test(smp): Agent vs Agent+ADP 1Hz dogstatsd benchmark

[jszwedko]

Summary

Adds an SMP regression benchmark comparing the Datadog Agent (baseline) against a converged Agent+Agent-Data-Plane image (comparison) for raw dogstatsd ingest at a 1-second aggregation bucket interval. Stacked on #1459 so the single `aggregator_bucket_size_seconds: 1` knob in `datadog.yaml` reaches both Agent core and ADP. Reuses `test/smp/regression/adp/` in place — OTLP and quality_gates cases removed; the 15 `dsd_uds_*` cases rewritten to drive the Agent entrypoint and a converged Dockerfile.

Based on PR that did comparison of tag filter experiments: #1327

Draft because the meaningful review signal is the SMP report comment, which CI posts after the benchmark runs.

Test plan

• CI green: `build-adp-baseline-image`, `build-adp-comparison-image`, `build-agent-adp-baseline-image`, `build-agent-adp-comparison-image`, `run-benchmarks-adp`, `binary-size-analysis`.
• SMP report posted as a PR comment under header "Regression Detector (Agent vs Agent+ADP, 1Hz dogstatsd)".
• Spot-check the report: comparison side shows ADP-side metrics (`sub_agent: adp` series), 1Hz bucketing visible (flush rate / output volume).

References

Stacked on #1459 (`luke/configurable-aggregation`). Modeled after #1327 but slimmer — no tag filtering, reuses existing dsd_uds cases.

🤖 Generated with Claude Code

[pr-commenter]

Binary Size Analysis (Agent Data Plane)

Target: dc8f4b9 (baseline) vs a25ec14 (comparison) diff
Analysis Type: Stripped binaries (debug symbols excluded)
Baseline Size: 37.11 MiB
Comparison Size: 37.00 MiB
Size Change: -111.49 KiB (-0.29%)
Pass/Fail Threshold: +5%
Result: PASSED ✅

Changes by Module

Module	File Size	Symbols
figment	-118.82 KiB	125
otlp_protos::otlp_include::opentelemetry	-41.79 KiB	103
hyper	+25.69 KiB	76
prost	+24.60 KiB	66
hyper_util	-14.05 KiB	13
hashbrown	+11.34 KiB	72
h2	-8.86 KiB	92
[sections]	-8.52 KiB	6
tonic	-7.37 KiB	33
core	-7.09 KiB	878
serde_core	+6.51 KiB	85
serde	+6.39 KiB	18
tower	+5.15 KiB	11
async_compression	+4.62 KiB	19
tokio_util	+3.94 KiB	14
alloc	+3.76 KiB	50
saluki_components::sources::otlp	+3.75 KiB	17
tokio	+3.67 KiB	124
saluki_core::data_model::event	-3.40 KiB	8
futures_channel	-3.35 KiB	7

Detailed Symbol Changes

    FILE SIZE        VM SIZE    
 --------------  -------------- 
  [NEW] +18.5Ki  [NEW] +18.3Ki    saluki_components::transforms::apm_stats::span_concentrator::SpanConcentrator::flush::h4cec187aab531472
  [NEW] +16.5Ki  [NEW] +16.4Ki    saluki_components::transforms::apm_stats::span_concentrator::SpanConcentrator::add_span::hf95b8e429cc5bbe0
  [NEW] +16.1Ki  [NEW] +16.0Ki    saluki_components::transforms::apm_stats::span_concentrator::SpanConcentrator::new_stat_span_from_span::h831751497d326aef
  +283% +15.9Ki  +288% +15.9Ki    h2::proto::connection::DynConnection<B>::recv_frame::h9d7adeb5727e1522
  [NEW] +12.3Ki  [NEW] +12.1Ki    _<core::marker::PhantomData<T> as serde_core::de::DeserializeSeed>::deserialize::hf9a832fc7767a946
  [NEW] +9.39Ki  [NEW] +9.23Ki    _<hyper::proto::h2::server::Server<T,S,B,E> as core::future::future::Future>::poll::h7fad89436d42473e
  +750% +6.74Ki  +828% +6.74Ki    prost::encoding::message::merge_repeated::hc52fda914c63fb75
  [NEW] +6.50Ki  [NEW] +6.28Ki    saluki_components::common::datadog::apm::_::_<impl serde_core::de::Deserialize for saluki_components::common::datadog::apm::ApmConfiguration>::deserialize::h2b55df90d15c8dc3
  +739% +6.39Ki  +819% +6.39Ki    prost::encoding::message::merge_repeated::h125609fe5afef278
  [DEL] -6.56Ki  [DEL] -6.41Ki    _<core::marker::PhantomData<T> as serde_core::de::DeserializeSeed>::deserialize::h14607bccbe25f0f5
 -24.0% -8.01Ki -24.1% -7.98Ki    _<saluki_components::transforms::apm_stats::ApmStats as saluki_core::components::transforms::Transform>::run::_{{closure}}::h63fe22416badd464
  [DEL] -8.13Ki  [DEL] -8.00Ki    figment::value::de::_<impl figment::value::value::Value>::deserialize_from::hc178e2144edf2db7
 -67.1% -9.24Ki -67.7% -9.24Ki    saluki_components::transforms::trace_obfuscation::sql::obfuscate_sql_string::hbc6c7c370aac7ff9
  [DEL] -9.49Ki  [DEL] -9.34Ki    _<figment::value::magic::Tagged<T> as figment::value::magic::Magic>::deserialize_from::h44f59c3078bb5bee
  [DEL] -9.74Ki  [DEL] -9.59Ki    _<figment::value::magic::RelativePathBuf as figment::value::magic::Magic>::deserialize_from::h795d206f112d7dfd
 -81.7% -10.1Ki -82.6% -10.1Ki    _<core::pin::Pin<P> as core::future::future::Future>::poll::h901e76ef802f2f4c
  [DEL] -11.6Ki  [DEL] -11.5Ki    _<figment::value::de::ConfiguredValueDe<I> as serde_core::de::Deserializer>::deserialize_struct::h262edabf8ad9b351
  [DEL] -15.4Ki  [DEL] -15.3Ki    _<figment::value::magic::Tagged<T> as figment::value::magic::Magic>::deserialize_from::hf4491bd4db3bec82
  [DEL] -15.9Ki  [DEL] -15.7Ki    _<figment::value::magic::RelativePathBuf as figment::value::magic::Magic>::deserialize_from::hf41c2cc956726b9d
  [DEL] -32.1Ki  [DEL] -32.0Ki    saluki_components::transforms::apm_stats::ApmStats::process_trace::h7d2f794f20a992a4
  -1.4% -83.5Ki  -1.4% -69.0Ki    [4675 Others]
  -0.3%  -111Ki  -0.3% -96.7Ki    TOTAL

