Load balancer

Distribute visitors across multiple servers

A Load Balancer provides one entry point for your service and distributes requests across multiple cloud servers.

Get startedPricing calculator
  • Pay by the hour for the time you use
  • Estimate costs before provisioning
  • Tehran data centers
Example

Request routing

Server 1HealthyServer 2HealthyServer 3Unhealthy

2 servers are healthy. Requests reach only these servers; unhealthy servers are excluded.

Select a server to see how routing responds.

How it works

How traffic is balanced

Choose one of two methods when creating a load balancer. Every listener has its own protocol and port.

Round robin

Round Robin

Requests cycle through servers in order: first to server 1, second to server 2, third to server 3, then repeat. The simplest choice for similarly sized servers.

ExampleSix consecutive requests

Least connections

Least Connection

Each request goes to the server with the fewest active connections. This balances load better when requests take different amounts of time.

ExampleActive connections per server; the next request goes to server 2

Listeners

Each listener defines the incoming protocol and port and the target server port. Web listeners can also enable automatic HTTPS redirects and TLS encryption.

HTTP808080
HTTPS4438080
TCP54325432
Example
Reliability

Server health

The load balancer checks target servers regularly. After repeated failures, it stops sending new requests to an unhealthy server while the others continue serving traffic.

  • Set the health check path and expected response code for each listener.
  • Configure check interval, response timeout and failure threshold.
  • Unhealthy servers are removed from request routing to avoid visitor errors.

Health check settings

Example
Check path
/health
Expected response code
200
Check interval
5 seconds
Response timeout
3 seconds
Failure threshold
3

A server's health check results

  1. Successful
  2. Successful
  3. Failed
  4. Failed
  5. Failed
  6. Three consecutive failures exclude the server
Startup

Set up in four steps

Everything happens from the project page.

  1. Select a private network

    The load balancer connects to one of your project's private networks.

  2. Define listeners and balancing method

    Enter the listener type (HTTP, HTTPS or TCP) and ports, and choose one of the two balancing methods.

  3. Add target servers

    Select the cloud servers in your project that should receive requests.

  4. Configure health checks

    Set check path, interval, timeout and failure threshold. You can edit servers and settings later.

Position in the network

Alongside your private network

The load balancer connects to a project private network, with your cloud servers behind it. Requests enter through one endpoint and are distributed across servers.

About Private Networking
Features

What is included

The main capabilities of Load balancer at a glance.

  • Service entry point

    Choose the incoming protocol (HTTP, HTTPS or TCP) and port.

  • Target servers

    Select the servers that should receive requests.

  • Service management

    Manage load balancer settings and status within your project.

Use cases

Who it is for

  • Services with multiple servers
  • High-traffic websites
  • Decouple the entry point from servers
User guide forLoad balancer
Recommended

Services commonly used alongside Load balancer.

Frequently asked questions

Before you start

What problem does Load balancer solve?

When every request reaches one server, increasing traffic puts it under pressure and adding another server becomes cumbersome.

What is Load balancer suitable for?

Load balancer supports use cases such as Services with multiple servers, High-traffic websites, Decouple the entry point from servers. The right choice depends on your project's technical and operational requirements.

How do I get started with Load balancer?

Read the product guide and estimate costs in the calculator, then create the service in your chosen project in the client panel.

Get started with Load balancer

Register, estimate costs before provisioning and pay hourly for the time you use the service.