feat(dogstatsd): Add post aggregation histogram filter component by aqian01 · Pull Request #1558 · DataDog/saluki

aqian01 · 2026-05-01T19:26:03Z

Summary

DogStatsD histograms are expanded after aggregation into metrics like .max, .avg, .count, and .95percentile.

This filter runs after aggregation so we can drop those generated histogram metrics, since they do not exist yet when the original DogStatsD sample is received.

This PR adds an ADP-only DogStatsD post-aggregate filter transform that runs immediately after dsd_agg.

Change Type

Bug fix
New feature
Non-functional (chore, refactoring, docs)
Performance

How did you test this PR?

Unit tests

References

pr-commenter · 2026-05-01T19:33:16Z

Binary Size Analysis (Agent Data Plane)

Target: 507d913 (baseline) vs fae4876 (comparison) diff
Analysis Type: Stripped binaries (debug symbols excluded)
Baseline Size: 36.78 MiB
Comparison Size: 36.95 MiB
Size Change: +179.28 KiB (+0.48%)
Pass/Fail Threshold: +5%
Result: PASSED ✅

Changes by Module

Module	File Size	Symbols
`figment`	+73.50 KiB	33
`core`	+30.34 KiB	1636
`agent_data_plane::components::dogstatsd_post_aggregate_filter`	+26.61 KiB	23
`axum`	+21.98 KiB	31
`saluki_app::api::APIBuilder`	-21.39 KiB	8
`alloc`	+15.10 KiB	116
`serde_core`	+13.50 KiB	68
`[sections]`	+12.65 KiB	8
`agent_data_plane::cli::run`	+12.48 KiB	5
`prost`	+11.12 KiB	33
`saluki_env::workload::providers`	-10.10 KiB	2
`bytes`	-8.18 KiB	19
`tower`	+7.68 KiB	26
`saluki_core::state::reflector`	+6.54 KiB	3
`saluki_config::dynamic::watcher`	+6.14 KiB	3
`agent_data_plane::components::tag_filterlist`	-5.88 KiB	7
`saluki_config::GenericConfiguration::try_get_typed`	+5.14 KiB	4
`ottl::parser::arena`	-5.12 KiB	4
`tonic_prost`	-5.07 KiB	7
`quick_cache`	+5.04 KiB	8

Detailed Symbol Changes

    FILE SIZE        VM SIZE    
 --------------  -------------- 
  +1.9%  +111Ki  +1.5% +65.9Ki    [6719 Others]
  [NEW] +34.7Ki  [NEW] +34.6Ki    _<figment::value::de::ConfiguredValueDe<I> as serde_core::de::Deserializer>::deserialize_struct::hd550f8201c462891
  [NEW] +21.2Ki  [NEW] +21.0Ki    _<figment::value::de::ConfiguredValueDe<I> as serde_core::de::Deserializer>::deserialize_struct::h722cdd348b224acf
  +100% +18.9Ki  +100% +18.8Ki    saluki_env::workload::providers::remote_agent::build_collector::_{{closure}}::hc014c2e4b066be19
  [NEW] +14.2Ki  [NEW] +14.1Ki    _<figment::value::de::ConfiguredValueDe<I> as serde_core::de::Deserializer>::deserialize_struct::h9d779790736af4a3
  [NEW] +12.6Ki  [NEW] +12.5Ki    h2::server::Connection<T,B>::poll_closed::h6ef09ff3c0430f9e
  [NEW] +12.5Ki  [NEW] +12.4Ki    h2::server::Connection<T,B>::poll_closed::hd75f66186bf27b9e
  [NEW] +11.5Ki  [NEW] +11.2Ki    _<agent_data_plane::components::dogstatsd_post_aggregate_filter::DogStatsDPostAggregateFilter as saluki_core::components::transforms::Transform>::run::_{{closure}}::ha89a4503494ea75e
  +6.5% +9.66Ki  +6.5% +9.66Ki    agent_data_plane::cli::run::handle_run_command::_{{closure}}::hcb488865a6b60fa9
 +17e2% +7.04Ki +28e2% +7.04Ki    _<hyper::server::conn::http1::Connection<I,S> as core::future::future::Future>::poll::h536fc097bc7c8eea
 +17e2% +6.97Ki +28e2% +6.97Ki    _<hyper::server::conn::http1::Connection<I,S> as core::future::future::Future>::poll::h586c4cdd2c008b4d
  +0.6% +6.30Ki  +0.6% +6.30Ki    [section .gcc_except_table]
 +11e2% +5.92Ki +13e2% +5.92Ki    saluki_core::state::reflector::Store<P>::process::h46df878273936e68
  [DEL] -5.70Ki  [DEL] -5.44Ki    agent_data_plane::components::ottl_filter_processor::config::_::_<impl serde_core::de::Deserialize for agent_data_plane::components::ottl_filter_processor::config::OttlFilterConfig>::deserialize::hc32e333e236cfcf0
  [DEL] -5.93Ki  [DEL] -5.75Ki    _<agent_data_plane::state::metrics::AggregatedMetricsProcessor as saluki_core::state::reflector::Processor>::process::hbdc920fddb89b26a
  [DEL] -7.29Ki  [DEL] -7.19Ki    h2::proto::connection::Connection<T,P,B>::poll::h8850b0c08737abb9
  [DEL] -7.45Ki  [DEL] -7.34Ki    h2::proto::connection::Connection<T,P,B>::poll::h2e92fe94476bb271
 -63.9% -8.29Ki -64.7% -8.29Ki    _<alloc::collections::btree::map::BTreeMap<K,V> as figment::coalesce::Coalescible>::coalesce::hf23dc7a18c0e6819
  [DEL] -14.9Ki  [DEL] -14.8Ki    hyper::proto::h1::dispatch::Dispatcher<D,Bs,I,T>::poll_catch::h86548adaece2a975
  [DEL] -15.1Ki  [DEL] -15.0Ki    hyper::proto::h1::dispatch::Dispatcher<D,Bs,I,T>::poll_catch::hf80a26e82cdb47d5
 -50.7% -29.0Ki -50.8% -29.0Ki    saluki_env::workload::providers::remote_agent::RemoteAgentWorkloadProvider::from_configuration::_{{closure}}::h0143b1bd6ee51cad
  +0.5%  +179Ki  +0.4%  +133Ki    TOTAL

