Managed Load Balancing

Load balancing is a technique that distributes incoming network traffic across a pool of servers so no single server is overloaded. The load balancer sits in front of the application fleet, decides which server should answer each request, and routes around servers that are unhealthy or saturated. The result is better performance, higher availability, and more predictable scaling under load.

Any application that needs to stay online during a server failure, scale beyond a single server's capacity, or absorb traffic spikes is a candidate for load balancing. Common drivers are seasonal peaks, marketing-driven traffic surges, regulatory uptime requirements, and the need to perform zero-downtime maintenance on individual servers.

Round Robin, Static (Weighted) Round Robin, Least Connections, and IP Source. Round Robin and Static Round Robin distribute by sequence and weight; Least Connections sends each new request to the least-busy server; IP Source pins traffic from a given client IP to a specific server, useful for geolocation-aware routing or session affinity at the network layer.

Round Robin distributes traffic by simple sequence regardless of how busy each server is, which is fine for short, similar requests but can overload a server stuck on a slow request. Least Connections looks at the actual open-connection count and routes to the least-loaded server, which handles long-lived or uneven workloads more gracefully.

Yes. SSL/TLS termination on the load balancer decrypts incoming HTTPS traffic before it reaches your backend servers. This centralizes certificate management and reduces the cryptographic workload on each web server, improving throughput. End-to-end encryption between the load balancer and backends is supported when policy requires it.

Sticky sessions (also called session affinity) keep a given user pinned to the same backend server for the duration of their session. They are useful for applications that store session state in local memory and cannot share that state across servers. Stateless applications ‐ those that store session data in a shared cache or database ‐ do not need sticky sessions and benefit from more even load distribution without them.

The load balancer continuously runs health checks against every server in the pool. When a server fails its check ‐ for example, no response on the configured port within the configured timeout ‐ it is taken out of rotation automatically and skipped until subsequent checks confirm it has recovered. Atlantic.Net's NOC is also alerted so engineers can investigate the underlying issue.

Yes. Servers can be added to or drained from the pool on request, which is what makes the service well suited to seasonal peaks, marketing launches, or rolling maintenance. Atlantic.Net engineers handle the change so existing sessions are not disrupted.

Yes. Managed Load Balancing is delivered alongside Atlantic.Net's Managed Firewall on the same FortiGate-based platform, so you get traffic distribution and perimeter security under a single managed contract, with one policy framework and one support relationship. GeoIP load balancing is also available via the Network Edge Protection service.

Pricing is based on the size of the load-balanced pool, the throughput required, and the bundled features (SSL termination, GeoIP, integration with Managed Firewall). Contact our sales team for a quote tailored to your environment.