Disable service discovery in CI Visibility agentless mode

[daniel-mohedano]

What Does This Do

Disables service discovery by default when CI Visibility runs in agentless mode. Setting DD_TRACE_SERVICE_DISCOVERY_ENABLED=true still turns it back on.

Motivation

On Linux, 1 second after startup, service discovery loads JNA on a background thread to write the tracer metadata file. JNA adds new system properties (jnidispatch.path, jna.loaded, jna.platform.library.path).

On Java 8, system properties are a Hashtable with fail-fast iterators. Any application thread iterating them at that moment throws ConcurrentModificationException. The Gradle launcher does exactly that at startup, so Gradle builds instrumented with the tracer can fail with:

java.util.ConcurrentModificationException
  at java.util.Hashtable$Enumerator.next
  at org.gradle.launcher.cli.converter.LayoutToPropertiesConverter.configureFromSystemPropertiesOfThisJvm

This caused 13 failures of GradleLauncherSmokeTest on Java 8 on master over the last 3 weeks.

The metadata file is only read by an Agent running on the same host. In CI Visibility agentless mode there is usually no Agent, and the instrumented processes are short-lived build tools and test JVMs. So the file brings little value there.

Additional Notes

The race still exists for other Java 8 applications on Linux that iterate system properties around startup. That would need a separate fix in service discovery.

test-environment-trigger: skip

[cit-pr-commenter-54b7da]

CI Visibility Test Environment - sbt-scalatest

Job Status: 🟢 success

Scenario	This PR (%)	7d median	Δ 7d	30d median	Δ 30d	runs (7d/30d)
agent	54.44	57.69	$\color{green}{\blacktriangledown}$ -3.25	56.55	$\color{green}{\blacktriangledown}$ -2.11	64/312
agentEvpProxy	54.07	n/a	n/a	n/a	n/a	-

Baseline: median of @test.tracer_overhead on main (gitlab) over the last 7/30 days, per OSS project & scenario. Δ = this PR − baseline median; red ▲ = more overhead, green ▽ = less overhead than baseline.

Tests run to verify that CI Visibility behavior has not regressed in the current PR.

[cit-pr-commenter-54b7da]

CI Visibility Test Environment - nebula-release-plugin

Job Status: 🟢 success

Scenario	This PR (%)	7d median	Δ 7d	30d median	Δ 30d	runs (7d/30d)
agent	38.82	37.15	$\color{red}{\blacktriangle}$ +1.67	37.15	$\color{red}{\blacktriangle}$ +1.67	33/170
agentless	37.98	37.15	$\color{red}{\blacktriangle}$ +0.83	36.42	$\color{red}{\blacktriangle}$ +1.56	33/143
agentlessCodeCoverage	46.72	45.38	$\color{red}{\blacktriangle}$ +1.34	45.38	$\color{red}{\blacktriangle}$ +1.34	33/142
agentlessLineCoverage	58.01	56.55	$\color{red}{\blacktriangle}$ +1.46	56.55	$\color{red}{\blacktriangle}$ +1.46	32/141

Baseline: median of @test.tracer_overhead on main (gitlab) over the last 7/30 days, per OSS project & scenario. Δ = this PR − baseline median; red ▲ = more overhead, green ▽ = less overhead than baseline.

Tests run to verify that CI Visibility behavior has not regressed in the current PR.

[cit-pr-commenter-54b7da]

CI Visibility Test Environment - netflix-zuul

Job Status: 🟢 success

Scenario	This PR (%)	7d median	Δ 7d	30d median	Δ 30d	runs (7d/30d)
agent	87.16	91.39	$\color{green}{\blacktriangledown}$ -4.23	89.58	$\color{green}{\blacktriangledown}$ -2.42	33/171
agentless	81.04	82.69	$\color{green}{\blacktriangledown}$ -1.65	82.69	$\color{green}{\blacktriangledown}$ -1.65	33/141
agentlessCodeCoverage	95.49	99.00	$\color{green}{\blacktriangledown}$ -3.51	97.04	$\color{green}{\blacktriangledown}$ -1.55	33/141
agentlessLineCoverage	110.83	113.88	$\color{green}{\blacktriangledown}$ -3.05	113.88	$\color{green}{\blacktriangledown}$ -3.05	33/142

Baseline: median of @test.tracer_overhead on main (gitlab) over the last 7/30 days, per OSS project & scenario. Δ = this PR − baseline median; red ▲ = more overhead, green ▽ = less overhead than baseline.

Tests run to verify that CI Visibility behavior has not regressed in the current PR.