[pr-commenter]

Regression Detector (Agent vs Agent+ADP, 1Hz dogstatsd)

Regression Detector Results

Run ID: 6e12d29a-e079-4e47-b394-db9d8f0fb254

Baseline: 15f1e04a
Comparison: a25ec14
Diff

Optimization Goals: ❌ Regression(s) detected

perf	experiment	goal	Δ mean %	Δ mean % CI	trials	links
❌	dsd_uds_512kb_3k_contexts_memory	memory utilization	+5.96	[+5.74, +6.18]	1	bounds checks dashboard
✅	dsd_uds_10mb_3k_contexts_memory	memory utilization	-15.26	[-15.43, -15.09]	1	bounds checks dashboard
❌	dsd_uds_500mb_3k_contexts_throughput	ingress throughput	-21.53	[-21.68, -21.39]	1	bounds checks dashboard
✅	dsd_uds_100mb_3k_contexts_cpu	% cpu utilization	-46.76	[-70.35, -23.18]	1	bounds checks dashboard
✅	dsd_uds_500mb_3k_contexts_cpu	% cpu utilization	-56.79	[-65.52, -48.07]	1	bounds checks dashboard
✅	dsd_uds_100mb_3k_contexts_memory	memory utilization	-56.99	[-57.14, -56.84]	1	bounds checks dashboard
✅	dsd_uds_500mb_3k_contexts_memory	memory utilization	-76.10	[-76.21, -75.99]	1	bounds checks dashboard

Fine details of change detection per experiment

perf	experiment	goal	Δ mean %	Δ mean % CI	trials	links
❌	dsd_uds_512kb_3k_contexts_memory	memory utilization	+5.96	[+5.74, +6.18]	1	bounds checks dashboard
➖	dsd_uds_1mb_3k_contexts_memory	memory utilization	+3.90	[+3.68, +4.12]	1	bounds checks dashboard
➖	dsd_uds_512kb_3k_contexts_throughput	ingress throughput	+0.00	[-0.06, +0.06]	1	bounds checks dashboard
➖	dsd_uds_10mb_3k_contexts_throughput	ingress throughput	+0.00	[-0.06, +0.06]	1	bounds checks dashboard
➖	dsd_uds_1mb_3k_contexts_throughput	ingress throughput	-0.00	[-0.06, +0.06]	1	bounds checks dashboard
➖	dsd_uds_100mb_3k_contexts_throughput	ingress throughput	-0.05	[-0.21, +0.10]	1	bounds checks dashboard
✅	dsd_uds_10mb_3k_contexts_memory	memory utilization	-15.26	[-15.43, -15.09]	1	bounds checks dashboard
➖	dsd_uds_10mb_3k_contexts_cpu	% cpu utilization	-18.44	[-79.85, +42.98]	1	bounds checks dashboard
❌	dsd_uds_500mb_3k_contexts_throughput	ingress throughput	-21.53	[-21.68, -21.39]	1	bounds checks dashboard
➖	dsd_uds_512kb_3k_contexts_cpu	% cpu utilization	-26.11	[-92.52, +40.30]	1	bounds checks dashboard
➖	dsd_uds_1mb_3k_contexts_cpu	% cpu utilization	-42.31	[-106.56, +21.93]	1	bounds checks dashboard
✅	dsd_uds_100mb_3k_contexts_cpu	% cpu utilization	-46.76	[-70.35, -23.18]	1	bounds checks dashboard
✅	dsd_uds_500mb_3k_contexts_cpu	% cpu utilization	-56.79	[-65.52, -48.07]	1	bounds checks dashboard
✅	dsd_uds_100mb_3k_contexts_memory	memory utilization	-56.99	[-57.14, -56.84]	1	bounds checks dashboard
✅	dsd_uds_500mb_3k_contexts_memory	memory utilization	-76.10	[-76.21, -75.99]	1	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".