Skip to content

[Feature Flags] Add agentless EVP fallback transport - #9235

Open
leoromanovsky wants to merge 1 commit into
leo.romanovsky/fflsdk-189-dotnet-exposure-lifecyclefrom
leo.romanovsky/fflsdk-189-evp-fallback-dotnet-master
Open

leoromanovsky wants to merge 1 commit into
leo.romanovsky/fflsdk-189-dotnet-exposure-lifecyclefrom
leo.romanovsky/fflsdk-189-evp-fallback-dotnet-master

Conversation

@leoromanovsky

@leoromanovsky leoromanovsky commented Sep 12, 2026 •

Copy link
Copy Markdown
Contributor

Motivation

Agentless configuration lets customers evaluate flags without running an Agent, but exposure delivery still requires its local EVP proxy. Evaluations can succeed while their experiment exposures are lost.

flowchart LR
    C[Agentless configuration] --> E[Flag evaluation]
    E --> X[Exposure batch]
    X --> A[Local Agent required]
    X -. missing route .-> D[Direct EVP intake]
Loading

Changes and Decisions

This is the single-commit final activation after endpoint-bound discovery #9352 and bounded exposure lifecycle #9353. It keeps Remote Config on its historical fixed EVP v2 behavior and does not add flag-evaluation emission.

Agentless delivery prefers local EVP v4, then v2, only when discovery confirms forwarding of both SDK identity headers. Configured Agent path prefixes are preserved. Otherwise it uses canonical HTTPS intake; API keys remain direct-only, relay headers local-only, and redirects are disabled for Agentless delivery.

Direct selection is sticky. An unavailable route has bounded recovery; each batch has bounded discovery waiting. The current batch can move to direct only after local 404/405 or a proven pre-send connection failure. Ambiguous failures never replay that batch, preventing duplicate exposures. Identity, credentials and this replay policy ship together.

flowchart LR
    E[Agentless exposure batch] --> R{Selected route}
    R -->|Identity-capable v4 or v2| A[Local relay]
    R -->|No eligible relay| D[Direct HTTPS intake]
    A -->|404 / 405 or proven pre-send failure| D
    A -->|Ambiguous failure| N[No current-batch replay]
    D --> S[Direct remains selected]
Loading

@linear-code

linear-code Bot commented Sep 12, 2026

Copy link
Copy Markdown

FFLSDK-189

@pr-commenter

pr-commenter Bot commented Sep 12, 2026 •

Copy link
Copy Markdown

Benchmarks

Benchmark execution time: 2026-10-01 23:22:37

Comparing candidate commit 8d55124 in PR branch leo.romanovsky/fflsdk-189-evp-fallback-dotnet-master with baseline commit 0610b13 in branch master.

📊 Benchmarking dashboard

Found 1 performance improvements and 11 performance regressions! Performance is the same for 60 metrics, 0 unstable metrics, 78 known flaky benchmarks, 48 flaky benchmarks without significant changes.

Explanation

This is an A/B test comparing a candidate commit's performance against that of a baseline commit. Performance changes are noted in the tables below as:

  • 🟩 = significantly better candidate vs. baseline
  • 🟥 = significantly worse candidate vs. baseline

We compute a confidence interval (CI) over the relative difference of means between metrics from the candidate and baseline commits, considering the baseline as the reference.

If the CI is entirely outside the configured SIGNIFICANT_IMPACT_THRESHOLD (or the deprecated UNCONFIDENCE_THRESHOLD), the change is considered significant.

Feel free to reach out to #apm-benchmarking-platform on Slack if you have any questions.

More details about the CI and significant changes

You can imagine this CI as a range of values that is likely to contain the true difference of means between the candidate and baseline commits.

CIs of the difference of means are often centered around 0%, because often changes are not that big:

---------------------------------(------|---^--------)-------------------------------->
                              -0.6%    0%  0.3%     +1.2%
                                 |          |        |
         lower bound of the CI --'          |        |
sample mean (center of the CI) -------------'        |
         upper bound of the CI ----------------------'

As described above, a change is considered significant if the CI is entirely outside the configured SIGNIFICANT_IMPACT_THRESHOLD (or the deprecated UNCONFIDENCE_THRESHOLD).

For instance, for an execution time metric, this confidence interval indicates a significantly worse performance:

----------------------------------------|---------|---(---------^---------)---------->
                                       0%        1%  1.3%      2.2%      3.1%
                                                  |   |         |         |
       significant impact threshold --------------'   |         |         |
                      lower bound of CI --------------'         |         |
       sample mean (center of the CI) --------------------------'         |
                      upper bound of CI ----------------------------------'

scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.TracerBenchmark.StartActiveSpan net6.0

