TESTING Make DogStatsD stats collection lazy

[aqian01]

Summary

Replace the always-connected DogStatsD stats destination with a shared lazy collector. The /dogstatsd/stats API remains available, but normal DogStatsD ingest no longer fans out every metric batch into a stats-only topology destination.

What changed

• Added DogStatsDStatsCollector with an atomic inactive fast path and mutex-protected active collection state.
• Kept /dogstatsd/stats response behavior, max-duration validation, 429 AlreadyRunning, and cancellation cleanup.
• Wired the collector into the DogStatsD source so decoded metric events are recorded only when a stats request is active.
• Removed the permanent dsd_stats_out destination and topology connection.
• Added collector unit tests and DogStatsD source wiring coverage.

Why

The previous topology connected dsd_stats_out as a second consumer of dsd_in.metrics, which forced dispatcher fanout cloning on normal DogStatsD traffic even when no stats request was collecting. This keeps the API available while reducing normal hot-path work to an inactive atomic check.

Validation

• cargo check --workspace && cargo check --workspace --tests
• cargo nextest run -p saluki-components (584 passed, 1 skipped)
• make fmt
• git diff --check
• make check-all reached formatting and Clippy successfully, then stopped because local vale is not installed: Please install Vale: https://vale.sh/docs/install

[datadog-datadog-prod-us1]

✨ Fix all issues with BitsAI

⚠️ Warnings

🚦 1 Pipeline job failed

Semantic PR Title Check | Check For Semantic PR Title     

🛟 This job is unlikely to succeed on retry. Please review your pipeline configuration.

No release type found in pull request title. Add a prefix to indicate the type of release.

Useful? React with 👍 / 👎

_{This comment will be updated automatically if new data arrives.
🔗 Commit SHA: 2175437 | Docs | Datadog PR Page | Give us feedback!}

[pr-commenter]

Binary Size Analysis (Agent Data Plane)

Baseline: 54dc37e · Comparison: 2175437 · diff
Analysis Configuration: stripped binaries · Pass/Fail Threshold: +5%
Sizes: 37.68 MiB (baseline) vs 37.64 MiB (comparison)
Size Change: -37.46 KiB (-0.10%)

✅ Binary size difference within threshold

Changes by Module

Module	File Size	Symbols
hyper	-62.07 KiB	218
hyper_util	+41.92 KiB	57
alloc	-31.02 KiB	445
anyhow	-26.92 KiB	445
core	+20.28 KiB	2176
prost	+19.97 KiB	236
serde_with	-17.60 KiB	20
tonic	-17.06 KiB	196
saluki_components::transforms::dogstatsd_mapper	+15.87 KiB	8
figment	-15.33 KiB	144
http_body_util	+14.79 KiB	41
[sections]	-13.52 KiB	7
rustls	+12.68 KiB	139
h2	+12.11 KiB	126
std	+12.04 KiB	130
tokio	-11.74 KiB	1137
saluki_components::forwarders::datadog	+8.86 KiB	6
saluki_components::common::datadog	-8.24 KiB	77
saluki_components::destinations::prometheus	-8.21 KiB	8
tokio_rustls	-7.93 KiB	12