…-aggregation histogram scalar series matching the metric filterlist

pr-commenter · 2026-05-01T19:50:05Z

Regression Detector (Agent Data Plane)

Regression Detector Results

Run ID: e2dde32e-85f2-470a-88b9-1b71ce2d75f4

Baseline: 507d913
Comparison: ac12914
Diff

Optimization Goals: ✅ No significant changes detected

Experiments ignored for regressions

Regressions in experiments with settings containing erratic: true are ignored.

perf	experiment	goal	Δ mean %	Δ mean % CI	trials	links
➖	otlp_ingest_logs_5mb_memory	memory utilization	+2.86	[+2.43, +3.30]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_logs_5mb_throughput	ingress throughput	-0.00	[-0.12, +0.12]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_logs_5mb_cpu	% cpu utilization	-0.68	[-5.53, +4.18]	1	(metrics) (profiles) (logs)

Fine details of change detection per experiment

perf	experiment	goal	Δ mean %	Δ mean % CI	trials	links
➖	otlp_ingest_logs_5mb_memory	memory utilization	+2.86	[+2.43, +3.30]	1	(metrics) (profiles) (logs)
➖	dsd_uds_500mb_3k_contexts_throughput	ingress throughput	+2.85	[+2.74, +2.97]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_traces_ottl_transform_5mb_cpu	% cpu utilization	+1.96	[-0.20, +4.13]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_metrics_5mb_cpu	% cpu utilization	+0.88	[-5.18, +6.94]	1	(metrics) (profiles) (logs)
➖	dsd_uds_500mb_3k_contexts_memory	memory utilization	+0.78	[+0.63, +0.93]	1	(metrics) (profiles) (logs)
➖	quality_gates_rss_dsd_medium	memory utilization	+0.73	[+0.56, +0.90]	1	(metrics) (profiles) (logs)
➖	dsd_uds_10mb_3k_contexts_memory	memory utilization	+0.66	[+0.51, +0.82]	1	(metrics) (profiles) (logs)
➖	dsd_uds_512kb_3k_contexts_memory	memory utilization	+0.47	[+0.32, +0.61]	1	(metrics) (profiles) (logs)
➖	quality_gates_rss_dsd_ultraheavy	memory utilization	+0.45	[+0.31, +0.58]	1	(metrics) (profiles) (logs)
➖	quality_gates_rss_idle	memory utilization	+0.35	[+0.31, +0.40]	1	(metrics) (profiles) (logs)
➖	quality_gates_rss_dsd_heavy	memory utilization	+0.31	[+0.18, +0.44]	1	(metrics) (profiles) (logs)
➖	dsd_uds_1mb_3k_contexts_memory	memory utilization	+0.30	[+0.15, +0.45]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_traces_ottl_transform_5mb_memory	memory utilization	+0.21	[+0.05, +0.38]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_traces_ottl_filtering_5mb_memory	memory utilization	+0.19	[-0.05, +0.43]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_traces_5mb_memory	memory utilization	+0.09	[-0.07, +0.25]	1	(metrics) (profiles) (logs)
➖	quality_gates_rss_dsd_low	memory utilization	+0.06	[-0.10, +0.22]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_traces_5mb_throughput	ingress throughput	+0.02	[-0.05, +0.10]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_metrics_5mb_throughput	ingress throughput	+0.01	[-0.16, +0.17]	1	(metrics) (profiles) (logs)
➖	dsd_uds_100mb_3k_contexts_throughput	ingress throughput	+0.00	[-0.04, +0.04]	1	(metrics) (profiles) (logs)
➖	dsd_uds_1mb_3k_contexts_throughput	ingress throughput	+0.00	[-0.06, +0.06]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_logs_5mb_throughput	ingress throughput	-0.00	[-0.12, +0.12]	1	(metrics) (profiles) (logs)
➖	dsd_uds_512kb_3k_contexts_throughput	ingress throughput	-0.00	[-0.05, +0.05]	1	(metrics) (profiles) (logs)