  • 🟩 throughput [+16434.627op/s; +20784.801op/s] or [+6.364%; +8.048%]

scenario:Benchmarks.Trace.DbCommandBenchmark.ExecuteNonQuery net472

  • 🟥 allocated_mem [+95 bytes; +96 bytes] or [+10.133%; +10.144%]
  • 🟥 throughput [-41467.045op/s; -37773.608op/s] or [-11.679%; -10.639%]

scenario:Benchmarks.Trace.DbCommandBenchmark.ExecuteNonQuery net6.0

  • 🟥 allocated_mem [+95 bytes; +96 bytes] or [+9.750%; +9.761%]

scenario:Benchmarks.Trace.DbCommandBenchmark.ExecuteNonQuery netcoreapp3.1

  • 🟥 allocated_mem [+95 bytes; +96 bytes] or [+9.830%; +9.841%]
  • 🟥 throughput [-41735.097op/s; -33372.621op/s] or [-10.316%; -8.249%]

scenario:Benchmarks.Trace.HttpClientBenchmark.SendAsync net472

  • 🟥 allocated_mem [+271 bytes; +272 bytes] or [+9.338%; +9.348%]
  • 🟥 throughput [-9799.337op/s; -9536.078op/s] or [-11.187%; -10.886%]

scenario:Benchmarks.Trace.HttpClientBenchmark.SendAsync net6.0

  • 🟥 allocated_mem [+199 bytes; +200 bytes] or [+9.426%; +9.438%]
  • 🟥 throughput [-9043.909op/s; -8222.582op/s] or [-6.144%; -5.586%]

scenario:Benchmarks.Trace.HttpClientBenchmark.SendAsync netcoreapp3.1

  • 🟥 allocated_mem [+199 bytes; +200 bytes] or [+7.569%; +7.581%]
  • 🟥 throughput [-12806.313op/s; -11593.332op/s] or [-10.162%; -9.200%]

Known flaky benchmarks

These benchmarks are marked as flaky and will not trigger a failure. Modify FLAKY_BENCHMARKS_REGEX to control which benchmarks are marked as flaky.

scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.TelemetrySpanBenchmark.StartSpan_AddEvent_Sampled net472

  • 🟥 throughput [-10301.569op/s; -8939.038op/s] or [-6.245%; -5.419%]

scenario:Benchmarks.Trace.ActivityBenchmark.StartStopWithChild net472

  • 🟥 throughput [-10736.615op/s; -10111.840op/s] or [-12.730%; -11.990%]

scenario:Benchmarks.Trace.ActivityBenchmark.StartStopWithChild net6.0

  • 🟥 throughput [-10147.442op/s; -7914.734op/s] or [-8.529%; -6.653%]

scenario:Benchmarks.Trace.ActivityBenchmark.StartStopWithChild netcoreapp3.1

  • 🟥 throughput [-16078.637op/s; -14249.945op/s] or [-16.348%; -14.489%]

scenario:Benchmarks.Trace.AgentWriterBenchmark.WriteAndFlushEnrichedTraces net472

  • 🟥 allocated_mem [+1.637KB; +1.637KB] or [+49.731%; +49.747%]
  • 🟥 execution_time [+306.701ms; +309.783ms] or [+152.196%; +153.725%]
  • 🟥 throughput [-55.767op/s; -51.349op/s] or [-10.034%; -9.239%]

scenario:Benchmarks.Trace.AgentWriterBenchmark.WriteAndFlushEnrichedTraces net6.0

  • 🟥 allocated_mem [+1.012KB; +1.012KB] or [+37.524%; +37.537%]
  • 🟥 execution_time [+379.482ms; +385.288ms] or [+299.814%; +304.401%]
  • 🟩 throughput [+69.833op/s; +73.212op/s] or [+9.207%; +9.653%]

scenario:Benchmarks.Trace.AgentWriterBenchmark.WriteAndFlushEnrichedTraces netcoreapp3.1

  • 🟥 allocated_mem [+1.090KB; +1.090KB] or [+40.417%; +40.429%]
  • 🟥 execution_time [+390.272ms; +396.587ms] or [+345.376%; +350.964%]

scenario:Benchmarks.Trace.Asm.AppSecBodyBenchmark.AllCycleMoreComplexBody net472

  • 🟥 allocated_mem [+4.758KB; +4.758KB] or [+100.159%; +100.174%]
  • 🟥 throughput [-60687.353op/s; -60284.467op/s] or [-47.218%; -46.904%]

scenario:Benchmarks.Trace.Asm.AppSecBodyBenchmark.AllCycleMoreComplexBody net6.0

  • 🟥 allocated_mem [+3.880KB; +3.880KB] or [+82.058%; +82.072%]
  • 🟩 execution_time [-16.021ms; -11.845ms] or [-7.482%; -5.532%]
  • 🟥 throughput [-61366.814op/s; -58605.222op/s] or [-44.794%; -42.779%]

scenario:Benchmarks.Trace.Asm.AppSecBodyBenchmark.AllCycleMoreComplexBody netcoreapp3.1

  • 🟥 allocated_mem [+4.608KB; +4.608KB] or [+99.641%; +99.654%]
  • 🟥 throughput [-49656.303op/s; -47394.561op/s] or [-44.895%; -42.850%]

scenario:Benchmarks.Trace.Asm.AppSecBodyBenchmark.AllCycleSimpleBody net472

  • 🟥 allocated_mem [+1.380KB; +1.380KB] or [+111.648%; +111.662%]
  • 🟥 throughput [-298530.430op/s; -293568.755op/s] or [-30.481%; -29.975%]

scenario:Benchmarks.Trace.Asm.AppSecBodyBenchmark.AllCycleSimpleBody net6.0

  • 🟥 allocated_mem [+543 bytes; +544 bytes] or [+44.437%; +44.446%]
  • 🟩 execution_time [-26.765ms; -21.901ms] or [-11.936%; -9.767%]
  • 🟥 throughput [-97761.064op/s; -75309.804op/s] or [-10.444%; -8.045%]

scenario:Benchmarks.Trace.Asm.AppSecBodyBenchmark.AllCycleSimpleBody netcoreapp3.1

