Optimizing Performance Without the Price Tag

When performance is slow(er), the typical response is to add more resources, but this often raises the question, "Can this be cheap(er)?" The challenge is optimizing performance without incurring higher costs. With over a decade of performance engineering experience, Dharmita and Dake demonstrate how even small data volumes or load can uncover optimization opportunities. Using Datadog APM, they identified anti-patterns and the code fixes resulted in a drastic reduction of resources. Beyond code fixes, performance tuning plays a critical role. By combining load testing with OpenTelemetry metrics, they optimized database connection settings, reducing connection waits and SQL latency. These improvements led to a more efficient system without needing additional resources.