➖	dsd_uds_10mb_3k_contexts_throughput	ingress throughput	-0.00	[-0.19, +0.19]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_traces_ottl_transform_5mb_throughput	ingress throughput	-0.02	[-0.09, +0.05]	1	(metrics) (profiles) (logs)
➖	dsd_uds_100mb_3k_contexts_memory	memory utilization	-0.06	[-0.22, +0.09]	1	(metrics) (profiles) (logs)
➖	dsd_uds_500mb_3k_contexts_cpu	% cpu utilization	-0.16	[-1.50, +1.19]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_traces_ottl_filtering_5mb_throughput	ingress throughput	-0.23	[-0.31, -0.15]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_traces_5mb_cpu	% cpu utilization	-0.24	[-2.19, +1.71]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_logs_5mb_cpu	% cpu utilization	-0.68	[-5.53, +4.18]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_metrics_5mb_memory	memory utilization	-0.79	[-0.95, -0.64]	1	(metrics) (profiles) (logs)
➖	dsd_uds_100mb_3k_contexts_cpu	% cpu utilization	-1.69	[-7.26, +3.87]	1	(metrics) (profiles) (logs)
➖	dsd_uds_512kb_3k_contexts_cpu	% cpu utilization	-1.71	[-57.33, +53.92]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_traces_ottl_filtering_5mb_cpu	% cpu utilization	-2.18	[-4.44, +0.08]	1	(metrics) (profiles) (logs)
➖	dsd_uds_10mb_3k_contexts_cpu	% cpu utilization	-4.48	[-32.46, +23.49]	1	(metrics) (profiles) (logs)
➖	dsd_uds_1mb_3k_contexts_cpu	% cpu utilization	-9.24	[-59.01, +40.53]	1	(metrics) (profiles) (logs)

Bounds Checks: ✅ Passed

perf	experiment	bounds_check_name	replicates_passed	observed_value	links
✅	quality_gates_rss_dsd_heavy	memory_usage	10/10	122.12MiB ≤ 140MiB	(metrics) (profiles) (logs)
✅	quality_gates_rss_dsd_low	memory_usage	10/10	40.33MiB ≤ 50MiB	(metrics) (profiles) (logs)
✅	quality_gates_rss_dsd_medium	memory_usage	10/10	60.45MiB ≤ 75MiB	(metrics) (profiles) (logs)
✅	quality_gates_rss_dsd_ultraheavy	memory_usage	10/10	177.57MiB ≤ 200MiB	(metrics) (profiles) (logs)
✅	quality_gates_rss_idle	memory_usage	10/10	27.50MiB ≤ 40MiB	(metrics) (profiles) (logs)

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:

Its estimated |Δ mean %| ≥ 5.00%, indicating the change is big enough to merit a closer look.
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.
Its configuration does not mark it "erratic".

aqian01 · 2026-05-04T15:27:04Z

There is also a rounding change to match Core Agent behavior. The Agent parses histogram percentiles with int(i*100 + 0.5), so values like 0.299 map to 30percentile, not 29percentile.

Agent reference: https://github.com/DataDog/datadog-agent/blob/12213fe95538f47d98d73bd945a87b3e24189285/pkg/metrics/histogram.go#L53-L70

tobz

Mostly nits, but one change related to one of the unit tests.

tobz · 2026-05-04T15:11:05Z

+struct EffectiveFilterlist {
+    metric_filterlist: Vec<String>,
+    metric_filterlist_match_prefix: bool,
+    metric_blocklist: Vec<String>,
+    metric_blocklist_match_prefix: bool,
+}


Just saying this out loud so I don't forget it, but it's not a change we need to/should make here: since this code is now duplicated between the DSD Prefix Filter transform and the Tag Filterlist transform and this transform... it could potentially make sense for us to pull the DSD Prefix Filter transform into the agent-data-plane crate and stick these helper types in a common module where they can all be depended on.

