feat(otel-metrics): materialize SDK trace histograms as APM stats

[mabdinur]

Summary

Convert the Datadog SDK delta histogram traces.span.sdk.metrics.duration into the exact V4 APM-stats payload expected by the trace-stats pipeline.

This PR is stacked on #54906, which owns SDK metric recognition, field mapping, validation, normalized tags, time windows, error isolation, and sparse-histogram construction. This PR keeps only the Agent protobuf adapter and delivery path.

Motivation

The Agent metrics pipeline cannot deliver the derived trace.* series through generic metrics intake because that namespace is owned by APM trace stats. The SDK histogram therefore needs to be materialized as an APM-stats payload and sent through the existing trace-agent stats path.

Flow

1. sdktracestats maps the OTLP histogram into transport-neutral V4 fields.
2. otlp_client_side_stats.go adapts those fields to the Agent V4 protobuf.
3. The serializer exporter posts the protobuf to local /v0.6/stats with Dd-Protocol: otlp.
4. The trace agent forwards it to backend /api/v0.2/stats.
5. ddoghq/dd-go#8800 recognizes the protocol and emits Kafka message version 4.

Stack and shipping order

1. Merge #54906, then rebase this PR onto main.
2. Merge and deploy ddoghq/dd-go#8800 so trace-edge supports the V4 protocol.
3. Roll out this PR only after [runtime-security] add security-agent.yaml to flare #8800 is deployed. ddoghq/dd-source#56526 can merge and roll out independently after feat(otel-metrics): share SDK trace stats mapping #54906.

Behavior and scope

• Supports Agent OTLP ingest and DDOT; the OSS exporter remains on its existing pass-through behavior because it does not opt into this remapping.
• Leaves the Datadog connector and legacy V2 stats paths unchanged.
• Processes delta histograms only; invalid and cumulative histograms are skipped because this path keeps no cumulative-to-delta state.
• Marks SDK-derived stats with the existing span_source:datadog convention.
• Treats derived APM-stats delivery as best effort. A local APM error is logged without failing or retrying the metrics batch.

Validation

The affected Bazel suites pass:

//pkg/opentelemetry-mapping-go/otlp/metrics:metrics_test
//pkg/opentelemetry-mapping-go/otlp/metrics/sdktracestats:sdktracestats_test
//comp/otelcol/otlp/components/exporter/serializerexporter:serializerexporter_test
//comp/otelcol/otlp/components/exporter/datadogexporter:datadogexporter_test
//comp/otelcol/ddprofilingextension/impl:impl_test
//comp/trace/agent/impl:impl_test
//pkg/trace/api:api_test
//pkg/trace/writer:writer_test

The compact smoke suite covers 95.6% of executable lines in the new Agent adapter. A local cross-repository E2E also passed:

OTLP HTTP histogram -> DDOT -> Agent V4 trace writer -> dd-go trace-edge -> Kafka-bound raw stats message version 4

Review focus

• Agent protobuf adaptation versus the shared mapping contract.
• Additive behavior alongside legacy APM stats protocols.
• Agent/DDOT delivery and failure isolation.
• V4 receiver/writer headers and forwarding.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 8c46d17544

ℹ️ About Codex in GitHub

Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".

[mabdinur]

@codex review

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 8c46d17544

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[datadog-datadog-prod-us1-2]

🎯 Code Coverage (details)
• Patch Coverage: 77.99%
• Overall Coverage: 52.42% (+0.01%)

_{This comment will be updated automatically if new data arrives.
🔗 Commit SHA: d76074a | Docs | Datadog PR Page | Give us feedback!}

[dd-octo-sts]

Files inventory check summary

File checks results against ancestor b6a50788:

Results for datadog-agent_7.84.0~devel.git.23.0fe5dfc.pipeline.130041772-1_amd64.deb:

No change detected

Results for datadog-iot-agent_7.84.0~devel.git.23.0fe5dfc.pipeline.130041772-1_amd64.deb:

No change detected

[dd-octo-sts]

Static quality checks

✅ Please find below the results from static quality gates
Comparison made with ancestor 3b2688f
📊 Static Quality Gates Dashboard
🔗 SQG Job
SOME SIZE DELTAS ARE N/A (ANCESTOR METRICS NOT YET AVAILABLE). RETRY JOB