  • 🟥 allocated_mem [+1.344KB; +1.344KB] or [+111.253%; +111.267%]
  • 🟥 throughput [-177706.607op/s; -161508.888op/s] or [-25.533%; -23.206%]

scenario:Benchmarks.Trace.Asm.AppSecBodyBenchmark.ObjectExtractorMoreComplexBody net472

  • 🟥 allocated_mem [+3.378KB; +3.378KB] or [+89.003%; +89.017%]
  • 🟥 throughput [-73536.112op/s; -72717.099op/s] or [-49.489%; -48.938%]

scenario:Benchmarks.Trace.Asm.AppSecBodyBenchmark.ObjectExtractorMoreComplexBody net6.0

  • 🟥 allocated_mem [+3.336KB; +3.336KB] or [+88.150%; +88.161%]
  • 🟥 throughput [-74727.843op/s; -71844.212op/s] or [-47.548%; -45.713%]

scenario:Benchmarks.Trace.Asm.AppSecBodyBenchmark.ObjectExtractorMoreComplexBody netcoreapp3.1

  • 🟥 allocated_mem [+3.264KB; +3.264KB] or [+88.493%; +88.506%]
  • 🟥 throughput [-57085.415op/s; -54457.167op/s] or [-45.476%; -43.382%]

scenario:Benchmarks.Trace.Asm.AppSecBodyBenchmark.ObjectExtractorSimpleBody net6.0

  • 🟩 throughput [+220843.794op/s; +255307.941op/s] or [+7.364%; +8.513%]

scenario:Benchmarks.Trace.Asm.AppSecBodyBenchmark.ObjectExtractorSimpleBody netcoreapp3.1

  • 🟩 execution_time [-18.921ms; -14.589ms] or [-8.722%; -6.725%]

scenario:Benchmarks.Trace.Asm.AppSecEncoderBenchmark.EncodeArgs net472

  • 🟩 allocated_mem [-13.759KB; -13.757KB] or [-42.326%; -42.318%]
  • 🟥 execution_time [+298.827ms; +299.809ms] or [+149.314%; +149.804%]
  • 🟩 throughput [+1009.452op/s; +1026.413op/s] or [+11.149%; +11.337%]

scenario:Benchmarks.Trace.Asm.AppSecEncoderBenchmark.EncodeArgs net6.0

  • 🟩 allocated_mem [-13.722KB; -13.718KB] or [-42.341%; -42.329%]
  • 🟥 execution_time [+300.405ms; +303.949ms] or [+151.495%; +153.282%]
  • 🟩 throughput [+1998.466op/s; +2288.803op/s] or [+15.285%; +17.506%]

scenario:Benchmarks.Trace.Asm.AppSecEncoderBenchmark.EncodeArgs netcoreapp3.1

  • 🟩 allocated_mem [-13.722KB; -13.718KB] or [-42.341%; -42.329%]
  • 🟥 execution_time [+300.356ms; +302.980ms] or [+151.296%; +152.618%]
  • 🟩 throughput [+1861.474op/s; +1989.897op/s] or [+17.972%; +19.211%]

scenario:Benchmarks.Trace.Asm.AppSecEncoderBenchmark.EncodeLegacyArgs net472

  • 🟥 execution_time [+295.754ms; +296.579ms] or [+145.263%; +145.668%]
  • 🟩 throughput [+525.954op/s; +535.492op/s] or [+13.943%; +14.196%]

scenario:Benchmarks.Trace.Asm.AppSecEncoderBenchmark.EncodeLegacyArgs net6.0

  • 🟥 execution_time [+298.854ms; +300.997ms] or [+146.099%; +147.146%]
  • 🟩 throughput [+2752.879op/s; +2815.770op/s] or [+39.994%; +40.908%]

scenario:Benchmarks.Trace.Asm.AppSecEncoderBenchmark.EncodeLegacyArgs netcoreapp3.1

  • 🟥 execution_time [+301.830ms; +302.820ms] or [+150.854%; +151.349%]
  • 🟩 throughput [+1380.018op/s; +1397.466op/s] or [+27.392%; +27.739%]

scenario:Benchmarks.Trace.Asm.AppSecWafBenchmark.RunWafRealisticBenchmark net472

  • 🟩 execution_time [-146.840µs; -142.268µs] or [-30.148%; -29.210%]
  • 🟩 throughput [+851.505op/s; +883.395op/s] or [+41.472%; +43.025%]

scenario:Benchmarks.Trace.Asm.AppSecWafBenchmark.RunWafRealisticBenchmark net6.0

  • 🟩 execution_time [-127.730µs; -100.881µs] or [-29.295%; -23.137%]
  • 🟩 throughput [+747.097op/s; +872.645op/s] or [+32.481%; +37.939%]

scenario:Benchmarks.Trace.Asm.AppSecWafBenchmark.RunWafRealisticBenchmark netcoreapp3.1

