experiment(context): amortized/deferred mutable view over tag set filtering (#1293 cherry pick)

[tobz]

Summary

Cherry-pick of #1292 on top of #1296 to test the performance improvements (or, hopefully not: regressions) of the mutable view stuff on top of the tag_filterlist transform.

Change Type

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

How did you test this PR?

References

[Copilot]

Pull request overview

This is a cherry-pick of PR #1292 that implements a lazy copy-on-write mutable view pattern for tag set filtering operations. The feature defers expensive operations like Arc clones and context key recomputation until mutations are actually recorded, potentially improving performance when filtering metrics.

Changes:

• Add collect_removals and apply_removals methods to TagSet to support deferred mutation
• Create TagSetMutView and TagSetMutViewState to implement lazy copy-on-write semantics
• Refactor tag_filterlist component to use the new mutable view API and store reusable state
• Update all tests to pass the new required parameter

Reviewed changes

Copilot reviewed 3 out of 4 changed files in this pull request and generated no comments.

File	Description
lib/saluki-context/src/tags/tagset/owned.rs	Adds two new methods for collecting and applying tag removal indices in a deferred manner
lib/saluki-context/src/lib.rs	Exports new public types TagSetMutView and TagSetMutViewState
lib/saluki-context/src/context.rs	Implements TagSetMutViewState for scratch space and TagSetMutView for lazy copy-on-write tag mutations
bin/agent-data-plane/src/components/tag_filterlist/mod.rs	Refactors filtering logic to use the new mutable view pattern, removing the has_removable_tags function and updating all test calls

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[tobz]

Warning

This pull request is not mergeable via GitHub because a downstack PR is open. Once all requirements are satisfied, merge this PR as a stack on Graphite.
Learn more

• experiment(tag_filterlist): try trie-based tag membership set approach #1304
• experiment(context): amortized/deferred mutable view over tag set filtering (#1293 cherry pick) #1295 👈 (View in Graphite)
• chore(ci): add smp experiments for tag filtering #1296
• enhancement(agent-data-plane): add metric tag filterlist #1293
• main

This stack of pull requests is managed by Graphite. Learn more about stacking.

[Copilot]

Pull request overview

Copilot reviewed 3 out of 4 changed files in this pull request and generated no new comments.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[pr-commenter]

Binary Size Analysis (Agent Data Plane)

Target: 3e775e7 (baseline) vs 6866597 (comparison) diff
Analysis Type: Stripped binaries (debug symbols excluded)
Baseline Size: 26.47 MiB
Comparison Size: 26.46 MiB
Size Change: -3.74 KiB (-0.01%)
Pass/Fail Threshold: +5%
Result: PASSED ✅

Changes by Module

Module	File Size	Symbols
agent_data_plane::components::tag_filterlist	-8.41 KiB	5
core	+3.78 KiB	68
saluki_context::tags::tagset	+1.36 KiB	5
[Unmapped]	+1.32 KiB	1
smallvec	-960 B	1
hashbrown	-650 B	7
saluki_context::tags::Tag	-420 B	1
saluki_context::context::TagSetMutView	+311 B	1
alloc	+207 B	107
[sections]	-189 B	6
idna	-77 B	1
axum	-66 B	1
unicode_segmentation	+8 B	1
ddsketch::agent::sketch	+5 B	1
std	-4 B	25
serde_yaml	+2 B	5
protobuf	-1 B	2
agent_data_plane::cli::run	+1 B	1
saluki_core::data_model::event	+0 B	2
hickory_proto	+0 B	18