Detailed Symbol Changes

    FILE SIZE        VM SIZE    
 --------------  -------------- 
  [NEW] +66.6Ki  [NEW] +66.4Ki    agent_data_plane::cli::run::create_topology::_{{closure}}::h1b011ad4ddc53a6d
  +0.2% +18.5Ki  +0.6% +56.0Ki    [11708 Others]
  [NEW] +25.1Ki  [NEW] +24.9Ki    saluki_components::sources::dogstatsd::drive_stream::_{{closure}}::h5c096335e4534134
  [NEW] +22.4Ki  [NEW] +22.2Ki    _<saluki_components::sources::dogstatsd::DogStatsDConfiguration as saluki_core::components::sources::builder::SourceBuilder>::build::_{{closure}}::hf2af4e9852048fee
  [NEW] +15.3Ki  [NEW] +15.2Ki    saluki_components::transforms::trace_sampler::TraceSampler::process_trace::h83e3ece73dd4ddf4
  [NEW] +14.3Ki  [NEW] +14.2Ki    _<tracing::instrument::Instrumented<T> as core::future::future::Future>::poll::hf0b9bc13d27eb0ad
 +28e2% +13.8Ki +33e2% +13.8Ki    prost::message::Message::encode_to_vec::h38c2d0079ee57f8f
  [NEW] +13.5Ki  [NEW] +13.4Ki    saluki_components::transforms::apm_stats::span_concentrator::SpanConcentrator::new_stat_span_from_span::h87cd34edb479fa6c
 +28e2% +13.4Ki +33e2% +13.4Ki    prost::message::Message::encode::hf34018aeb1a7a39a
 +53e2% +12.2Ki +90e2% +12.2Ki    std::sys::backtrace::__rust_begin_short_backtrace::h770c6768cbe27e37
 -31.8% -11.5Ki -32.0% -11.5Ki    _<saluki_components::transforms::apm_stats::ApmStats as saluki_core::components::transforms::Transform>::run::_{{closure}}::h84b9d860b81c1dae
 -93.5% -11.7Ki -94.3% -11.7Ki    agent_data_plane::state::metrics::rules::get_datadog_agent_remappings::h377928bc659a9572
  -1.2% -12.8Ki  -1.2% -12.8Ki    [section .gcc_except_table]
  [DEL] -13.2Ki  [DEL] -13.1Ki    saluki_components::sources::dogstatsd::replay::writer::run_capture_loop::h03554b186c4c1ae5
  [DEL] -13.8Ki  [DEL] -13.6Ki    _<tracing::instrument::Instrumented<T> as core::future::future::Future>::poll::h59e4ebbb57eb7f91
 -98.3% -14.2Ki -99.7% -14.2Ki    _<saluki_components::transforms::trace_sampler::TraceSampler as saluki_core::components::transforms::SynchronousTransform>::transform_buffer::h95266a14ee7bdae6
  [DEL] -19.5Ki  [DEL] -19.3Ki    _<saluki_components::sources::dogstatsd::DogStatsDConfiguration as saluki_core::components::sources::builder::SourceBuilder>::build::_{{closure}}::hf04f3aa1c0fae428
  [DEL] -21.2Ki  [DEL] -21.0Ki    _<figment::value::de::ConfiguredValueDe<I> as serde_core::de::Deserializer>::deserialize_struct::h1495d50bfb28ee14
  [DEL] -22.0Ki  [DEL] -21.9Ki    saluki_components::sources::dogstatsd::drive_stream::_{{closure}}::h22208f5cc9094b54
  [DEL] -44.5Ki  [DEL] -44.4Ki    saluki_components::transforms::apm_stats::ApmStats::process_trace::h319e78c858c0338a
  [DEL] -68.2Ki  [DEL] -68.1Ki    agent_data_plane::cli::run::create_topology::_{{closure}}::h41fee3c69708d7c8
  -0.1% -37.5Ki  +0.0%    +120    TOTAL

[pr-commenter]

Regression Detector (Agent Data Plane)

Run ID: e40dc9d8-a48e-4254-b856-386dbd95c4ea
Baseline: 54dc37e7 · Comparison: 21754377 · diff

Optimization Goals: ✅ No significant changes detected

Fine details of change detection per experiment (35)

Experiments configured erratic: true are tagged (ignored) and skipped when determining which experiments regressed or improved. Experiments which are detected as erratic at runtime are tagged (erratic) to flag that the run's sample dispersion was high, but their regression / improvement signal still counts.