  • 🟩 execution_time [-142.185µs; -120.179µs] or [-30.464%; -25.749%]
  • 🟩 throughput [+772.810op/s; +855.827op/s] or [+35.674%; +39.507%]

scenario:Benchmarks.Trace.Asm.AppSecWafBenchmark.RunWafRealisticBenchmarkWithAttack net472

  • 🟩 execution_time [-127.762µs; -123.458µs] or [-34.495%; -33.333%]
  • 🟩 throughput [+1361.511op/s; +1412.392op/s] or [+50.424%; +52.308%]

scenario:Benchmarks.Trace.Asm.AppSecWafBenchmark.RunWafRealisticBenchmarkWithAttack net6.0

  • 🟩 execution_time [-93.382µs; -69.425µs] or [-29.812%; -22.164%]
  • 🟩 throughput [+1003.337op/s; +1218.381op/s] or [+31.277%; +37.981%]

scenario:Benchmarks.Trace.Asm.AppSecWafBenchmark.RunWafRealisticBenchmarkWithAttack netcoreapp3.1

  • 🟩 execution_time [-133.342µs; -106.718µs] or [-36.477%; -29.194%]
  • 🟩 throughput [+1211.509op/s; +1422.544op/s] or [+43.476%; +51.049%]

scenario:Benchmarks.Trace.AspNetCoreBenchmark.SendRequest net472

  • 🟥 execution_time [+299.712ms; +300.385ms] or [+149.587%; +149.923%]

scenario:Benchmarks.Trace.AspNetCoreBenchmark.SendRequest net6.0

  • 🟥 execution_time [+416.634ms; +424.522ms] or [+452.690%; +461.261%]

scenario:Benchmarks.Trace.AspNetCoreBenchmark.SendRequest netcoreapp3.1

  • unstable execution_time [+295.587ms; +349.423ms] or [+224.437%; +265.314%]

scenario:Benchmarks.Trace.CIVisibilityProtocolWriterBenchmark.WriteAndFlushEnrichedTraces net472

  • 🟥 allocated_mem [+2.834KB; +2.839KB] or [+5.035%; +5.044%]
  • unstable execution_time [+298.805ms; +331.172ms] or [+137.388%; +152.270%]
  • 🟥 throughput [-529.479op/s; -484.668op/s] or [-47.976%; -43.916%]

scenario:Benchmarks.Trace.CIVisibilityProtocolWriterBenchmark.WriteAndFlushEnrichedTraces net6.0

  • unstable execution_time [+207.523ms; +340.784ms] or [+88.438%; +145.228%]
  • 🟥 throughput [-673.095op/s; -589.574op/s] or [-44.896%; -39.325%]

scenario:Benchmarks.Trace.CIVisibilityProtocolWriterBenchmark.WriteAndFlushEnrichedTraces netcoreapp3.1

  • 🟥 execution_time [+341.035ms; +352.019ms] or [+203.978%; +210.548%]
  • 🟥 throughput [-400.724op/s; -364.782op/s] or [-27.902%; -25.399%]

scenario:Benchmarks.Trace.CharSliceBenchmark.OptimizedCharSlice net472

  • 🟥 execution_time [+205.431µs; +216.143µs] or [+10.338%; +10.877%]
  • 🟥 throughput [-49.374op/s; -47.095op/s] or [-9.811%; -9.358%]

scenario:Benchmarks.Trace.CharSliceBenchmark.OriginalCharSlice net472

  • 🟥 execution_time [+249.988µs; +261.201µs] or [+9.765%; +10.203%]
  • 🟥 throughput [-36.157op/s; -34.725op/s] or [-9.257%; -8.890%]

scenario:Benchmarks.Trace.CharSliceBenchmark.OriginalCharSlice net6.0

  • 🟩 execution_time [-130.098µs; -101.123µs] or [-6.590%; -5.123%]
  • 🟩 throughput [+27.753op/s; +36.093op/s] or [+5.479%; +7.125%]

scenario:Benchmarks.Trace.ElasticsearchBenchmark.CallElasticsearch net472

  • 🟥 allocated_mem [+96 bytes; +97 bytes] or [+12.092%; +12.103%]
  • 🟥 execution_time [+302.700ms; +304.361ms] or [+152.434%; +153.270%]

scenario:Benchmarks.Trace.ElasticsearchBenchmark.CallElasticsearch net6.0

  • 🟥 allocated_mem [+95 bytes; +96 bytes] or [+12.114%; +12.124%]
  • 🟥 execution_time [+300.000ms; +302.961ms] or [+150.331%; +151.814%]

scenario:Benchmarks.Trace.ElasticsearchBenchmark.CallElasticsearch netcoreapp3.1

  • 🟥 allocated_mem [+95 bytes; +96 bytes] or [+12.114%; +12.124%]
  • 🟥 execution_time [+302.358ms; +306.654ms] or [+151.892%; +154.050%]
  • 🟥 throughput [-41425.758op/s; -32879.985op/s] or [-8.726%; -6.926%]

