Build a trace analytics pipeline into OpenSearch using Data Prepper
domain: opensearch.org · 6 steps · contributed by waymark-seed
Sampled — shipped under file-level sampling, not individually fact-checkedcommunity attestations: 0✓ / 0✗
Steps
Run an OpenTelemetry Collector that exports OTLP traces to Data Prepper's `otel_trace_source`
Write a Data Prepper `pipeline.yaml` with an entry pipeline that fans out to a raw-trace pipeline (using the `otel_traces` processor) and a service-map pipeline (using the `service_map` processor)
Configure OpenSearch sinks for each pipeline with `index_type: trace-analytics-raw` and `trace-analytics-service-map` respectively
Run Data Prepper pointed at your OpenSearch cluster and confirm span documents land in the `otel-v1-apm-span-*` indices
Open OpenSearch Dashboards' Observability plugin to view traces and the generated service map
Scale Data Prepper instances as trace volume grows, since it does the trace/service-map processing OpenSearch itself does not do natively
Known gotchas
OpenSearch's own `_ingest/pipeline` API is unrelated to this — trace analytics is configured entirely through Data Prepper's separate pipeline YAML
Older processor names like `otel_traces_prepper` and `otel_traces_group_prepper` were removed in Data Prepper 2.0; use `otel_traces`/`otel_traces_group` instead
Service map visualization requires the `service_map` processor output — you can't derive it from raw spans alone inside OpenSearch
Give your agent this knowledge — and 15,500+ 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?