Decouple Kafka DSM instrumentation from APM tracing enablement by ericfirth · Pull Request #12350 · DataDog/dd-trace-java

ericfirth · 2026-08-31T18:43:56Z

Summary

Switches Kafka clients (0.11, 3.8) and kafka-streams-0.11 instrumentation modules to InstrumenterModule.DataStreams, so DSM checkpoints are created even when APM tracing/integrations are disabled.
Forces DSM-only produce/consume/poll spans to PrioritySampling.USER_DROP via SamplingMechanism.DATA_STREAMS when they are the local root span, so they aren't billed as APM — while never dropping an ambient customer trace.
Dedupes each KafkaDecorator's hardcoded integration-name list into shared constants reused by TRACING_ENABLED and every module constructor, delegates Config.isDataStreamsEnabled() to InstrumenterConfig, extracts a duplicated DSM-drop guard in KafkaStreamTaskInstrumentation into a helper, and rewrites InstrumenterModuleTest to use @WithConfig instead of raw reflection.

Motivation

Using dd-trace-rb with JRuby doesn't support DSM. This explores whether Java's DSM instrumentation can run standalone (independent of APM tracing) as a path back to DSM support when tracing is handled by a different tracer, such as dd-trace-rb

Jira ticket: [DSM2-278]

Test plan

./gradlew :dd-java-agent:instrumentation:kafka-clients-0.11:test
./gradlew :dd-java-agent:instrumentation:kafka-clients-3.8:test
./gradlew :dd-java-agent:instrumentation:kafka-streams-0.11:test
./gradlew :internal-api:test
./gradlew :dd-java-agent:agent-tooling:test
Manual verification against a real Java Kafka client test app with DSM enabled / APM tracing disabled
Techdebt and perf review passes run over the branch; findings fixed

🤖 Generated with Claude Code

dd-octo-sts · 2026-08-31T19:14:12Z

🟢 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	14.83 s	14.73 s	[-0.5%; +1.8%] (no difference)
startup:insecure-bank:tracing:Agent	13.68 s	13.71 s	[-1.1%; +0.5%] (no difference)
startup:petclinic:appsec:Agent	17.64 s	17.32 s	[+0.9%; +2.7%] (maybe worse)
startup:petclinic:iast:Agent	17.57 s	17.64 s	[-1.5%; +0.7%] (no difference)
startup:petclinic:profiling:Agent	16.66 s	17.42 s	[-8.4%; -0.3%] (maybe better)
startup:petclinic:sca:Agent	17.59 s	17.51 s	[-0.6%; +1.4%] (no difference)
startup:petclinic:tracing:Agent	16.58 s	16.29 s	[-2.5%; +6.1%] (no difference)

Commit: be3a989d · 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.

pr-commenter · 2026-08-31T19:41:06Z

Kafka / consumer-benchmark

Parameters

	Baseline	Candidate
baseline_or_candidate	baseline	candidate
git_branch	master	worktree-dsm-kafka-decoupling
git_commit_date	1789566771	1789673168
git_commit_sha	`c595628`	`be3a989`

See matching parameters

	Baseline	Candidate
ci_job_date	1789674756	1789674756
ci_job_id	2055294161	2055294161
ci_pipeline_id	138296209	138296209
cpu_model	Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz	Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz
jdkVersion	11.0.31	11.0.31
jmhVersion	1.36	1.36
jvm	/usr/lib/jvm/java-11-openjdk-amd64/bin/java	/usr/lib/jvm/java-11-openjdk-amd64/bin/java
jvmArgs	-Dhttp.proxyHost=127.0.0.1 -Dhttp.proxyPort=15002 -Dhttps.proxyHost=127.0.0.1 -Dhttps.proxyPort=15002 -Dhttp.nonProxyHosts=localhost	*.localhost
kernel_version	Linux runner-zfyrx7zua-project-304-concurrent-0-sinxfg73 6.8.0-1031-aws #33~22.04.1-Ubuntu SMP Thu Jun 26 14:22:30 UTC 2025 x86_64 x86_64 x86_64 GNU/Linux	Linux runner-zfyrx7zua-project-304-concurrent-0-sinxfg73 6.8.0-1031-aws #33~22.04.1-Ubuntu SMP Thu Jun 26 14:22:30 UTC 2025 x86_64 x86_64 x86_64 GNU/Linux
vmName	OpenJDK 64-Bit Server VM	OpenJDK 64-Bit Server VM
vmVersion	11.0.31+11-post-1ubuntu1-22.04.2-Ubuntu	11.0.31+11-post-1ubuntu1-22.04.2-Ubuntu

Summary

Found 0 performance improvements and 0 performance regressions! Performance is the same for 3 metrics, 0 unstable metrics.