scenario:Benchmarks.Trace.ElasticsearchBenchmark.CallElasticsearchAsync net472

  • 🟥 allocated_mem [+95 bytes; +96 bytes] or [+11.165%; +11.177%]
  • 🟥 execution_time [+301.186ms; +302.471ms] or [+151.245%; +151.891%]
  • 🟥 throughput [-19759.808op/s; -17646.881op/s] or [-6.620%; -5.912%]

scenario:Benchmarks.Trace.ElasticsearchBenchmark.CallElasticsearchAsync net6.0

  • 🟥 allocated_mem [+95 bytes; +96 bytes] or [+12.493%; +12.506%]
  • 🟥 execution_time [+297.600ms; +299.364ms] or [+147.150%; +148.022%]

scenario:Benchmarks.Trace.ElasticsearchBenchmark.CallElasticsearchAsync netcoreapp3.1

  • 🟥 allocated_mem [+95 bytes; +96 bytes] or [+11.422%; +11.434%]
  • 🟥 execution_time [+302.561ms; +306.195ms] or [+153.351%; +155.193%]
  • 🟥 throughput [-42697.519op/s; -32583.769op/s] or [-9.220%; -7.036%]

scenario:Benchmarks.Trace.GraphQLBenchmark.ExecuteAsync net472

  • 🟥 allocated_mem [+96 bytes; +97 bytes] or [+11.197%; +11.208%]
  • 🟥 execution_time [+300.793ms; +303.946ms] or [+150.971%; +152.554%]
  • 🟥 throughput [-29677.148op/s; -26583.083op/s] or [-7.699%; -6.896%]

scenario:Benchmarks.Trace.GraphQLBenchmark.ExecuteAsync net6.0

  • 🟥 allocated_mem [+95 bytes; +96 bytes] or [+10.618%; +10.631%]
  • 🟥 execution_time [+298.324ms; +302.232ms] or [+148.687%; +150.635%]

scenario:Benchmarks.Trace.GraphQLBenchmark.ExecuteAsync netcoreapp3.1

  • 🟥 allocated_mem [+95 bytes; +96 bytes] or [+10.618%; +10.631%]
  • 🟥 execution_time [+299.859ms; +302.191ms] or [+149.177%; +150.338%]
  • 🟥 throughput [-34192.010op/s; -28556.162op/s] or [-8.093%; -6.759%]

scenario:Benchmarks.Trace.ILoggerBenchmark.EnrichedLog net472

  • 🟥 allocated_mem [+96 bytes; +97 bytes] or [+6.002%; +6.011%]
  • 🟥 throughput [-18882.765op/s; -16811.584op/s] or [-7.593%; -6.760%]

scenario:Benchmarks.Trace.ILoggerBenchmark.EnrichedLog net6.0

  • 🟥 allocated_mem [+95 bytes; +96 bytes] or [+5.627%; +5.636%]

scenario:Benchmarks.Trace.ILoggerBenchmark.EnrichedLog netcoreapp3.1

  • 🟥 allocated_mem [+95 bytes; +96 bytes] or [+5.605%; +5.617%]

scenario:Benchmarks.Trace.Iast.StringAspectsBenchmark.StringConcatAspectBenchmark net472

  • unstable execution_time [+15.283µs; +59.491µs] or [+3.775%; +14.695%]

scenario:Benchmarks.Trace.Iast.StringAspectsBenchmark.StringConcatAspectBenchmark net6.0

  • 🟩 allocated_mem [-18.557KB; -18.534KB] or [-6.769%; -6.761%]
  • unstable execution_time [-31.358µs; +26.621µs] or [-6.198%; +5.261%]
  • unstable throughput [-93.877op/s; +109.045op/s] or [-4.684%; +5.441%]

scenario:Benchmarks.Trace.Iast.StringAspectsBenchmark.StringConcatAspectBenchmark netcoreapp3.1

  • unstable execution_time [-80.066µs; -19.083µs] or [-13.875%; -3.307%]

scenario:Benchmarks.Trace.Iast.StringAspectsBenchmark.StringConcatBenchmark net472

  • 🟥 allocated_mem [+8.190KB; +8.196KB] or [+16.663%; +16.675%]

scenario:Benchmarks.Trace.Iast.StringAspectsBenchmark.StringConcatBenchmark net6.0

  • 🟥 execution_time [+5.490µs; +9.604µs] or [+12.978%; +22.702%]
  • 🟥 throughput [-4478.770op/s; -2704.447op/s] or [-18.854%; -11.385%]

scenario:Benchmarks.Trace.Iast.StringAspectsBenchmark.StringConcatBenchmark netcoreapp3.1

