Concurrent Scaling: Evaluating AWS Lambda Performance through Load Testing

Abstract In the dynamic environment of serverless computing, efficient concurrency management and reasonable utilization of load testing techniques closely correlate with performance improvement. This paper, aiming to offer practitioners practical insights and contribute to the theory of serverless architectures, intends to illuminate the intricate connection between concurrency settings and load testing results within AWS Lambda. The paper thoroughly explores load testing and concurrency control in AWS Lambda. We specified a range of concurrency settings, including provisioned and unreserved, and meticulously selected representative Lambda functions. We subjected these functions to diverse simulated workloads using industry-standard load testing methods. We methodically measured and recorded performance metrics, including response times, throughput, and resource consumption, to identify significant patterns and correlations. The article shows discrete performance quirks associated with various concurrency settings by exposing necessary trade-offs and optimization techniques. In addition to providing helpful advice for developers and architects, this research advances our theoretical knowledge of how concurrency affects the performance of serverless functions. Focusing on the unique features of AWS Lambda, the need for customized load-testing approaches is emphasized, offering practitioners practical advice on how to improve the efficiency and scalability of their serverless apps. This article uniquely enriches the confluence of theory, practice, and policy in serverless computing. Defining the complex relationship between concurrency and load testing and providing a platform for future research enhances theoretical underpinnings. The results provide developers and architects with explicit guidelines on configuring concurrency settings to meet workload needs. Additionally, the study supports a sophisticated approach to load testing and concurrency control in the larger context of cloud computing, which adds to policy discussions surrounding serverless best practices.