Skip to content

[DSM] Switch from ManualResetEventSlim to AsyncManualResetEvent#6356

Merged
andrewlock merged 1 commit into
masterfrom
andrew/dsm-thread-use
Nov 27, 2024
Merged

[DSM] Switch from ManualResetEventSlim to AsyncManualResetEvent#6356
andrewlock merged 1 commit into
masterfrom
andrew/dsm-thread-use

Conversation

@andrewlock

@andrewlock andrewlock commented Nov 26, 2024

Copy link
Copy Markdown
Member

Summary of changes

Use AsyncManualResetEvent instead of ManualResetEventSlim in the DSM writer

Reason for change

Using ManualResetEventSlim means we hog a ThreadPool thread for a potentially long time. Using async/await releases it.

Implementation details

We have already vendored a AsyncManualResetEvent based on the AsyncEx version so just switched to that

Test coverage

Covered by existing, if it passes 👍

@andrewlock
andrewlock requested review from a team as code owners November 26, 2024 14:25
@github-actions github-actions Bot added the area:tracer The core tracer library (Datadog.Trace, does not include OpenTracing, native code, or integrations) label Nov 26, 2024
bouwkast
bouwkast approved these changes Nov 26, 2024
View reviewed changes
Comment thread tracer/src/Datadog.Trace/DataStreamsMonitoring/DataStreamsWriter.cs
}

_processingMutex.Wait();
await _processingMutex.WaitAsync().ConfigureAwait(false);

@bouwkast bouwkast Nov 26, 2024

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

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

So just to clarify, we sit on the thread pool for this function (compared to SerializeTracesLoop) because we start this function with _processTask = Task.Run(ProcessQueueLoopAsync); where the other one we do a _serializationTask = automaticFlush ? Task.Factory.StartNew(SerializeTracesLoop, TaskCreationOptions.LongRunning)

Which gets it out of the thread pool.

Should we do the same for both and update that one to this as well?

@andrewlock andrewlock Nov 26, 2024

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

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

Should we do the same for both and update that one to this as well?

As I understand it, that wouldn't work here because we're using async/await. Using LongRunning makes sense for the non-async method, as it (potentially) stays out of the pool entirely, but when you use async/await you lose the benefit seeing as you're going to be resuming on different threads potentially anyway.

cc @kevingosse to sense check me 😄

@datadog-ddstaging

datadog-ddstaging Bot commented Nov 26, 2024 •
edited
Loading

Copy link
Copy Markdown

Datadog Report

Branch report: andrew/dsm-thread-use
Commit report: 2aa1558
Test service: dd-trace-dotnet

✅ 0 Failed, 448371 Passed, 2721 Skipped, 19h 16m 35.41s Total Time

@andrewlock

andrewlock commented Nov 26, 2024

Copy link
Copy Markdown
Member Author

Execution-Time Benchmarks Report ⏱️

Execution-time results for samples comparing the following branches/commits:

Execution-time benchmarks measure the whole time it takes to execute a program. And are intended to measure the one-off costs. Cases where the execution time results for the PR are worse than latest master results are shown in red. The following thresholds were used for comparing the execution times:

  • Welch test with statistical test for significance of 5%
  • Only results indicating a difference greater than 5% and 5 ms are considered.

Note that these results are based on a single point-in-time result for each branch. For full results, see the dashboard.

Graphs show the p99 interval based on the mean and StdDev of the test run, as well as the mean value of the run (shown as a diamond below the graph).