  • unstable execution_time [-14.699µs; -7.735µs] or [-22.805%; -12.001%]
  • 🟩 throughput [+1857.689op/s; +3357.566op/s] or [+11.398%; +20.600%]

scenario:Benchmarks.Trace.Log4netBenchmark.EnrichedLog net472

  • 🟥 execution_time [+302.797ms; +304.713ms] or [+153.050%; +154.019%]

scenario:Benchmarks.Trace.Log4netBenchmark.EnrichedLog net6.0

  • 🟥 execution_time [+301.681ms; +304.747ms] or [+153.555%; +155.115%]

scenario:Benchmarks.Trace.Log4netBenchmark.EnrichedLog netcoreapp3.1

  • 🟥 execution_time [+295.430ms; +298.618ms] or [+147.899%; +149.495%]

scenario:Benchmarks.Trace.RedisBenchmark.SendReceive net472

  • 🟥 allocated_mem [+95 bytes; +96 bytes] or [+9.197%; +9.208%]
  • 🟥 throughput [-38912.987op/s; -36720.311op/s] or [-10.772%; -10.165%]

scenario:Benchmarks.Trace.RedisBenchmark.SendReceive net6.0

  • 🟥 allocated_mem [+95 bytes; +96 bytes] or [+9.226%; +9.238%]

scenario:Benchmarks.Trace.RedisBenchmark.SendReceive netcoreapp3.1

  • 🟥 allocated_mem [+95 bytes; +96 bytes] or [+9.226%; +9.238%]
  • 🟥 throughput [-43019.720op/s; -34845.750op/s] or [-10.181%; -8.247%]

scenario:Benchmarks.Trace.SerilogBenchmark.EnrichedLog net472

  • 🟥 execution_time [+296.537ms; +299.040ms] or [+147.797%; +149.044%]
  • 🟥 throughput [-8802.847op/s; -7696.102op/s] or [-5.813%; -5.082%]

scenario:Benchmarks.Trace.SerilogBenchmark.EnrichedLog net6.0

  • 🟥 allocated_mem [+96 bytes; +96 bytes] or [+6.000%; +6.010%]
  • 🟥 execution_time [+301.843ms; +303.177ms] or [+151.571%; +152.241%]

scenario:Benchmarks.Trace.SerilogBenchmark.EnrichedLog netcoreapp3.1

  • 🟥 allocated_mem [+95 bytes; +96 bytes] or [+5.819%; +5.828%]
  • 🟥 execution_time [+302.967ms; +305.474ms] or [+153.645%; +154.917%]

scenario:Benchmarks.Trace.SingleSpanAspNetCoreBenchmark.SingleSpanAspNetCore net472

  • 🟥 execution_time [+300.350ms; +301.329ms] or [+149.816%; +150.305%]
  • 🟩 throughput [+65189515.934op/s; +65506079.558op/s] or [+47.475%; +47.706%]

scenario:Benchmarks.Trace.SingleSpanAspNetCoreBenchmark.SingleSpanAspNetCore net6.0

  • unstable execution_time [+325.740ms; +395.198ms] or [+405.116%; +491.498%]

scenario:Benchmarks.Trace.SingleSpanAspNetCoreBenchmark.SingleSpanAspNetCore netcoreapp3.1

  • 🟥 execution_time [+299.040ms; +300.111ms] or [+149.154%; +149.689%]
  • 🟩 throughput [+18260536.371op/s; +19238744.636op/s] or [+8.088%; +8.522%]

scenario:Benchmarks.Trace.SpanBenchmark.StartFinishScope net472

  • 🟥 allocated_mem [+31 bytes; +32 bytes] or [+5.244%; +5.256%]
  • 🟥 throughput [-78380.948op/s; -70470.384op/s] or [-8.747%; -7.865%]

scenario:Benchmarks.Trace.SpanBenchmark.StartFinishScope net6.0

  • 🟩 throughput [+64702.857op/s; +74775.210op/s] or [+6.041%; +6.982%]

scenario:Benchmarks.Trace.SpanBenchmark.StartFinishSpan net472

  • 🟥 allocated_mem [+31 bytes; +32 bytes] or [+6.037%; +6.047%]
  • 🟥 throughput [-106104.780op/s; -86500.857op/s] or [-9.713%; -7.918%]

scenario:Benchmarks.Trace.SpanBenchmark.StartFinishSpan net6.0

  • 🟥 allocated_mem [+31 bytes; +32 bytes] or [+6.057%; +6.065%]

scenario:Benchmarks.Trace.SpanBenchmark.StartFinishSpan netcoreapp3.1

  • 🟥 allocated_mem [+31 bytes; +32 bytes] or [+6.059%; +6.069%]

scenario:Benchmarks.Trace.SpanBenchmark.StartFinishTwoScopes net6.0

  • 🟩 throughput [+43844.496op/s; +49183.839op/s] or [+7.961%; +8.931%]

scenario:Benchmarks.Trace.TraceAnnotationsBenchmark.RunOnMethodBegin net472

  • 🟥 allocated_mem [+95 bytes; +96 bytes] or [+15.733%; +15.745%]
  • 🟥 throughput [-86168.430op/s; -82647.649op/s] or [-12.611%; -12.096%]

scenario:Benchmarks.Trace.TraceAnnotationsBenchmark.RunOnMethodBegin net6.0

  • 🟥 allocated_mem [+95 bytes; +96 bytes] or [+14.810%; +14.823%]

scenario:Benchmarks.Trace.TraceAnnotationsBenchmark.RunOnMethodBegin netcoreapp3.1

  • 🟥 allocated_mem [+95 bytes; +96 bytes] or [+14.810%; +14.821%]
  • 🟥 throughput [-52619.588op/s; -37973.276op/s] or [-7.347%; -5.302%]

