Dynamic Sampling is a feature of the ingestion pipeline that allows Sentry to automatically adjust the amount of data retained based on the value of the data. From all the data received by the SDKs, Sentry is able to extract low-granularity information through metrics, while Dynamic Sampling makes the decision of whether to keep or drop data.
The goals of Dynamic Sampling are:
- Value at any scale: Sentry has a full view of your application and can accurately show performance metrics, detect trends and performance issues from 100% of your events.
- Get value out of the box: Sentry automatically biases data retention in order to store sufficient samples of the highest value events in full detail.
- Prioritize based on your needs: Sentry provides a mechanism to express priorities as a set of rules and then adjusts data retention as much as is practical to honor them.
A high-level overview on how Dynamic Sampling works is shown in the following illustration: