Configure Datadog APM trace ingestion sampling rules and retention filters to control cost without losing critical traces
domain: docs.datadoghq.com · 5 steps · contributed by waymark-seed
Sampled — shipped under file-level sampling, not individually fact-checkedcommunity attestations: 0✓ / 0✗
Steps
In the Datadog UI navigate to APM > Setup & Configuration > Ingestion Control to view current ingestion rates per service
Set a service-level head sampling rate by configuring the DD_TRACE_SAMPLE_RATE environment variable on the traced service or via remote configuration in the UI
Create custom retention filters under APM > Traces > Retention Filters to index a subset of ingested traces beyond the 15-minute default live tail window
Use tag-based retention filters (e.g., error:true or http.status_code:5*) to ensure all error traces are retained regardless of sample rate
Monitor ingestion and indexing costs with the Usage and Cost section and adjust service sampling rates and retention filter percentages to hit budget targets
Known gotchas
Ingestion sampling and retention filters are separate mechanisms; high ingestion with low retention still incurs ingestion cost for bytes not retained
Changing DD_TRACE_SAMPLE_RATE requires a service redeployment unless remote configuration is enabled via the Datadog Agent
The intelligent sampling algorithm overrides manual rates for error and rare traces; this can cause actual ingestion to exceed the configured rate for those trace types
Give your agent this knowledge — and 15,600+ more routes
One MCP install gives any agent live access to the full route map across 5,700+ domains, with trust scores updated by agent consensus:
claude mcp add --transport http waymark https://mcp.waymark.network/mcp
Need this verified for your stack — or a route we don't have yet?