Detailed Symbol Changes

    FILE SIZE        VM SIZE    
 --------------  -------------- 
  [NEW] +27.2Ki  [NEW] +27.0Ki    _<agent_data_plane::components::tag_filterlist::TagFilterlist as saluki_core::components::transforms::Transform>::run::_{{closure}}::h743d6c09bd780764
  [NEW] +10.6Ki  [NEW] +10.5Ki    <saluki_core::data_model::event::Event as core::clone::Clone>::clone.9926
  [NEW] +5.52Ki  [NEW] +5.38Ki    <hickory_proto::rr::record_data::RData as core::clone::Clone>::clone.8377
  [NEW] +5.13Ki  [NEW] +5.00Ki    <hickory_proto::rr::record_data::RData as core::clone::Clone>::clone.11165
  +0.3% +4.71Ki  +0.3% +3.14Ki    [469 Others]
  [NEW] +4.71Ki  [NEW] +4.59Ki    <rustls::error::Error as core::clone::Clone>::clone.9904
  [NEW] +4.71Ki  [NEW] +4.59Ki    <rustls::error::Error as core::clone::Clone>::clone.11180
  [NEW] +3.83Ki  [NEW] +3.71Ki    <webpki::error::Error as core::fmt::Debug>::fmt.11358
  [NEW] +3.66Ki  [NEW] +3.54Ki    <rustls::error::Error as core::fmt::Debug>::fmt.9533
  [NEW] +3.57Ki  [NEW] +3.45Ki    <rustls::error::Error as core::fmt::Debug>::fmt.11181
  [NEW] +2.81Ki  [NEW] +2.69Ki    <figment::error::Error as core::clone::Clone>::clone.7763
  [DEL] -2.81Ki  [DEL] -2.69Ki    <figment::error::Error as core::clone::Clone>::clone.7801
  [DEL] -3.57Ki  [DEL] -3.45Ki    <rustls::error::Error as core::fmt::Debug>::fmt.11217
  [DEL] -3.66Ki  [DEL] -3.54Ki    <rustls::error::Error as core::fmt::Debug>::fmt.9571
  [DEL] -3.83Ki  [DEL] -3.71Ki    <webpki::error::Error as core::fmt::Debug>::fmt.11394
  [DEL] -4.71Ki  [DEL] -4.59Ki    <rustls::error::Error as core::clone::Clone>::clone.11216
  [DEL] -4.71Ki  [DEL] -4.59Ki    <rustls::error::Error as core::clone::Clone>::clone.9946
  [DEL] -5.13Ki  [DEL] -5.00Ki    <hickory_proto::rr::record_data::RData as core::clone::Clone>::clone.11201
  [DEL] -5.52Ki  [DEL] -5.38Ki    <hickory_proto::rr::record_data::RData as core::clone::Clone>::clone.8415
  [DEL] -10.6Ki  [DEL] -10.5Ki    <saluki_core::data_model::event::Event as core::clone::Clone>::clone.9968
  [DEL] -35.7Ki  [DEL] -35.5Ki    _<agent_data_plane::components::tag_filterlist::TagFilterlist as saluki_core::components::transforms::Transform>::run::_{{closure}}::h468c85120b3a918a
  -0.0% -3.74Ki  -0.0% -5.32Ki    TOTAL

[pr-commenter]

Regression Detector (Agent Data Plane)

Regression Detector Results

Run ID: e251b62d-4bd3-4124-843f-fd8a0d3f7fb1

Baseline: 3e775e7
Comparison: 6866597
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	+4.03	[+3.62, +4.44]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_logs_5mb_cpu	% cpu utilization	+1.56	[-3.07, +6.19]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_logs_5mb_throughput	ingress throughput	-0.01	[-0.13, +0.11]	1	(metrics) (profiles) (logs)