Successful checks

Info

	Quality gate	Change	Size (prev → curr → max)
✅	agent_deb_amd64	N/A	N/A → 759.706 → 764.500
✅	agent_deb_amd64_fips	N/A	N/A → 712.566 → 715.740
✅	agent_heroku_amd64	N/A	N/A → 312.045 → 319.050
✅	agent_msi	N/A	N/A → 643.685 → 660.960
✅	agent_rpm_amd64	N/A	N/A → 759.689 → 764.470
✅	agent_rpm_amd64_fips	N/A	N/A → 712.550 → 715.740
✅	agent_rpm_arm64	N/A	N/A → 735.446 → 736.540
✅	agent_rpm_arm64_fips	N/A	N/A → 691.633 → 693.690
✅	agent_suse_amd64	N/A	N/A → 759.689 → 764.470
✅	agent_suse_amd64_fips	N/A	N/A → 712.550 → 715.740
✅	agent_suse_arm64	N/A	N/A → 735.446 → 736.540
✅	agent_suse_arm64_fips	N/A	N/A → 691.633 → 693.690
✅	docker_agent_amd64	N/A	N/A → 818.321 → 820.040
✅	docker_agent_arm64	N/A	N/A → 819.122 → 820.740
✅	docker_agent_jmx_amd64	N/A	N/A → 1009.219 → 1010.800
✅	docker_agent_jmx_arm64	N/A	N/A → 998.672 → 1000.420
✅	docker_cluster_agent_amd64	N/A	N/A → 210.495 → 211.300
✅	docker_cluster_agent_arm64	N/A	N/A → 223.518 → 223.810
✅	docker_cws_instrumentation_amd64	N/A	N/A → 7.439 → 7.480
✅	docker_cws_instrumentation_arm64	N/A	N/A → 6.877 → 7.110
✅	docker_dogstatsd_amd64	N/A	N/A → 39.482 → 39.960
✅	docker_dogstatsd_arm64	N/A	N/A → 37.560 → 38.270
✅	docker_host_profiler_amd64	N/A	N/A → 305.866 → 317.690
✅	docker_host_profiler_arm64	N/A	N/A → 317.144 → 328.970
✅	dogstatsd_deb_amd64	N/A	N/A → 30.224 → 31.200
✅	dogstatsd_deb_arm64	N/A	N/A → 28.247 → 29.580
✅	dogstatsd_rpm_amd64	N/A	N/A → 30.224 → 31.200
✅	dogstatsd_suse_amd64	N/A	N/A → 30.224 → 31.200
✅	iot_agent_deb_amd64	N/A	N/A → 46.467 → 47.530
✅	iot_agent_deb_arm64	N/A	N/A → 43.124 → 44.200
✅	iot_agent_deb_armhf	N/A	N/A → 43.907 → 44.990
✅	iot_agent_rpm_amd64	N/A	N/A → 46.468 → 47.530
✅	iot_agent_suse_amd64	N/A	N/A → 46.467 → 47.530

[cit-pr-commenter-54b7da]

Regression Detector

Regression Detector Results

Metrics dashboard
Target profiles
Job ID: 5c003e45-fd93-4720-8cb4-82ade4049b23

Baseline: 3b2688f
Comparison: d76074a
Diff

Optimization Goals: ✅ No significant changes detected

Fine details of change detection per experiment

perf	experiment	goal	Δ mean %	Δ mean % CI	trials	links
➖	quality_gate_security_idle	memory utilization	+0.24	[+0.13, +0.34]	1	Logs bounds checks dashboard
➖	quality_gate_security_mean_fs_load	memory utilization	+0.11	[+0.03, +0.18]	1	Logs bounds checks dashboard
➖	quality_gate_security_no_fs_load	memory utilization	+0.08	[-0.07, +0.23]	1	Logs bounds checks dashboard
➖	quality_gate_logs	% cpu utilization	-0.05	[-0.91, +0.81]	1	Logs bounds checks dashboard
➖	quality_gate_private_action_runner	memory utilization	-0.13	[-0.25, -0.01]	1	Logs bounds checks dashboard
➖	quality_gate_idle	memory utilization	-0.27	[-0.39, -0.15]	1	Logs bounds checks dashboard
➖	quality_gate_idle_all_features	memory utilization	-0.30	[-0.34, -0.26]	1	Logs bounds checks dashboard
➖	quality_gate_metrics_logs	memory utilization	-0.48	[-0.72, -0.23]	1	Logs bounds checks dashboard

