Check the DBM comment region in place, dropping the duplicate-comment substring by dougqh · Pull Request #11737 · DataDog/dd-trace-java

dougqh · 2026-06-24T22:57:47Z

What Does This Do

Replaces the per-call sql.substring used by SQLCommenter.hasDDComment.
In place of a sub-string, the code now uses a SubSequence which has lower overhead

Motivation

SubSequence avoids the allocation and copying of a regular sub-string.
And frequently, SubSequence is completely optimized away by escape analysis.

For this code path, the net effect is zero allocation and a modest ~12% improvement in throughput.

Additional Notes

Measured — SQLCommenterDuplicateCommentBenchmark (JDK 17, @Threads(8), @Fork(5), -prof gc)

	throughput	`gc.alloc.rate.norm`
before (substring)	23.5M ± 1.1M ops/s	140 B/op
after (range/view)	26.2M ± 1.5M ops/s	~0 (10⁻⁵)

The allocation delta (140 → 0 B/op) is the win and is exact/fork-stable. This path is scan-CPU-dominated, so the ~12% throughput is a modest bonus — the real, compounding win is the per-inject allocation removed across comment-bearing queries.

🤖 Generated with Claude Code

dd-octo-sts · 2026-06-24T23:20:45Z

🟢 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	13.96 s	14.05 s	[-1.3%; +0.0%] (no difference)
startup:insecure-bank:tracing:Agent	12.94 s	13.02 s	[-1.4%; +0.2%] (no difference)
startup:petclinic:appsec:Agent	16.37 s	16.70 s	[-6.4%; +2.4%] (no difference)
startup:petclinic:iast:Agent	16.40 s	16.91 s	[-7.2%; +1.2%] (no difference)
startup:petclinic:profiling:Agent	16.85 s	16.33 s	[-1.1%; +7.4%] (no difference)
startup:petclinic:sca:Agent	16.81 s	16.59 s	[+0.2%; +2.5%] (maybe worse)
startup:petclinic:tracing:Agent	15.59 s	16.09 s	[-7.1%; +1.0%] (no difference)

Commit: 2eeff802 · 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.

dougqh · 2026-06-25T11:35:29Z

Benchmark results (zulu-17, JDK 17.0.7, @Threads(8), @Fork(5), -prof gc)

SQLCommenterDuplicateCommentBenchmark.alreadyCommented:

	throughput	gc.alloc.rate.norm
before (substring)	23.5M ± 1.1M ops/s	140 B/op
after (range/view)	26.2M ± 1.5M ops/s	~0 B/op (10⁻⁵)

The extractCommentContent substring (140 B/op) is gone — the in-place range scan and the SubSequence view it flows through are both EA-elided. The headline win is the allocation: this path is dominated by the nine indexOf scans (CPU the alloc-removal doesn't touch), so throughput is a modest ~1.1× uplift, now resolved at @Fork(5). Benchmark Javadoc refreshed in bd0983f.

…ld the view Per review on #11737: move SubSequence.of to the call sites (SQLCommenter builds the comment-region view; the String overload wraps the whole content) and make SharedDBCommenter.containsTraceComment take a SubSequence instead of (String, from, to). Behavior unchanged (SubSequence.contains -> Strings.regionContains); test + dup-comment-benchmark Javadoc updated to the new shape. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

…tent substring SQLCommenter.hasDDComment materialized sql.substring(commentStart, commentEnd) on every duplicate-comment check just to feed containsTraceComment. Add a SharedDBCommenter.containsTraceComment(sql, from, to) range overload that scans the comment body in place, and a Strings.regionContains primitive it (and the String delegate) build on -- no per-call substring. The String overload now delegates to the range form, so Mongo behavior is unchanged. regionContains is the allocation-free, copy-free primitive; the natural-reading SubSequence.contains layer can delegate to it later. Boundary semantics unit- tested on Strings.regionContains; DB needle behavior on the overloads; existing SQLCommenter/SharedDBCommenter suites unchanged and green. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

…diom) Add SubSequence.contains (delegating to the Strings.regionContains primitive) and rewrite containsTraceComment(sql, from, to) as SubSequence.of(sql, from, to).contains(...) -- a 1-to-1 substitution for what you'd idiomatically write on a substring, with no copy. EA elides the view in the transient consumption; even if it doesn't, the string copy is still avoided. Couples this branch to the SubSequence base (#10640); regionContains stays as the allocation-free primitive the view delegates to. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

@threads

Isolates the duplicate-comment guard (dbType=null skips the first-word scan; already-DD-commented SQL makes inject return early). The extractCommentContent substring allocated 140 B/op; the in-place range/view scan is EA-elided (~0). @threads(8), @fork(2), -prof gc; numbers in the class Javadoc. Throughput is flat-to-slightly-up but within @fork(2) noise -- this path is scan-CPU-dominated, so the win is the allocation, not throughput. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

zulu-17 @fork(5), -prof gc: 23.5M -> 26.2M ops/s (~1.1x), 140 -> ~0 B/op. Headline win is the allocation; @fork(5) tightens the earlier bimodal spread. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

…scan-overload # Conflicts: # internal-api/src/main/java/datadog/trace/util/SubSequence.java # internal-api/src/test/java/datadog/trace/util/SubSequenceTest.java

datadog-official · 2026-06-30T21:40:18Z

🎯 Code Coverage (details)
• Patch Coverage: 57.14%
• Overall Coverage: 56.72% (-0.46%)

_{This comment will be updated automatically if new data arrives.