Known flaky benchmarks without significant changes:

  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.ActivityBenchmark.StartSpan net472
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.ActivityBenchmark.StartSpan net6.0
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.ActivityBenchmark.StartSpan netcoreapp3.1
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.ActivityBenchmark.StartSpan_AddEvent_Sampled net472
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.ActivityBenchmark.StartSpan_AddEvent_Sampled net6.0
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.ActivityBenchmark.StartSpan_AddEvent_Sampled netcoreapp3.1
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.ActivityBenchmark.StartSpan_GetContext_Sampled net472
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.ActivityBenchmark.StartSpan_GetContext_Sampled net6.0
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.ActivityBenchmark.StartSpan_GetContext_Sampled netcoreapp3.1
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.ActivityBenchmark.StartSpan_SetAttributes_Sampled net472
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.ActivityBenchmark.StartSpan_SetAttributes_Sampled net6.0
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.ActivityBenchmark.StartSpan_SetAttributes_Sampled netcoreapp3.1
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.ActivityBenchmark.StartSpan_SetStatus_Sampled net472
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.ActivityBenchmark.StartSpan_SetStatus_Sampled net6.0
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.ActivityBenchmark.StartSpan_SetStatus_Sampled netcoreapp3.1
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.ActivityBenchmark.StartSpan_UpdateName_Sampled net472
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.ActivityBenchmark.StartSpan_UpdateName_Sampled net6.0
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.ActivityBenchmark.StartSpan_UpdateName_Sampled netcoreapp3.1
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.TelemetrySpanBenchmark.StartSpan net472
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.TelemetrySpanBenchmark.StartSpan net6.0
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.TelemetrySpanBenchmark.StartSpan netcoreapp3.1
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.TelemetrySpanBenchmark.StartSpan_AddEvent_Sampled net6.0
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.TelemetrySpanBenchmark.StartSpan_AddEvent_Sampled netcoreapp3.1
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.TelemetrySpanBenchmark.StartSpan_GetContext_Sampled net472
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.TelemetrySpanBenchmark.StartSpan_GetContext_Sampled net6.0
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.TelemetrySpanBenchmark.StartSpan_GetContext_Sampled netcoreapp3.1
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.TelemetrySpanBenchmark.StartSpan_RecordException_Sampled net472
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.TelemetrySpanBenchmark.StartSpan_RecordException_Sampled net6.0
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.TelemetrySpanBenchmark.StartSpan_RecordException_Sampled netcoreapp3.1
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.TelemetrySpanBenchmark.StartSpan_SetAttributes_Sampled net472
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.TelemetrySpanBenchmark.StartSpan_SetAttributes_Sampled net6.0
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.TelemetrySpanBenchmark.StartSpan_SetAttributes_Sampled netcoreapp3.1
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.TelemetrySpanBenchmark.StartSpan_SetStatus_Sampled net472
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.TelemetrySpanBenchmark.StartSpan_SetStatus_Sampled net6.0
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.TelemetrySpanBenchmark.StartSpan_SetStatus_Sampled netcoreapp3.1
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.TelemetrySpanBenchmark.StartSpan_UpdateName_Sampled net472
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.TelemetrySpanBenchmark.StartSpan_UpdateName_Sampled net6.0
  • scenario:Benchmarks.OpenTelemetry.InstrumentedApi.Trace.TelemetrySpanBenchmark.StartSpan_UpdateName_Sampled netcoreapp3.1
  • scenario:Benchmarks.Trace.Asm.AppSecBodyBenchmark.ObjectExtractorSimpleBody net472
  • scenario:Benchmarks.Trace.CharSliceBenchmark.OptimizedCharSlice net6.0
  • scenario:Benchmarks.Trace.CharSliceBenchmark.OptimizedCharSlice netcoreapp3.1
  • scenario:Benchmarks.Trace.CharSliceBenchmark.OptimizedCharSliceWithPool net472
  • scenario:Benchmarks.Trace.CharSliceBenchmark.OptimizedCharSliceWithPool net6.0
  • scenario:Benchmarks.Trace.CharSliceBenchmark.OptimizedCharSliceWithPool netcoreapp3.1
  • scenario:Benchmarks.Trace.CharSliceBenchmark.OriginalCharSlice netcoreapp3.1
  • scenario:Benchmarks.Trace.SpanBenchmark.StartFinishScope netcoreapp3.1
  • scenario:Benchmarks.Trace.SpanBenchmark.StartFinishTwoScopes net472
  • scenario:Benchmarks.Trace.SpanBenchmark.StartFinishTwoScopes netcoreapp3.1

@dd-trace-dotnet-ci-bot

dd-trace-dotnet-ci-bot Bot commented Sep 12, 2026 •

Copy link
Copy Markdown

Execution-Time Benchmarks Report ⏱️

Execution-time results for samples comparing This PR (9235) and master.