Bounds Checks: ✅ Passed

perf	experiment	bounds_check_name	replicates_passed	observed_value	links
✅	quality_gate_idle	intake_connections	10/10	4 = 4	bounds checks dashboard
✅	quality_gate_idle	memory_usage	10/10	173.07MiB ≤ 178MiB	bounds checks dashboard
✅	quality_gate_idle	total_bytes_received	10/10	741.78KiB ≤ 819.20KiB	bounds checks dashboard
✅	quality_gate_idle_all_features	intake_connections	10/10	4 = 4	bounds checks dashboard
✅	quality_gate_idle_all_features	memory_usage	10/10	511.84MiB ≤ 538MiB	bounds checks dashboard
✅	quality_gate_idle_all_features	total_bytes_received	10/10	1.13MiB ≤ 1.25MiB	bounds checks dashboard
✅	quality_gate_logs	intake_connections	10/10	20 ≤ 40	bounds checks dashboard
✅	quality_gate_logs	memory_usage	10/10	206.56MiB ≤ 229MiB	bounds checks dashboard
✅	quality_gate_logs	missed_bytes	10/10	0B = 0B	bounds checks dashboard
✅	quality_gate_logs	total_bytes_received	10/10	264.13MiB ≤ 292MiB	bounds checks dashboard
✅	quality_gate_metrics_logs	cpu_usage	10/10	402.49 ≤ 2000	bounds checks dashboard
✅	quality_gate_metrics_logs	intake_connections	10/10	19 ≤ 40	bounds checks dashboard
✅	quality_gate_metrics_logs	memory_usage	10/10	395.18MiB ≤ 439MiB	bounds checks dashboard
✅	quality_gate_metrics_logs	missed_bytes	10/10	0B = 0B	bounds checks dashboard
✅	quality_gate_metrics_logs	total_bytes_received	10/10	0.94GiB ≤ 1.04GiB	bounds checks dashboard
✅	quality_gate_private_action_runner	memory_usage	10/10	72.00MiB ≤ 76MiB	bounds checks dashboard
✅	quality_gate_security_idle	cpu_usage	10/10	28.10 ≤ 100	bounds checks dashboard
✅	quality_gate_security_idle	memory_usage	10/10	324.21MiB ≤ 335MiB	bounds checks dashboard
✅	quality_gate_security_mean_fs_load	cpu_usage	10/10	73.43 ≤ 200	bounds checks dashboard
✅	quality_gate_security_mean_fs_load	memory_usage	10/10	303.28MiB ≤ 314MiB	bounds checks dashboard
✅	quality_gate_security_no_fs_load	cpu_usage	10/10	20.28 ≤ 100	bounds checks dashboard
✅	quality_gate_security_no_fs_load	memory_usage	10/10	310.51MiB ≤ 343MiB	bounds checks dashboard

Explanation

Confidence level: 90.00%
Effect size tolerance: |Δ mean %| ≥ 5.00%

Performance changes are noted in the perf column of each table:

• ✅ = significantly better comparison variant performance
• ❌ = significantly worse comparison variant performance
• ➖ = no significant change in performance

A regression test is an A/B test of target performance in a repeatable rig, where "performance" is measured as "comparison variant minus baseline variant" for an optimization goal (e.g., ingress throughput). Due to intrinsic variability in measuring that goal, we can only estimate its mean value for each experiment; we report uncertainty in that value as a 90.00% confidence interval denoted "Δ mean % CI".

For each experiment, we decide whether a change in performance is a "regression" -- a change worth investigating further -- if all of the following criteria are true:

1. Its estimated |Δ mean %| ≥ 5.00%, indicating the change is big enough to merit a closer look.

2. Its 90.00% confidence interval "Δ mean % CI" does not contain zero, indicating that if our statistical model is accurate, there is at least a 90.00% chance there is a difference in performance between baseline and comparison variants.

3. Its configuration does not mark it "erratic".

CI Pass/Fail Decision

✅ Passed. All Quality Gates passed.