Fine details of change detection per experiment

perf	experiment	goal	Δ mean %	Δ mean % CI	trials	links
➖	dsd_uds_512kb_3k_contexts_cpu	% cpu utilization	+7.40	[-52.51, +67.31]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_logs_5mb_memory	memory utilization	+4.03	[+3.62, +4.44]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_traces_ottl_filtering_5mb_cpu	% cpu utilization	+2.07	[-0.55, +4.69]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_logs_5mb_cpu	% cpu utilization	+1.56	[-3.07, +6.19]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_traces_ottl_transform_5mb_cpu	% cpu utilization	+0.48	[-1.66, +2.62]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_metrics_5mb_memory	memory utilization	+0.43	[+0.23, +0.64]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_traces_5mb_memory	memory utilization	+0.29	[+0.05, +0.54]	1	(metrics) (profiles) (logs)
➖	dsd_uds_500mb_3k_contexts_cpu	% cpu utilization	+0.29	[-1.06, +1.65]	1	(metrics) (profiles) (logs)
➖	dsd_uds_512kb_3k_contexts_memory	memory utilization	+0.24	[+0.07, +0.41]	1	(metrics) (profiles) (logs)
➖	quality_gates_rss_dsd_medium	memory utilization	+0.23	[+0.04, +0.42]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_traces_ottl_filtering_5mb_memory	memory utilization	+0.19	[-0.14, +0.51]	1	(metrics) (profiles) (logs)
➖	dsd_uds_500mb_3k_contexts_memory	memory utilization	+0.15	[-0.02, +0.32]	1	(metrics) (profiles) (logs)
➖	quality_gates_rss_dsd_low	memory utilization	+0.11	[-0.08, +0.30]	1	(metrics) (profiles) (logs)
➖	quality_gates_rss_dsd_ultraheavy	memory utilization	+0.08	[-0.05, +0.21]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_metrics_5mb_throughput	ingress throughput	+0.05	[-0.07, +0.18]	1	(metrics) (profiles) (logs)
➖	dsd_uds_10mb_3k_contexts_memory	memory utilization	+0.03	[-0.16, +0.21]	1	(metrics) (profiles) (logs)
➖	dsd_uds_100mb_3k_contexts_throughput	ingress throughput	+0.01	[-0.02, +0.04]	1	(metrics) (profiles) (logs)
➖	dsd_uds_10mb_3k_contexts_throughput	ingress throughput	+0.01	[-0.13, +0.15]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_traces_ottl_transform_5mb_throughput	ingress throughput	+0.00	[-0.02, +0.02]	1	(metrics) (profiles) (logs)
➖	dsd_uds_1mb_3k_contexts_throughput	ingress throughput	+0.00	[-0.05, +0.06]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_traces_ottl_filtering_5mb_throughput	ingress throughput	-0.00	[-0.02, +0.02]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_traces_5mb_throughput	ingress throughput	-0.00	[-0.02, +0.02]	1	(metrics) (profiles) (logs)
➖	dsd_uds_512kb_3k_contexts_throughput	ingress throughput	-0.00	[-0.05, +0.05]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_logs_5mb_throughput	ingress throughput	-0.01	[-0.13, +0.11]	1	(metrics) (profiles) (logs)
➖	quality_gates_rss_dsd_heavy	memory utilization	-0.05	[-0.19, +0.10]	1	(metrics) (profiles) (logs)
➖	quality_gates_rss_idle	memory utilization	-0.10	[-0.13, -0.07]	1	(metrics) (profiles) (logs)
➖	dsd_uds_1mb_3k_contexts_memory	memory utilization	-0.11	[-0.27, +0.06]	1	(metrics) (profiles) (logs)
➖	dsd_uds_500mb_3k_contexts_throughput	ingress throughput	-0.50	[-0.63, -0.36]	1	(metrics) (profiles) (logs)
➖	dsd_uds_100mb_3k_contexts_memory	memory utilization	-0.52	[-0.70, -0.35]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_metrics_5mb_cpu	% cpu utilization	-0.58	[-8.39, +7.23]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_traces_ottl_transform_5mb_memory	memory utilization	-0.68	[-0.93, -0.44]	1	(metrics) (profiles) (logs)
➖	otlp_ingest_traces_5mb_cpu	% cpu utilization	-0.84	[-3.07, +1.39]	1	(metrics) (profiles) (logs)
➖	dsd_uds_10mb_3k_contexts_cpu	% cpu utilization	-1.63	[-32.73, +29.47]	1	(metrics) (profiles) (logs)
➖	dsd_uds_100mb_3k_contexts_cpu	% cpu utilization	-3.70	[-10.01, +2.61]	1	(metrics) (profiles) (logs)
➖	dsd_uds_1mb_3k_contexts_cpu	% cpu utilization	-5.80	[-57.79, +46.19]	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	115.57MiB ≤ 140MiB	(metrics) (profiles) (logs)
✅	quality_gates_rss_dsd_low	memory_usage	10/10	34.20MiB ≤ 50MiB	(metrics) (profiles) (logs)
✅	quality_gates_rss_dsd_medium	memory_usage	10/10	54.15MiB ≤ 75MiB	(metrics) (profiles) (logs)
✅	quality_gates_rss_dsd_ultraheavy	memory_usage	10/10	170.70MiB ≤ 200MiB	(metrics) (profiles) (logs)
✅	quality_gates_rss_idle	memory_usage	10/10	21.45MiB ≤ 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:

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".