gantt
    title Execution time (ms) FakeDbCommand (.NET Framework 4.6.2) 
    dateFormat  X
    axisFormat %s
    todayMarker off
    section Baseline
    This PR (6356) - mean (76ms)  : 72, 80
     .   : milestone, 76,
    master - mean (69ms)  : 67, 70
     .   : milestone, 69,

    section CallTarget+Inlining+NGEN
    This PR (6356) - mean (1,033ms)  : crit, 1018, 1048
     .   : crit, milestone, 1033,
    master - mean (975ms)  : 954, 996
     .   : milestone, 975,
Loading 
gantt
    title Execution time (ms) FakeDbCommand (.NET Core 3.1) 
    dateFormat  X
    axisFormat %s
    todayMarker off
    section Baseline
    This PR (6356) - mean (119ms)  : 113, 124
     .   : milestone, 119,
    master - mean (108ms)  : 105, 110
     .   : milestone, 108,

    section CallTarget+Inlining+NGEN
    This PR (6356) - mean (720ms)  : crit, 708, 733
     .   : crit, milestone, 720,
    master - mean (679ms)  : 666, 691
     .   : milestone, 679,
Loading 
gantt
    title Execution time (ms) FakeDbCommand (.NET 6) 
    dateFormat  X
    axisFormat %s
    todayMarker off
    section Baseline
    This PR (6356) - mean (102ms)  : 95, 108
     .   : milestone, 102,
    master - mean (91ms)  : 88, 94
     .   : milestone, 91,

    section CallTarget+Inlining+NGEN
    This PR (6356) - mean (673ms)  : crit, 655, 692
     .   : crit, milestone, 673,
    master - mean (630ms)  : 615, 646
     .   : milestone, 630,
Loading 
gantt
    title Execution time (ms) HttpMessageHandler (.NET Framework 4.6.2) 
    dateFormat  X
    axisFormat %s
    todayMarker off
    section Baseline
    This PR (6356) - mean (191ms)  : 186, 196
     .   : milestone, 191,
    master - mean (190ms)  : 186, 194
     .   : milestone, 190,

    section CallTarget+Inlining+NGEN
    This PR (6356) - mean (1,096ms)  : 1070, 1122
     .   : milestone, 1096,
    master - mean (1,094ms)  : 1064, 1124
     .   : milestone, 1094,
Loading 
gantt
    title Execution time (ms) HttpMessageHandler (.NET Core 3.1) 
    dateFormat  X
    axisFormat %s
    todayMarker off
    section Baseline
    This PR (6356) - mean (278ms)  : 271, 284
     .   : milestone, 278,
    master - mean (277ms)  : 272, 282
     .   : milestone, 277,

    section CallTarget+Inlining+NGEN
    This PR (6356) - mean (870ms)  : 842, 898
     .   : milestone, 870,
    master - mean (871ms)  : 837, 906
     .   : milestone, 871,
Loading 
gantt
    title Execution time (ms) HttpMessageHandler (.NET 6) 
    dateFormat  X
    axisFormat %s
    todayMarker off
    section Baseline
    This PR (6356) - mean (267ms)  : 262, 273
     .   : milestone, 267,
    master - mean (266ms)  : 261, 271
     .   : milestone, 266,

    section CallTarget+Inlining+NGEN
    This PR (6356) - mean (851ms)  : 816, 886
     .   : milestone, 851,
    master - mean (851ms)  : 810, 893
     .   : milestone, 851,
Loading

@lucaspimentel lucaspimentel changed the title Switch to AsyncManualResetEvent instead [DSM] Switch from ManualResetEventSlim to AsyncManualResetEvent Nov 26, 2024
@andrewlock

andrewlock commented Nov 26, 2024

Copy link
Copy Markdown
Member Author

Benchmarks Report for tracer 🐌

Benchmarks for #6356 compared to master:

  • 1 benchmarks are faster, with geometric mean 1.121
  • 1 benchmarks are slower, with geometric mean 1.132
  • All benchmarks have the same allocations

The following thresholds were used for comparing the benchmark speeds:

  • Mann–Whitney U test with statistical test for significance of 5%
  • Only results indicating a difference greater than 10% and 0.3 ns are considered.

Allocation changes below 0.5% are ignored.