experiment	goal	Δ mean %	links
otlp_ingest_traces_ottl_transform_5mb_cpu (erratic)	cpu	⚪ +1.36	metrics profiles logs
dsd_uds_512kb_3k_contexts_cpu (erratic)	cpu	⚪ +0.78	metrics profiles logs
otlp_ingest_traces_5mb_cpu (erratic)	cpu	⚪ +0.70	metrics profiles logs
dsd_uds_1mb_3k_contexts_memory	memory	⚪ +0.53	metrics profiles logs
otlp_ingest_traces_ottl_filtering_5mb_cpu (erratic)	cpu	⚪ +0.53	metrics profiles logs
otlp_ingest_traces_ottl_filtering_5mb_throughput	throughput	⚪ -0.45	metrics profiles logs
dsd_uds_500mb_3k_contexts_memory	memory	⚪ +0.23	metrics profiles logs
quality_gates_rss_dsd_low	memory	⚪ +0.15	metrics profiles logs
quality_gates_rss_dsd_heavy	memory	⚪ +0.15	metrics profiles logs
otlp_ingest_traces_ottl_filtering_5mb_memory	memory	⚪ +0.13	metrics profiles logs
otlp_ingest_traces_5mb_throughput	throughput	⚪ -0.10	metrics profiles logs
dsd_uds_500mb_3k_contexts_cpu (erratic)	cpu	⚪ +0.07	metrics profiles logs
otlp_ingest_metrics_5mb_throughput	throughput	⚪ -0.01	metrics profiles logs
quality_gates_rss_dsd_ultraheavy	memory	⚪ +0.01	metrics profiles logs
dsd_uds_512kb_3k_contexts_throughput	throughput	⚪ -0.00	metrics profiles logs
dsd_uds_1mb_3k_contexts_throughput	throughput	⚪ -0.00	metrics profiles logs
dsd_uds_100mb_3k_contexts_throughput	throughput	⚪ +0.00	metrics profiles logs
otlp_ingest_logs_5mb_throughput (ignored)	throughput	⚪ +0.01	metrics profiles logs
otlp_ingest_logs_5mb_memory (ignored)	memory	⚪ -0.02	metrics profiles logs
dsd_uds_10mb_3k_contexts_throughput	throughput	⚪ +0.03	metrics profiles logs
otlp_ingest_traces_ottl_transform_5mb_memory	memory	⚪ -0.13	metrics profiles logs
dsd_uds_512kb_3k_contexts_memory	memory	⚪ -0.14	metrics profiles logs
quality_gates_rss_idle	memory	⚪ -0.34	metrics profiles logs
otlp_ingest_traces_ottl_transform_5mb_throughput	throughput	⚪ +0.34	metrics profiles logs
otlp_ingest_traces_5mb_memory	memory	⚪ -0.38	metrics profiles logs
dsd_uds_10mb_3k_contexts_memory	memory	⚪ -0.42	metrics profiles logs
dsd_uds_10mb_3k_contexts_cpu (erratic)	cpu	⚪ -0.67	metrics profiles logs
dsd_uds_100mb_3k_contexts_memory	memory	⚪ -0.89	metrics profiles logs
otlp_ingest_logs_5mb_cpu (ignored)	cpu	⚪ -0.89	metrics profiles logs
quality_gates_rss_dsd_medium	memory	⚪ -1.04	metrics profiles logs
otlp_ingest_metrics_5mb_cpu (erratic)	cpu	⚪ -2.52	metrics profiles logs
dsd_uds_500mb_3k_contexts_throughput	throughput	⚪ +2.59	metrics profiles logs
otlp_ingest_metrics_5mb_memory	memory	⚪ -2.88	metrics profiles logs
dsd_uds_100mb_3k_contexts_cpu (erratic)	cpu	🟢 -5.68	metrics profiles logs
dsd_uds_1mb_3k_contexts_cpu (erratic)	cpu	🟢 -7.84	metrics profiles logs

Bounds Checks: ✅ Passed (5)

experiment	check	replicates	observed	links
quality_gates_rss_dsd_heavy	memory_usage	10/10	✅ 123 MiB ≤ 140 MiB	metrics profiles logs
quality_gates_rss_dsd_low	memory_usage	10/10	✅ 40.2 MiB ≤ 50 MiB	metrics profiles logs
quality_gates_rss_dsd_medium	memory_usage	10/10	✅ 59.9 MiB ≤ 75 MiB	metrics profiles logs
quality_gates_rss_dsd_ultraheavy	memory_usage	10/10	✅ 179 MiB ≤ 200 MiB	metrics profiles logs
quality_gates_rss_idle	memory_usage	10/10	✅ 26.9 MiB ≤ 40 MiB	metrics profiles logs

Explanation

A change is flagged as a regression when |Δ mean %| > 5.00% in the regressing direction for its optimization goal AND SMP marks the experiment as a regression (is_regression: true). Improvements use the matching criteria for the improving direction. Experiments configured erratic: true (tagged (ignored)) are skipped outright; experiments detected as erratic at runtime (tagged (erratic)) still count, since that flag describes sample dispersion rather than directional certainty. The Δ mean % cell is colored accordingly: 🟢 = improvement, 🔴 = regression, ⚪ = neutral. Reduction in CPU or memory is an improvement; reduction in ingress throughput is a regression.