✅ No regressions detected

📄 View the full report (charts + all metrics) →

@leoromanovsky
leoromanovsky marked this pull request as ready for review September 21, 2026 14:40
@leoromanovsky
leoromanovsky requested review from a team as code owners September 21, 2026 14:40
@leoromanovsky
leoromanovsky requested review from anna-git, danyal002 and vjfridge and removed request for a team September 21, 2026 14:40
@chatgpt-codex-connector

chatgpt-codex-connector Bot commented Sep 21, 2026 •

Copy link
Copy Markdown

Codex Review Summary

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

Review Status Commit Review trigger
📝 Code Review ✅ Completed 2026-09-21T14:52:51.012960Z 7eba078 Draft marked ready
🔒 Security Review ✅ Completed 2026-09-21T14:46:44.892738Z 7eba078 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.

@leoromanovsky
leoromanovsky requested review from pavlokhrebto and removed request for danyal002 September 21, 2026 14:40

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 7eba078530

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".

Comment on lines +162 to +166
[
new(TelemetryConstants.ApiKeyHeader, apiKey),
new(FeatureFlagsEvpHeaderHelper.EvpOriginHeader, FeatureFlagsEvpHeaderHelper.EvpOrigin),
new(FeatureFlagsEvpHeaderHelper.EvpOriginVersionHeader, TracerConstants.ThreePartVersion),
];

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Disable tracing on direct EVP requests

When agentless delivery falls back to direct intake, these are the only default headers installed on the new HttpClient/HttpWebRequest, so the request omits x-datadog-tracing-enabled: false. The HTTP integrations explicitly use that header to suppress instrumentation, and the existing agentless configuration client adds it for the same reason; without it, every direct exposure upload can create an internal HTTP span, potentially attached to the customer trace whose execution context started the send loop. Add the tracing-suppression header to the direct header set as well.

Useful? React with 👍 / 👎.

Comment on lines +516 to +518
if (response.StatusCode is < 200 or >= 300)
{
Log.Warning<int>("Feature Flags direct EVP request failed with HTTP status code {StatusCode}", response.StatusCode);

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P1 Badge Report terminal direct failures at error level

When direct intake returns a non-retryable response such as HTTP 400, this method performs no retry and the exposure batch has already been removed from the queue, yet the terminal failure is logged only as a warning. This hides payload bugs and permanent event loss from error-level diagnostics; log the final response as an error (with the failed endpoint and attempt context), especially for 400 responses.

AGENTS.md reference: AGENTS.md:L192-L197

Useful? React with 👍 / 👎.

Comment on lines +102 to +107
_settingsSubscription = settings.Manager.SubscribeToChanges(changes =>
{
if (changes.UpdatedExporter is { } exporter)
{
Interlocked.Exchange(ref _localRequestFactory!, CreateLocalRequestFactory(exporter));
}

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Revalidate EVP capabilities after changing the agent endpoint

When runtime configuration changes the agent URL or transport, this callback swaps _localRequestFactory but retains _localProxyEndpoint and _discoveryKnown from the previous agent. Until the shared discovery loop polls again, exposures are therefore sent to the new relay using capabilities established for the old one; if the new agent exposes the route but does not forward the two origin headers, it silently strips the producer identity that this transport otherwise requires before selecting a local route. Invalidate the cached route/discovery state when replacing the factory and wait for discovery of the new endpoint.

Useful? React with 👍 / 👎.

Comment on lines +381 to +385
? configuration.EventPlatformProxyEndpoint switch
{
EventPlatformProxyV4 => EventPlatformProxyV4,
EventPlatformProxyV2 => EventPlatformProxyV2,
_ => null,

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Match discovered EVP routes case-insensitively

If /info advertises a supported route with different casing, DiscoveryService accepts it using OrdinalIgnoreCase and preserves the advertised spelling in EventPlatformProxyEndpoint, but this case-sensitive switch then rejects it and unnecessarily falls back to direct intake or drops events when credentials are unavailable. Classify v2/v4 with the same case-insensitive comparison used during discovery.

Useful? React with 👍 / 👎.

Comment on lines +87 to +91
// The shared discovery service owns its own bounded retry/backoff loop. Event flushes
// wait for its first result once and then consume later callbacks; they never start an
// independent /info request on every flush.
_discoveryService.SubscribeToChanges(_discoveryCallback);
_discoverySubscribed = true;

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Skip the discovery wait when polling is disabled

When DD_AGENT_FEATURE_POLLING_ENABLED=false, TracerManagerFactory supplies NullDiscoveryService, whose subscription is a no-op, but this constructor still leaves _discoveryKnown unset. The first direct exposure consequently waits the full five-second discovery timeout even though no callback can ever arrive; in short-lived or abruptly frozen processes this delays or loses the first batch unnecessarily. Detect the null discovery implementation and mark discovery complete so direct delivery can start immediately.

Useful? React with 👍 / 👎.

Prefer identity-capable local relays and keep a credential-isolated direct route with conservative replay and sticky selection.

Environment: Datadog workspace
@leoromanovsky
leoromanovsky force-pushed the leo.romanovsky/fflsdk-189-evp-fallback-dotnet-master branch from f439192 to 8d55124 Compare October 1, 2026 22:29
@leoromanovsky
leoromanovsky changed the base branch from master to leo.romanovsky/fflsdk-189-dotnet-exposure-lifecycle October 1, 2026 22:31

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant