Change adaptive metric collection for low-traffic Dynatrace hosts

For a Dynatrace deployment assumed to be on version 1.2023 or later, the goal is to decide whether adaptive metric collection should remain automatic for low-traffic hosts or be supplemented by explicit sampling controls to reduce ingestion and storage costs.

The documented feature lowers collection frequency for minimally active hosts. The open configuration boundary is whether current releases expose a per-host or host-group minimum sampling interval; available research indicates that they do not, but the version boundary and current “Metric sampling” documentation require confirmation. Custom Metrics API polling and tag-based retention are separate mechanisms, and their interaction with adaptive sampling is unclear. Changing sampling could also affect the temporal resolution available to anomaly detection and alert evaluation.

Which configuration is supported for setting a minimum sampling interval for selected hosts or groups?

If no such control exists, can custom API polling coexist with adaptive sampling without disabling it or increasing ingestion?

Does the adaptive-sampling behavior or configuration boundary differ across Dynatrace 1.2023+ releases?