Click any tag below to further narrow down your results
Links
Tokenflood is a tool designed for load testing instruction-tuned large language models (LLMs). It allows users to define various parameters like prompt lengths and request rates without needing specific prompt data, making it easier to assess latency and performance across different providers and configurations. Users should be cautious of potential costs when using pay-per-token services.
This article details Shopify's extensive preparations for the Black Friday Cyber Monday (BFCM) weekend, including year-round capacity planning and chaos engineering exercises. It outlines how the team conducts load testing and scale tests to identify and fix potential bottlenecks before peak traffic. The focus is on ensuring the infrastructure can handle unprecedented demand in 2025.
This article examines how many HTTP requests per second a single machine can handle using a simple setup. It details the testing process, configurations, and results from various load levels on different machine specifications. The findings highlight performance limits and response times under sustained loads.
Impulse is Airbnb's internal load-testing framework designed to enhance the reliability and efficiency of system-level testing. It features a range of tools including a load generator, dependency mocker, traffic collector, and testing API generator, enabling engineers to conduct self-service load tests that integrate seamlessly with CI/CD pipelines. This decentralized architecture allows for context-aware testing, accurate simulation of production conditions, and better management of system performance under varying traffic scenarios.