🔗 Commit SHA: 2eeff80 | Docs | Datadog PR Page | Give us feedback!}

mhlidd

Nice improvements!

dd-octo-sts · 2026-07-14T15:29:57Z

/merge

gh-worker-devflow-routing-ef8351 · 2026-07-14T15:30:01Z

View all feedbacks in Devflow UI.

2026-07-14 15:30:01 UTC ℹ️ Start processing command /merge

2026-07-14 15:30:07 UTC ℹ️ MergeQueue: pull request added to the queue

The expected merge time in master is approximately 2h (p90).

2026-07-14 16:31:49 UTC ℹ️ MergeQueue: This merge request was merged

dougqh added tag: performance Performance related changes tag: no release notes Changes to exclude from release notes inst: jdbc JDBC instrumentation comp: database Database Monitoring tag: ai generated Largely based on code generated by an AI or LLM labels Jun 24, 2026

dougqh force-pushed the dougqh/dbcommenter-scan-overload branch from bd0983f to 24d8ea0 Compare June 29, 2026 18:16

dougqh commented Jun 29, 2026

View reviewed changes

Comment thread ...gent/instrumentation/jdbc/src/main/java/datadog/trace/instrumentation/jdbc/SQLCommenter.java

dougqh commented Jun 29, 2026

View reviewed changes

Comment thread ...t-bootstrap/src/main/java/datadog/trace/bootstrap/instrumentation/dbm/SharedDBCommenter.java

dougqh marked this pull request as ready for review June 29, 2026 19:53

dougqh requested review from a team as code owners June 29, 2026 19:53

dougqh requested review from jordan-wong and mhlidd and removed request for a team June 29, 2026 19:53

dougqh and others added 4 commits June 29, 2026 20:54

Refresh SQLCommenterDuplicateCommentBenchmark results to JDK 17 @fork(5)

5904552

zulu-17 @fork(5), -prof gc: 23.5M -> 26.2M ops/s (~1.1x), 140 -> ~0 B/op. Headline win is the allocation; @fork(5) tightens the earlier bimodal spread. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

dougqh force-pushed the dougqh/dbcommenter-scan-overload branch from 71f3df8 to 5904552 Compare June 30, 2026 00:56

dougqh mentioned this pull request Jun 30, 2026

Add window-bounded String and char overloads to SubSequence #11796

Open

Merge remote-tracking branch 'origin/master' into dougqh/dbcommenter-…

3b96fdc

…scan-overload # Conflicts: # internal-api/src/main/java/datadog/trace/util/SubSequence.java # internal-api/src/test/java/datadog/trace/util/SubSequenceTest.java

na-ji approved these changes Jul 1, 2026

View reviewed changes

dougqh enabled auto-merge July 1, 2026 19:29

Merge branch 'master' into dougqh/dbcommenter-scan-overload

0e91753

mhlidd approved these changes Jul 9, 2026

View reviewed changes

Merge branch 'master' into dougqh/dbcommenter-scan-overload

2eeff80

dougqh added this pull request to the merge queue Jul 14, 2026

github-merge-queue Bot removed this pull request from the merge queue due to failed status checks Jul 14, 2026

gh-worker-dd-mergequeue-cf854d Bot merged commit 7e08161 into master Jul 14, 2026
587 checks passed

gh-worker-dd-mergequeue-cf854d Bot deleted the dougqh/dbcommenter-scan-overload branch July 14, 2026 16:31

github-actions Bot added this to the 1.65.0 milestone Jul 14, 2026

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Check the DBM comment region in place, dropping the duplicate-comment substring#11737

Check the DBM comment region in place, dropping the duplicate-comment substring#11737
gh-worker-dd-mergequeue-cf854d[bot] merged 7 commits into
masterfrom
dougqh/dbcommenter-scan-overload

dougqh commented Jun 24, 2026 •

edited

Loading

Uh oh!

dd-octo-sts Bot commented Jun 24, 2026 •

edited

Loading

Uh oh!

dougqh commented Jun 25, 2026 •

edited

Loading

Uh oh!

Uh oh!

Uh oh!

datadog-official Bot commented Jun 30, 2026 •

edited

Loading

Uh oh!

mhlidd left a comment

Uh oh!

dd-octo-sts Bot commented Jul 14, 2026

Uh oh!

gh-worker-devflow-routing-ef8351 Bot commented Jul 14, 2026 •

edited

Loading

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Uh oh!

Conversation

dougqh commented Jun 24, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

What Does This Do

Motivation

Additional Notes

Uh oh!

dd-octo-sts Bot commented Jun 24, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

🟢 Java Benchmark SLOs — All performance SLOs passed

Uh oh!

dougqh commented Jun 25, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

Uh oh!

datadog-official Bot commented Jun 30, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

mhlidd left a comment

Choose a reason for hiding this comment

Uh oh!

dd-octo-sts Bot commented Jul 14, 2026

Uh oh!

gh-worker-devflow-routing-ef8351 Bot commented Jul 14, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

dougqh commented Jun 24, 2026 •

edited

Loading

dd-octo-sts Bot commented Jun 24, 2026 •

edited

Loading

dougqh commented Jun 25, 2026 •

edited

Loading

datadog-official Bot commented Jun 30, 2026 •

edited

Loading

gh-worker-devflow-routing-ef8351 Bot commented Jul 14, 2026 •

edited

Loading