See unchanged results

scenario	Δ mean throughput
scenario:not-instrumented/KafkaConsumerBenchmark.benchConsume	same
scenario:only-tracing-dsm-disabled-benchmarks/KafkaConsumerBenchmark.benchConsume	same
scenario:only-tracing-dsm-enabled-benchmarks/KafkaConsumerBenchmark.benchConsume	unsure [+767.743op/s; +9862.954op/s] or [+0.415%; +5.332%]

dougqh · 2026-08-31T20:07:43Z

Heads up: the TRACING_CONFIG_NONDEFAULT system-tests scenario (main / End-to-end #8 / spring-boot 8) is currently failing because of this PR's core behavior change:

FAILED tests/test_config_consistency.py::Test_Config_IntegrationEnabled_False::test_integration_enabled_false
AssertionError: kafka.produce span was found in trace: [...]

That test disables the Kafka integration and asserts no kafka.produce span should reach the agent at all — the pre-PR contract. With this PR, when DSM is enabled, a kafka.produce/kafka.consume span is still created and sent (now carrying sampling_priority=USER_DROP) even though Kafka tracing itself is disabled — that's the intended decoupling. The scenario has DSM traffic (api_v2_data_streams_messages.json present in its logs), so this is a genuine interaction, not a flake.

This will need a system-tests-side update before merge: either exempt/adjust Test_Config_IntegrationEnabled_False for the DSM-enabled case, or have the scenario disable DSM alongside the integration so the old assertion still holds where DSM isn't in play. Worth confirming with whoever owns that shared parametric test (it's cross-language) before changing its semantics unilaterally.

pr-commenter · 2026-09-04T17:26:19Z

Kafka / producer-benchmark

Parameters

	Baseline	Candidate
baseline_or_candidate	baseline	candidate
git_branch	master	worktree-dsm-kafka-decoupling
git_commit_date	1789566771	1789673168
git_commit_sha	`c595628`	`be3a989`

See matching parameters

	Baseline	Candidate
ci_job_date	1789674712	1789674712
ci_job_id	2055294159	2055294159
ci_pipeline_id	138296209	138296209
cpu_model	Intel(R) Xeon(R) Platinum 8175M CPU @ 2.50GHz	Intel(R) Xeon(R) Platinum 8175M CPU @ 2.50GHz
jdkVersion	11.0.31	11.0.31
jmhVersion	1.36	1.36
jvm	/usr/lib/jvm/java-11-openjdk-amd64/bin/java	/usr/lib/jvm/java-11-openjdk-amd64/bin/java
jvmArgs	-Dhttp.proxyHost=127.0.0.1 -Dhttp.proxyPort=15002 -Dhttps.proxyHost=127.0.0.1 -Dhttps.proxyPort=15002 -Dhttp.nonProxyHosts=localhost	*.localhost
kernel_version	Linux runner-zfyrx7zua-project-304-concurrent-0-nsqnza35 6.8.0-1031-aws #33~22.04.1-Ubuntu SMP Thu Jun 26 14:22:30 UTC 2025 x86_64 x86_64 x86_64 GNU/Linux	Linux runner-zfyrx7zua-project-304-concurrent-0-nsqnza35 6.8.0-1031-aws #33~22.04.1-Ubuntu SMP Thu Jun 26 14:22:30 UTC 2025 x86_64 x86_64 x86_64 GNU/Linux
vmName	OpenJDK 64-Bit Server VM	OpenJDK 64-Bit Server VM
vmVersion	11.0.31+11-post-1ubuntu1-22.04.2-Ubuntu	11.0.31+11-post-1ubuntu1-22.04.2-Ubuntu

Summary

Found 0 performance improvements and 0 performance regressions! Performance is the same for 3 metrics, 0 unstable metrics.