[cit-pr-commenter-54b7da]

CI Visibility Test Environment - pass4s

Job Status: 🟢 success

Scenario	This PR (%)	7d median	Δ 7d	30d median	Δ 30d	runs (7d/30d)
agent	6.96	10.75	$\color{green}{\blacktriangledown}$ -3.79	10.13	$\color{green}{\blacktriangledown}$ -3.17	35/166
agentless	6.70	10.33	$\color{green}{\blacktriangledown}$ -3.63	10.13	$\color{green}{\blacktriangledown}$ -3.43	32/134
agentlessCodeCoverage	14.29	18.08	$\color{green}{\blacktriangledown}$ -3.79	17.73	$\color{green}{\blacktriangledown}$ -3.44	28/122

Baseline: median of @test.tracer_overhead on main (gitlab) over the last 7/30 days, per OSS project & scenario. Δ = this PR − baseline median; red ▲ = more overhead, green ▽ = less overhead than baseline.

Tests run to verify that CI Visibility behavior has not regressed in the current PR.

[cit-pr-commenter-54b7da]

CI Visibility Test Environment - reactive-streams-jvm

Job Status: 🟢 success

Scenario	This PR (%)	7d median	Δ 7d	30d median	Δ 30d	runs (7d/30d)
agent	22.99	22.54	$\color{red}{\blacktriangle}$ +0.45	22.09	$\color{red}{\blacktriangle}$ +0.90	35/171
agentless	18.96	19.59	$\color{green}{\blacktriangledown}$ -0.63	19.59	$\color{green}{\blacktriangledown}$ -0.63	33/137
agentlessCodeCoverage	20.44	20.39	$\color{red}{\blacktriangle}$ +0.05	19.99	$\color{red}{\blacktriangle}$ +0.45	33/137
agentlessLineCoverage	27.03	27.52	$\color{green}{\blacktriangledown}$ -0.49	27.52	$\color{green}{\blacktriangledown}$ -0.49	20/93

Baseline: median of @test.tracer_overhead on main (gitlab) over the last 7/30 days, per OSS project & scenario. Δ = this PR − baseline median; red ▲ = more overhead, green ▽ = less overhead than baseline.

Tests run to verify that CI Visibility behavior has not regressed in the current PR.

[cit-pr-commenter-54b7da]

CI Visibility Test Environment - heliboard

Job Status: 🟢 success

Scenario	This PR (%)	7d median	Δ 7d	30d median	Δ 30d	runs (7d/30d)
agent	9.05	9.92	$\color{green}{\blacktriangledown}$ -0.87	9.92	$\color{green}{\blacktriangledown}$ -0.87	32/139

Baseline: median of @test.tracer_overhead on main (gitlab) over the last 7/30 days, per OSS project & scenario. Δ = this PR − baseline median; red ▲ = more overhead, green ▽ = less overhead than baseline.

Tests run to verify that CI Visibility behavior has not regressed in the current PR.

[dd-octo-sts]

🟢 Java Benchmark SLOs — All performance SLOs passed

Suite	Status
Startup	🟢 pass

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	14.76 s	14.65 s	[+0.1%; +1.3%] (maybe worse)
startup:insecure-bank:tracing:Agent	13.64 s	13.65 s	[-0.9%; +0.7%] (no difference)
startup:petclinic:appsec:Agent	17.05 s	16.41 s	[-0.6%; +8.4%] (no difference)
startup:petclinic:iast:Agent	16.87 s	16.88 s	[-1.0%; +0.8%] (no difference)
startup:petclinic:profiling:Agent	16.81 s	16.60 s	[+0.3%; +2.3%] (maybe worse)
startup:petclinic:sca:Agent	16.98 s	16.88 s	[-0.4%; +1.6%] (no difference)
startup:petclinic:tracing:Agent	15.59 s	16.08 s	[-7.1%; +1.0%] (no difference)

Commit: baefd31a · 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.

[cit-pr-commenter-54b7da]

CI Visibility Test Environment - sonar-kotlin

Job Status: 🟢 success

Scenario	This PR (%)	7d median	Δ 7d	30d median	Δ 30d	runs (7d/30d)
agent	11.65	13.67	$\color{green}{\blacktriangledown}$ -2.02	13.40	$\color{green}{\blacktriangledown}$ -1.75	35/171
agentless	10.71	12.12	$\color{green}{\blacktriangledown}$ -1.41	12.12	$\color{green}{\blacktriangledown}$ -1.41	35/141
agentlessCodeCoverage	14.85	15.72	$\color{green}{\blacktriangledown}$ -0.87	15.41	$\color{green}{\blacktriangledown}$ -0.56	34/140
agentlessLineCoverage	16.56	17.38	$\color{green}{\blacktriangledown}$ -0.82	17.38	$\color{green}{\blacktriangledown}$ -0.82	33/139

Baseline: median of @test.tracer_overhead on main (gitlab) over the last 7/30 days, per OSS project & scenario. Δ = this PR − baseline median; red ▲ = more overhead, green ▽ = less overhead than baseline.

Tests run to verify that CI Visibility behavior has not regressed in the current PR.

