

High availability

The load balancer ensures the high availability of your applications:

• Health checks — Continuous verification of each instance's state. If an instance stops responding, it's removed from the distribution pool
• Zero downtime deployments — During a deployment, new instances are added to the pool before old ones are removed
• Intelligent distribution — Traffic is distributed evenly to avoid overloading any single instance

tip

You don't need to size or configure the load balancer. It's automatically adapted to the size of your cluster and the number of your applications. Focus on your code — Fransys handles the traffic.