Large Scale Load Tests

Discover the unparalleled power and affordability of LoadForge, one of the world's highest-capacity load test solutions.

Key Features

• Scalability: Achieve over 5 million SSL transactions per second (TPS) and beyond.
• Duration: Possibility to run tests continuously for several days if required.
• Affordability: Test at capacities that were previously either out of reach or prohibitively expensive.

The Technology Behind the Power

How Do We Achieve Such Scale?

LoadForge is a groundbreaking, cloud-native solution. It harnesses the capability to initiate an almost infinite number of individual cloud test servers to dispatch requests.

For instance, consider the scenario where you target 200,000 SSL users per second:

1. Distribute it across 20 servers.
2. Assign 10,000 users to each server.
3. This results in ~400,000 users or approximately 800,000 requests per second.

Amplify this further by launching multiple tests simultaneously from diverse data centers such as London, San Francisco, Singapore, and New York:

• 800,000 requests/second/test multiplied by 4 tests gives 3,200,000 SSL TPS.

Affordability: High Performance at a Fraction of the Cost

LoadForge leverages public cloud infrastructures, providing a transparent cost structure. For a large-scale test:

• Each test server is priced at approximately $0.24/test/server.
• A test with 3.2 million SSL TPS, involving 80 servers, would amount to just $19.20 for an entire hour.

Designing High Load Tests: An Overview

Designing tests for high load scenarios is often more straightforward than standard tests. In such scenarios, you typically assess individual components of your application, such as:

• Web servers
• Caches
• Load balancers
• CDNs

We recommend isolating components for focused testing, but the choice remains with the user.

Example of a High Throughput Test File:

from locust import task, between
from locust.contrib.fasthttp import FastHttpUser


class MyUser(FastHttpUser):
    @task
    def index(self):
        response = self.client.get("/")

Depending on your server's response rate, this test can simulate between 10,000 to 25,000 requests per second for each server.

Engage with Our Experts

Tests of this magnitude necessitate either our Enterprise or Premium packages. While the Enterprise package supports this large scale, the Premium package accommodates up to 200,000 users. Both packages come with dedicated support. For test design and scaling insights, we encourage you to connect with our team.