[cit-pr-commenter-54b7da]

CI Visibility Test Environment - jolokia

Job Status: 🟢 success

Scenario	This PR (%)	7d median	Δ 7d	30d median	Δ 30d	runs (7d/30d)
agent	94.39	95.12	$\color{green}{\blacktriangledown}$ -0.73	95.12	$\color{green}{\blacktriangledown}$ -0.73	36/172
agentless	90.22	91.39	$\color{green}{\blacktriangledown}$ -1.17	89.58	$\color{red}{\blacktriangle}$ +0.64	35/143
agentlessCodeCoverage	97.59	101.00	$\color{green}{\blacktriangledown}$ -3.41	99.00	$\color{green}{\blacktriangledown}$ -1.41	34/142
agentlessLineCoverage	100.28	103.04	$\color{green}{\blacktriangledown}$ -2.76	101.00	$\color{green}{\blacktriangledown}$ -0.72	39/150

Baseline: median of @test.tracer_overhead on main (gitlab) over the last 7/30 days, per OSS project & scenario. Δ = this PR − baseline median; red ▲ = more overhead, green ▽ = less overhead than baseline.

Tests run to verify that CI Visibility behavior has not regressed in the current PR.

[cit-pr-commenter-54b7da]

CI Visibility Test Environment - okhttp

Job Status: 🟢 success

Scenario	This PR (%)	7d median	Δ 7d	30d median	Δ 30d	runs (7d/30d)
agent	19.15	19.20	$\color{green}{\blacktriangledown}$ -0.05	19.99	$\color{green}{\blacktriangledown}$ -0.84	35/169
agentless	18.15	18.45	$\color{green}{\blacktriangledown}$ -0.30	19.20	$\color{green}{\blacktriangledown}$ -1.05	36/145
agentlessCodeCoverage	20.45	21.65	$\color{green}{\blacktriangledown}$ -1.20	22.99	$\color{green}{\blacktriangledown}$ -2.54	37/147
agentlessLineCoverage	37.37	37.91	$\color{green}{\blacktriangledown}$ -0.54	39.45	$\color{green}{\blacktriangledown}$ -2.08	33/141

Baseline: median of @test.tracer_overhead on main (gitlab) over the last 7/30 days, per OSS project & scenario. Δ = this PR − baseline median; red ▲ = more overhead, green ▽ = less overhead than baseline.

Tests run to verify that CI Visibility behavior has not regressed in the current PR.

[cit-pr-commenter-54b7da]

CI Visibility Test Environment - spring_boot

Job Status: 🟢 success

Scenario	This PR (%)	7d median	Δ 7d	30d median	Δ 30d	runs (7d/30d)
agent	15.96	16.36	$\color{green}{\blacktriangledown}$ -0.40	16.36	$\color{green}{\blacktriangledown}$ -0.40	35/167
agentless	10.49	9.92	$\color{red}{\blacktriangle}$ +0.57	9.92	$\color{red}{\blacktriangle}$ +0.57	35/140
agentlessCodeCoverage	14.36	13.67	$\color{red}{\blacktriangle}$ +0.69	13.67	$\color{red}{\blacktriangle}$ +0.69	35/142
agentlessLineCoverage	23.58	22.54	$\color{red}{\blacktriangle}$ +1.04	22.54	$\color{red}{\blacktriangle}$ +1.04	32/138

Baseline: median of @test.tracer_overhead on main (gitlab) over the last 7/30 days, per OSS project & scenario. Δ = this PR − baseline median; red ▲ = more overhead, green ▽ = less overhead than baseline.

Tests run to verify that CI Visibility behavior has not regressed in the current PR.

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-16T10:54:27.827872Z	baefd31	Draft marked ready
🔒 Security Review	✅ Completed 2026-10-16T10:54:26.489135Z	baefd31	Draft marked ready

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" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[datadog-prod-us1-5]

More details

The new default disables service discovery only when CI Visibility and agentless mode are both enabled, while preserving explicit overrides. The completed static review found no actionable regressions.

Was this helpful? React 👍 or 👎

_{Open Bits AI session}

_{🤖 Bits Code Review · Commit baefd31}

[cit-pr-commenter-54b7da]

CI Visibility Test Environment - sonar-java

Job Status: 🟢 success

Scenario	This PR (%)	7d median	Δ 7d	30d median	Δ 30d	runs (7d/30d)
agent	44.96	17.38	$\color{red}{\blacktriangle}$ +27.58	15.41	$\color{red}{\blacktriangle}$ +29.55	35/169
agentless	41.46	13.94	$\color{red}{\blacktriangle}$ +27.52	13.13	$\color{red}{\blacktriangle}$ +28.33	35/141
agentlessCodeCoverage	71.91	79.45	$\color{green}{\blacktriangledown}$ -7.54	79.45	$\color{green}{\blacktriangledown}$ -7.54	35/142
agentlessLineCoverage	109.61	123.36	$\color{green}{\blacktriangledown}$ -13.75	123.36	$\color{green}{\blacktriangledown}$ -13.75	31/141

Baseline: median of @test.tracer_overhead on main (gitlab) over the last 7/30 days, per OSS project & scenario. Δ = this PR − baseline median; red ▲ = more overhead, green ▽ = less overhead than baseline.

Tests run to verify that CI Visibility behavior has not regressed in the current PR.