What is the granularity of serverless metrics when only using the AWS integration?

The granularity of serverless metrics when using the AWS integration is set to 1 minute. This means that the data points are collected and reported in intervals of 1 minute, providing a balance between data resolution and performance.

Using a 1-minute granularity allows for an adequate level of detail to monitor serverless applications efficiently, such as AWS Lambda functions, without overwhelming the monitoring system. It offers a good compromise for analyzing trends and ensuring that performance metrics can inform decisions effectively, without generating excessive data that may hinder real-time analysis or visualization.

The design of serverless architectures often involves transient workloads that can lead to a fluctuating usage pattern. Therefore, having metrics reported at a 1-minute interval is useful for understanding usage patterns while providing sufficient temporal resolution for operational insights. Higher granularity metrics, like 1 second, can generate a significant amount of data, which might not be practical given the bursty nature of serverless functions. Lower granularity options would not capture enough detail for timely insights.