## Summary  DogStatsD histograms are expanded after aggregation into metrics like <name>.max, <name>.avg, <name>.count, and <name>.95percentile. This filter runs after aggregation so we can drop those generated histogram metrics, since they do not exist yet when the original DogStatsD sample is received. This PR adds an ADP-only DogStatsD post-aggregate filter transform that runs immediately after `dsd_agg`. ## Change Type - [ ] Bug fix - [x] New feature - [ ] Non-functional (chore, refactoring, docs) - [ ] Performance ## How did you test this PR?  Unit tests ## References    cf28497

…and share filterlist helpers (#1598) ## Summary  Follow up PR for this [comment](#1558) Moves the DogStatsD prefix filter into ADP and adds shared filterlist helpers. The prefix and post-aggregate filters now use the same blocklist matching and filterlist fallback logic instead of duplicating it. ## Change Type - [ ] Bug fix - [ ] New feature - [x] Non-functional (chore, refactoring, docs) - [ ] Performance ## How did you test this PR?  Existing tests still verify behaviour. ## References    Co-authored-by: andrew.qian <andrew.qian@datadoghq.com>

…and share filterlist helpers (#1598) ## Summary  Follow up PR for this [comment](#1558) Moves the DogStatsD prefix filter into ADP and adds shared filterlist helpers. The prefix and post-aggregate filters now use the same blocklist matching and filterlist fallback logic instead of duplicating it. ## Change Type - [ ] Bug fix - [ ] New feature - [x] Non-functional (chore, refactoring, docs) - [ ] Performance ## How did you test this PR?  Existing tests still verify behaviour. ## References    Co-authored-by: andrew.qian <andrew.qian@datadoghq.com> 6dbc4cf

aqian01 added 3 commits May 1, 2026 15:47

feat(adp): add dogstatsd_post_aggregate_filter transform to drop post…

b69c3b3

…-aggregation histogram scalar series matching the metric filterlist

update run.rs with new component

f1c5b79

add telemetry and remapper

79e72a5

aqian01 force-pushed the andrewq/add_missing_telemetry_amcc branch from b677a0d to 79e72a5 Compare May 1, 2026 19:54

make test logic match the agent

fae4876

dd-octo-sts Bot added area/components Sources, transforms, and destinations. transform/aggregate Aggregate transform. labels May 4, 2026

aqian01 marked this pull request as ready for review May 4, 2026 14:12

aqian01 requested a review from a team as a code owner May 4, 2026 14:12

tobz requested changes May 4, 2026

View reviewed changes

address comments

ac12914

tobz approved these changes May 4, 2026

View reviewed changes

gh-worker-dd-devflow-36fce6 Bot added mergequeue-status: queued mergequeue-status: in_progress and removed mergequeue-status: queued labels May 4, 2026

aqian01 merged commit cf28497 into main May 4, 2026
72 of 73 checks passed

gh-worker-dd-devflow-36fce6 Bot added mergequeue-status: done and removed mergequeue-status: in_progress labels May 4, 2026

aqian01 deleted the andrewq/add_missing_telemetry_amcc branch May 4, 2026 16:18

jszwedko mentioned this pull request May 6, 2026

Support filtering specific histogram aggregates #1546

Closed

aqian01 mentioned this pull request May 6, 2026

enhancement(agent-data-plane): Move DogStatsD prefix filter into ADP and share filterlist helpers #1598

Merged

4 tasks

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat(dogstatsd): Add post aggregation histogram filter component#1558

feat(dogstatsd): Add post aggregation histogram filter component#1558
aqian01 merged 5 commits into
mainfrom
andrewq/add_missing_telemetry_amcc

aqian01 commented May 1, 2026 •

edited

Loading

Uh oh!

pr-commenter Bot commented May 1, 2026 •

edited

Loading

Changes by Module

Detailed Symbol Changes

Uh oh!

pr-commenter Bot commented May 1, 2026 •

edited

Loading

Experiments ignored for regressions

Fine details of change detection per experiment

Bounds Checks: ✅ Passed

Explanation

Uh oh!

aqian01 commented May 4, 2026

Uh oh!

tobz left a comment

Uh oh!

Uh oh!

tobz May 4, 2026

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

aqian01 commented May 1, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Change Type

How did you test this PR?

References

Uh oh!

pr-commenter Bot commented May 1, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Binary Size Analysis (Agent Data Plane)

Changes by Module

Detailed Symbol Changes

Uh oh!

pr-commenter Bot commented May 1, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Regression Detector (Agent Data Plane)

Regression Detector Results

Optimization Goals: ✅ No significant changes detected

Experiments ignored for regressions

Fine details of change detection per experiment

Bounds Checks: ✅ Passed

Explanation

Uh oh!

aqian01 commented May 4, 2026

Uh oh!

tobz left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

tobz May 4, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

aqian01 commented May 1, 2026 •

edited

Loading

pr-commenter Bot commented May 1, 2026 •

edited

Loading

pr-commenter Bot commented May 1, 2026 •

edited

Loading