See unchanged results

scenario	Δ mean throughput
scenario:not-instrumented/KafkaProduceBenchmark.benchProduce	same
scenario:only-tracing-dsm-disabled-benchmarks/KafkaProduceBenchmark.benchProduce	unsure [-7302.610op/s; -268.767op/s] or [-4.171%; -0.154%]
scenario:only-tracing-dsm-enabled-benchmarks/KafkaProduceBenchmark.benchProduce	unsure [-4576.868op/s; -948.949op/s] or [-3.092%; -0.641%]

Implements the pre-instrumentation config accessor for the data streams feature flag, mirroring the existing usmEnabled pattern. Reads the same data.streams.enabled key that Config.isDataStreamsEnabled() uses, enabling DSM to work independently of the general DD_INTEGRATIONS_ENABLED APM gate. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

… class

…ledOverridesFalse Save and restore the original dataStreamsEnabled value instead of hardcoding false in the finally block. This prevents test pollution where an incorrect state could leak to the next test.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

Add DATA_STREAMS sampling mechanism (byte value 14) to support decoupling Kafka DSM from APM tracing. This mechanism enforces USER_DROP priority for DSM-only spans and allows overriding locked priority when DSM is enabled. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

Use qualified PrioritySampling.USER_DROP form in validateWithSamplingPriority to match the DATA_JOBS case style (line 70) which uses qualified form. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

…reams base class Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

When Kafka tracing is disabled but Data Streams Monitoring is enabled, produce-path spans created without a propagated trace context should be marked with PrioritySampling.USER_DROP to avoid billing as APM spans. This implements the 'local-root spans only' scoping from the spec: spans with an extracted parent context (genuine distributed traces) are left completely untouched. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

…eams base class Change extends InstrumenterModule.Tracing to extends InstrumenterModule.DataStreams for all 8 instrumentation classes in the kafka-clients-3.8 module. This decouples Data Streams Monitoring from APM tracing and prepares for billing-suppression logic. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

…me paths) Add TRACING_ENABLED constant and billing-suppression logic for produce and consume paths to suppress unnecessary billing when APM tracing is disabled but DSM is enabled. The suppression only applies to local-root spans (no extracted parent context) per the DSM-Kafka decoupling plan. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

Ensures the cross-language contract holds: when APM tracing is disabled and DSM is enabled, no span is ever written to the agent for Kafka instrumentations. Replaces AgentTracer.startSpan with a lightweight Utils.newPathwayOnlySpan shim across kafka-clients-0.11, kafka-clients-3.8, and kafka-streams-0.11, and extracts shared DSM checkpoint/inject and transaction-tracking logic into helper methods to remove duplication. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

… mechanism InstrumenterConfig.java unconditionally disabled dataStreamsEnabled under GraalVM native-image, unlike its pre-PR behavior; DSM has no native-image restriction, so its read is moved outside that carve-out to match dataJobsEnabled. SamplingMechanism.DATA_STREAMS forced a USER_DROP sampling priority to keep DSM-only Kafka spans from being billed as APM traces. That approach is now fully superseded by Utils.newPathwayOnlySpan, which never creates a real span for the DSM-only case, so DATA_STREAMS had zero remaining production callers and is removed as dead code. Also collapses the near-identical TRACING_ENABLED computation duplicated across the three Kafka decorators (kafka-clients-0.11, kafka-clients-3.8, kafka-streams-0.11) into a single Utils.isTracingEnabled helper in kafka-common, per techdebt review. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

Several kafka-clients instrumentations referenced KafkaDecorator (whose TRACING_ENABLED field now calls Utils.isTracingEnabled) without declaring kafka_common.Utils as a helper class, causing muzzle to fail resolving that dependency. Separately, kafka-streams' StartSpanAdvice/StartSpanAdvice27 advice classes contained their own private static helper methods, which muzzle cannot validate since advice classes aren't part of the checked helper set. Moved that logic into KafkaStreamsDecorator, which is already a registered helper class. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

dougqh · 2026-09-16T14:26:26Z

Claude (re-review update): Checked back in on this since my 08-31 review — the three substantive findings from back then are all resolved via a genuine redesign rather than a patch:

Native-image regression (InstrumenterConfig.isDataStreamsEnabled()): now read unconditionally outside the Platform.isNativeImageBuilder() carve-out, with a comment explaining why DSM doesn't need that restriction.
Hardcoded USER_DROP guard duplicated 4x: replaced by shared TRACING_ENABLED/INTEGRATION_NAME/LEGACY_INTEGRATION_NAME constants each instrumentation's constructor now reuses.
TracingIterator / static-vs-dynamic isDataStreamsEnabled() mismatch: moot — the DSM-only path no longer creates a real span or forces sampling priority at all. It now uses a newPathwayOnlySpan shim that only carries the pathway context, so it doesn't route through CoreSpanBuilder.startSpan (and my original propagation-behavior concern with it) anymore.

Two other things worth noting for reviewers:

"Enforce Groovy migration" check is failing on the 3 new .groovy files, but all of them extend InstrumentationSpecification (genuine instrumented tests against real embedded Kafka) — the sanctioned Groovy/Spock exception per this repo's testing conventions. This looks like a false positive from the blanket new-file check rather than a real issue; the tag: override groovy enforcement label should clear it.
The TRACING_CONFIG_NONDEFAULT system-tests scenario I flagged on 08-31 has been green since the 09-09 14:37 system-tests run — that's fixed. A separate play 5 end-to-end failure showed up in the 09-09 15:21 run; unclear if it's related to the muzzle-validation fix pushed right before it or just transient — worth a glance if it recurs.

Scope has grown since my last pass to also cover kafka-streams-1.0 and kafka-connect-0.11.

BC_IMPOSSIBLE_INSTANCEOF flags the Object-cast instanceof check used to detect RecordMetadata support (only available from kafka-streams 2.7+), which SpotBugs can't reason about statically but is always true at runtime given this class's version constraint. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

…-trace-java into worktree-dsm-kafka-decoupling

chatgpt-codex-connector · 2026-09-17T17:01:01Z

Codex Review Summary

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

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-09-17T17:10:42.536272Z	`820ab35`	Draft marked ready
🔒 Security Review	✅ Completed 2026-09-17T17:07:34.048320Z	`820ab35`	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-datadog-prod-us1

The two most critical issues are: (1) AgentInstaller never installs CoreTracer when tracing is disabled but DSM is enabled, so every DSM checkpoint is silently dropped via the no-op implementation; and (2) the DSM-only span/carrier zeroes out trace and span IDs, severing continuity for any active customer trace and its producer callbacks.

_{Open Bits AI session}

_{🤖 Datadog Autotest · Commit 820ab35 · What is Autotest? · @DataDog review to ask questions · Any feedback? Reach out in #autotest}

chatgpt-codex-connector

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 820ab35e6d

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".

Several branch-selection checks used the dynamic, remote-config-aware traceConfig().isDataStreamsEnabled() alongside the static TRACING_ENABLED flag when deciding whether to create a real APM span vs. a DSM-only pathway shim. That let toggling DSM off at runtime silently re-enable real Kafka spans that should stay disabled. Also fixes producer callbacks losing the active trace, and restores strict span-write ordering between broker/queue spans and their consume-span children. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

…tale span If an exception was thrown between startTracedConsumeSpan setting pendingQueueSpanToFinish and it being finished/nulled in startNewRecordSpan, the field stayed set and a later, unrelated call would finish that stale queue span at the wrong time. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

dougqh

Overall, I think this looks like a good start at decoupled DSM & tracing.

I have once concern around clearing pendingSpan on an exception in the TracingIterators.

mcculls · 2026-09-18T12:51:05Z

This is exploratory — tracked in DSM2-278. It may or may not be accepted upstream.

@ericfirth is this still exploratory, or is this something you'd like to integrate before the next release?

ericfirth · 2026-09-18T13:40:35Z

This is exploratory — tracked in DSM2-278. It may or may not be accepted upstream.

@ericfirth is this still exploratory, or is this something you'd like to integrate before the next release?

@mcculls No, not really exploratory anymore. We'd like to merge it

ericfirth added type: feature Enhancements and improvements inst: kafka Kafka instrumentation comp: data streams Data Streams Monitoring tag: ai generated Largely based on code generated by an AI or LLM labels Aug 31, 2026

This comment has been minimized.

Sign in to view

dougqh reviewed Aug 31, 2026

View reviewed changes

Comment thread ...-clients-0.11/src/main/java/datadog/trace/instrumentation/kafka_clients/TracingIterator.java Outdated

dougqh reviewed Aug 31, 2026

View reviewed changes

Comment thread internal-api/src/main/java/datadog/trace/api/sampling/SamplingMechanism.java Outdated

dougqh reviewed Aug 31, 2026

View reviewed changes

Comment thread internal-api/src/main/java/datadog/trace/api/InstrumenterConfig.java Outdated

dougqh reviewed Aug 31, 2026

View reviewed changes

Comment thread ...a-clients-0.11/src/main/java/datadog/trace/instrumentation/kafka_clients/KafkaDecorator.java

ericfirth mentioned this pull request Sep 4, 2026

Eric.firth/backport pr 12350 1.65.x #12415

Closed

ericfirth and others added 12 commits September 9, 2026 09:49

Add TargetSystem.DATA_STREAMS and InstrumenterModule.DataStreams base…

601eec2

… class

Add isEnabled() test coverage for DataStreams

f3ace91

Fix test isolation bug in testDataStreamsIsEnabledWhenDataStreamsEnab…

26eb4ef

…ledOverridesFalse Save and restore the original dataStreamsEnabled value instead of hardcoding false in the finally block. This prevents test pollution where an incorrect state could leak to the next test.

Remove unused AfterEach import from InstrumenterModuleTest

0dc6b84

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

Wire AgentInstaller.getEnabledSystems() to DATA_STREAMS target system

b36a125

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

Fix DATA_STREAMS case formatting to match DATA_JOBS style

34b0202

Use qualified PrioritySampling.USER_DROP form in validateWithSamplingPriority to match the DATA_JOBS case style (line 70) which uses qualified form. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

Swap kafka-clients-0.11 instrumentations to InstrumenterModule.DataSt…

f216a2d

…reams base class Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

ericfirth and others added 2 commits September 9, 2026 09:50

ericfirth force-pushed the worktree-dsm-kafka-decoupling branch from bb5d6cf to c498c71 Compare September 9, 2026 14:37

ericfirth and others added 2 commits September 9, 2026 11:20

Merge branch 'master' into worktree-dsm-kafka-decoupling

bde0039

ericfirth and others added 2 commits September 16, 2026 11:39

Merge branch 'worktree-dsm-kafka-decoupling' of github.com:DataDog/dd…

820ab35

…-trace-java into worktree-dsm-kafka-decoupling

ericfirth added the tag: override groovy enforcement Override the "Enforce Groovy Migration" check label Sep 16, 2026

ericfirth marked this pull request as ready for review September 17, 2026 17:00

ericfirth requested review from a team as code owners September 17, 2026 17:00

ericfirth requested review from mcculls, sarahchen6 and vandonr and removed request for a team September 17, 2026 17:00

datadog-datadog-prod-us1 Bot reviewed Sep 17, 2026

View reviewed changes

chatgpt-codex-connector Bot reviewed Sep 17, 2026

View reviewed changes

dougqh reviewed Sep 17, 2026

View reviewed changes

Comment thread ...-clients-0.11/src/main/java/datadog/trace/instrumentation/kafka_clients/TracingIterator.java

dougqh reviewed Sep 17, 2026

View reviewed changes

Conversation

ericfirth commented Aug 31, 2026 • edited by mcculls Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Motivation

Test plan

Uh oh!

This comment has been minimized.

dd-octo-sts Bot commented Aug 31, 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!

pr-commenter Bot commented Aug 31, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Kafka / consumer-benchmark

Parameters

Summary

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

dougqh commented Aug 31, 2026

Uh oh!

pr-commenter Bot commented Sep 4, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Kafka / producer-benchmark

Parameters

Summary

Uh oh!

dougqh commented Sep 16, 2026

Uh oh!

chatgpt-codex-connector Bot commented Sep 17, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codex Review Summary

Uh oh!

datadog-datadog-prod-us1 Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

chatgpt-codex-connector Bot left a comment

Choose a reason for hiding this comment

💡 Codex Review

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

dougqh left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

mcculls commented Sep 18, 2026

Uh oh!

ericfirth commented Sep 18, 2026 • edited by mcculls Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

ericfirth commented Aug 31, 2026 •

edited by mcculls

Loading

dd-octo-sts Bot commented Aug 31, 2026 •

edited

Loading

pr-commenter Bot commented Aug 31, 2026 •

edited

Loading

pr-commenter Bot commented Sep 4, 2026 •

edited

Loading

chatgpt-codex-connector Bot commented Sep 17, 2026 •

edited

Loading

dougqh left a comment •

edited

Loading

ericfirth commented Sep 18, 2026 •

edited by mcculls

Loading