Unresolved Datadog Agent log collection mode for Docker sidecar containers in development environments

Goal

Establish a deterministic log collection strategy for Docker sidecar containers in repeatable development environments using Datadog Agent.

Constraints and Uncertainties

The Datadog Agent 7.45 introduced Docker container mode for automatic log discovery, but documentation does not clarify whether sidecar container logs are consistently collected across all deployment scenarios. The Terraform Provider (v1.31.0) officially supports only a subset of Datadog resources; custom metric and dashboard resources are missing, limiting reproducible infrastructure definitions. Additionally, mixing host-mode and container-mode log collection can lead to duplicate entries and increased storage costs.

Documentation does not specify a preferred mode (container vs host) for log collection, leaving teams to decide without clear guidance, which can affect environment repeatability.

Specific Questions

• Which log collection mode provides reliable sidecar log ingestion in Docker development environments?
• How can teams configure the Terraform provider to guarantee consistent log availability across different agent versions?