• quality_gate_security_no_fs_load, bounds check memory_usage: 10/10 replicas passed. Gate passed.
• quality_gate_security_no_fs_load, bounds check cpu_usage: 10/10 replicas passed. Gate passed.
• quality_gate_private_action_runner, bounds check memory_usage: 10/10 replicas passed. Gate passed.
• quality_gate_metrics_logs, bounds check missed_bytes: 10/10 replicas passed. Gate passed.
• quality_gate_metrics_logs, bounds check total_bytes_received: 10/10 replicas passed. Gate passed.
• quality_gate_metrics_logs, bounds check cpu_usage: 10/10 replicas passed. Gate passed.
• quality_gate_metrics_logs, bounds check memory_usage: 10/10 replicas passed. Gate passed.
• quality_gate_metrics_logs, bounds check intake_connections: 10/10 replicas passed. Gate passed.
• quality_gate_idle, bounds check intake_connections: 10/10 replicas passed. Gate passed.
• quality_gate_idle, bounds check memory_usage: 10/10 replicas passed. Gate passed.
• quality_gate_idle, bounds check total_bytes_received: 10/10 replicas passed. Gate passed.
• quality_gate_idle_all_features, bounds check total_bytes_received: 10/10 replicas passed. Gate passed.
• quality_gate_idle_all_features, bounds check intake_connections: 10/10 replicas passed. Gate passed.
• quality_gate_idle_all_features, bounds check memory_usage: 10/10 replicas passed. Gate passed.
• quality_gate_security_idle, bounds check cpu_usage: 10/10 replicas passed. Gate passed.
• quality_gate_security_idle, bounds check memory_usage: 10/10 replicas passed. Gate passed.
• quality_gate_logs, bounds check total_bytes_received: 10/10 replicas passed. Gate passed.
• quality_gate_logs, bounds check intake_connections: 10/10 replicas passed. Gate passed.
• quality_gate_logs, bounds check memory_usage: 10/10 replicas passed. Gate passed.
• quality_gate_logs, bounds check missed_bytes: 10/10 replicas passed. Gate passed.
• quality_gate_security_mean_fs_load, bounds check memory_usage: 10/10 replicas passed. Gate passed.
• quality_gate_security_mean_fs_load, bounds check cpu_usage: 10/10 replicas passed. Gate passed.

[mabdinur]

@codex review

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 15670a0a79

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[mabdinur]

@codex review

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: fc6fae1c62

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Preserve the resource container ID through the APM hop

For Agent OTLP ingestion with a container.id resource attribute, this field is encoded into the client stats payload but serializerConsumer posts it to /v0.6/stats without a container-origin header. The trace agent's processStats then unconditionally replaces in.ContainerID with the ID resolved from that header/context, which is empty for this TCP request, so all converted containerized stats lose container attribution and the associated container tags; preserve the payload value when no origin ID is resolved or forward it in the request metadata.

Useful? React with 👍 / 👎.

[IbraheemA]

CreateDDSketchFromHistogramOfDuration is an outdated path that we only built for an earlier PoC (results in inaccurate histograms) - we've since added support for exact 1:1 histogram mapping, which is what otlp-intake-metrics does. make sure the payload is tagged with the correct version, and the stats pipeline should be able to accept it.

[IbraheemA]

There are a handful of small differences between the mapping logic here and that in otlp-intake-metrics:

• no guard on bucket counts
• different parsing for rpc.response.status_code
• "other_tags" used vs "primary_tags" in the intake
• etc.

One of the goals behind using golang to build the otlp-intake-* services was so that the opentelemetry-mapping-go library could be the source of truth for all OTLP->DD translation. Sorry to ask for another big refactor but instead of duplicating code here, I'd really push for you to move as much as you can of the common translation logic between the two into opentelemetry-mapping-go, which we'll then import in both. (along the lines of processTraceMetricsDatapoint in dd-source). I'm happy to pair with you on that effort.

[dd-octo-sts]

Go Package Import Differences

Baseline: 3b2688f
Comparison: d76074a

binary	os	arch	change
agent	linux	amd64	+1, -0 +github.com/DataDog/datadog-agent/pkg/opentelemetry-mapping-go/otlp/metrics/sdktracestats
agent	linux	arm64	+1, -0 +github.com/DataDog/datadog-agent/pkg/opentelemetry-mapping-go/otlp/metrics/sdktracestats
agent	windows	amd64	+1, -0 +github.com/DataDog/datadog-agent/pkg/opentelemetry-mapping-go/otlp/metrics/sdktracestats
agent	darwin	amd64	+1, -0 +github.com/DataDog/datadog-agent/pkg/opentelemetry-mapping-go/otlp/metrics/sdktracestats
agent	darwin	arm64	+1, -0 +github.com/DataDog/datadog-agent/pkg/opentelemetry-mapping-go/otlp/metrics/sdktracestats
agent	aix	ppc64	+1, -0 +github.com/DataDog/datadog-agent/pkg/opentelemetry-mapping-go/otlp/metrics/sdktracestats
heroku-agent	linux	amd64	+1, -0 +github.com/DataDog/datadog-agent/pkg/opentelemetry-mapping-go/otlp/metrics/sdktracestats
otel-agent	linux	amd64	+1, -0 +github.com/DataDog/datadog-agent/pkg/opentelemetry-mapping-go/otlp/metrics/sdktracestats
otel-agent	linux	arm64	+1, -0 +github.com/DataDog/datadog-agent/pkg/opentelemetry-mapping-go/otlp/metrics/sdktracestats