Propagate sampling metadata across trace chunks

[sarahchen6]

What Does This Do

Propagate sampling metadata across trace chunks by copying over the sampling decision maker (_dd.p.dm) on trace chunks that do not contain the local root span.

Motivation

Keep sampling metadata consistent across trace chunks. Non-root trace chunks previously did not include _dd.p.dm, which led trace intake to classify the adaptive-sampled traces with the wrong ingestion_reason (i.e. defaulting to rule instead of adaptive_rule). This led to over-ingestion of traces when adaptive sampling was enabled.

Additional Notes

May be superseded by #12096

Contributor Checklist

• Format the title according to the contribution guidelines
• Assign the type: and (comp: or inst:) labels in addition to any other useful labels
• Avoid using close, fix, or any linking keywords when referencing an issue
  Use solves instead, and assign the PR milestone to the issue
• Update the CODEOWNERS file on source file addition, migration, or deletion
• Update public documentation with any new configuration flags or behaviors
• Once approved, use merge queue to merge the PR

Jira ticket: [PROJ-IDENT]

[datadog-prod-us1-6]

🎯 Code Coverage (details)
• Patch Coverage: 50.00%
• Overall Coverage: 57.74% (+0.04%)

_{This comment will be updated automatically if new data arrives.
🔗 Commit SHA: 71f3bbb | Docs | Datadog PR Page | Give us feedback!}

[dd-octo-sts]

🟡 Java Benchmark SLOs — Performance SLO warning (near threshold)

Suite	Status
Startup	🟡 warning

SLO thresholds are defined here based on automatically generated metrics. A warning is raised when results are within 5% of the threshold.

PR vs. master results

Scenario	Candidate	master	Δ (95% CI of mean)
startup:insecure-bank:iast:Agent	14.03 s	14.05 s	[-0.9%; +0.7%] (no difference)
startup:insecure-bank:tracing:Agent	12.93 s	13.06 s	[-1.8%; -0.2%] (maybe better)
startup:petclinic:appsec:Agent	16.92 s	16.76 s	[+0.1%; +1.8%] (maybe worse)
startup:petclinic:iast:Agent	16.83 s	16.89 s	[-1.1%; +0.5%] (no difference)
startup:petclinic:profiling:Agent	16.51 s	16.72 s	[-2.3%; -0.3%] (maybe better)
startup:petclinic:sca:Agent	16.90 s	16.82 s	[-0.5%; +1.4%] (no difference)
startup:petclinic:tracing:Agent	16.09 s	16.03 s	[-0.7%; +1.4%] (no difference)

Commit: 71f3bbb7 · CI Pipeline · Benchmarking Platform UI

---

Load and DaCapo benchmarks can be triggered manually in the GitLab pipeline. Results will appear in the Benchmarking Platform UI after completion.

[datadog-prod-us1-6]

More details

The changed write path copies the local root decision maker only onto the first span of a separately exported chunk, after sampling succeeds and before serialization. Propagation-tag locking and pending-trace ordering match that intent, and no diff-only behavioral regression was identified; execution was limited by the unavailable Java 25 toolchain.

Was this helpful? React 👍 or 👎

_{Open Bits AI session}

_{🤖 Datadog Autotest · Commit 71f3bbb · What is Autotest? · @DataDog review to ask questions · Any feedback? Reach out in #autotest}