Benchmark details

Benchmarks.Trace.ActivityBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master StartStopWithChild net6.0 7.78μs 42.5ns 240ns 0.0153 0.00763 0 5.61 KB
master StartStopWithChild netcoreapp3.1 10.1μs 54.5ns 308ns 0.0198 0.00988 0 5.8 KB
master StartStopWithChild net472 16.2μs 36.4ns 141ns 1.04 0.305 0.104 6.21 KB
#6356 StartStopWithChild net6.0 8.15μs 45.9ns 304ns 0.012 0.00402 0 5.61 KB
#6356 StartStopWithChild netcoreapp3.1 10μs 55.6ns 381ns 0.0197 0.00983 0 5.8 KB
#6356 StartStopWithChild net472 16.4μs 49.5ns 192ns 1.04 0.304 0.0986 6.22 KB
Benchmarks.Trace.AgentWriterBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master WriteAndFlushEnrichedTraces net6.0 484μs 202ns 729ns 0 0 0 2.7 KB
master WriteAndFlushEnrichedTraces netcoreapp3.1 662μs 256ns 992ns 0 0 0 2.7 KB
master WriteAndFlushEnrichedTraces net472 854μs 501ns 1.87μs 0.428 0 0 3.3 KB
#6356 WriteAndFlushEnrichedTraces net6.0 484μs 355ns 1.33μs 0 0 0 2.7 KB
#6356 WriteAndFlushEnrichedTraces netcoreapp3.1 672μs 377ns 1.46μs 0 0 0 2.7 KB
#6356 WriteAndFlushEnrichedTraces net472 834μs 354ns 1.37μs 0.417 0 0 3.3 KB
Benchmarks.Trace.AspNetCoreBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master SendRequest net6.0 149μs 850ns 6.47μs 0.168 0 0 14.47 KB
master SendRequest netcoreapp3.1 163μs 881ns 5.71μs 0.165 0 0 17.27 KB
master SendRequest net472 0.00132ns 0.000695ns 0.0026ns 0 0 0 0 b
#6356 SendRequest net6.0 141μs 615ns 2.3μs 0.149 0 0 14.47 KB
#6356 SendRequest netcoreapp3.1 168μs 970ns 9.35μs 0.157 0 0 17.27 KB
#6356 SendRequest net472 0ns 0ns 0ns 0 0 0 0 b
Benchmarks.Trace.CIVisibilityProtocolWriterBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master WriteAndFlushEnrichedTraces net6.0 558μs 2.02μs 7μs 0.573 0 0 41.62 KB
master WriteAndFlushEnrichedTraces netcoreapp3.1 678μs 3.49μs 17.1μs 0.338 0 0 41.82 KB
master WriteAndFlushEnrichedTraces net472 847μs 4.16μs 17.7μs 8.13 2.57 0.428 53.29 KB
#6356 WriteAndFlushEnrichedTraces net6.0 580μs 2.98μs 15.5μs 0.573 0 0 41.69 KB
#6356 WriteAndFlushEnrichedTraces netcoreapp3.1 696μs 3.95μs 30.1μs 0.357 0 0 41.96 KB
#6356 WriteAndFlushEnrichedTraces net472 857μs 4.33μs 20.3μs 8.13 2.57 0.428 53.29 KB
Benchmarks.Trace.DbCommandBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master ExecuteNonQuery net6.0 1.23μs 0.817ns 2.94ns 0.0141 0 0 1.02 KB
master ExecuteNonQuery netcoreapp3.1 1.76μs 2.84ns 9.85ns 0.0138 0 0 1.02 KB
master ExecuteNonQuery net472 2.16μs 1.99ns 7.71ns 0.157 0.00108 0 987 B
#6356 ExecuteNonQuery net6.0 1.33μs 1.19ns 4.61ns 0.0139 0 0 1.02 KB
#6356 ExecuteNonQuery netcoreapp3.1 1.78μs 1.37ns 5.31ns 0.0133 0 0 1.02 KB
#6356 ExecuteNonQuery net472 2.1μs 1.68ns 6.06ns 0.157 0.00105 0 987 B
Benchmarks.Trace.ElasticsearchBenchmark - Faster 🎉 Same allocations ✔️

Faster 🎉 in #6356

Benchmark base/diff Base Median (ns) Diff Median (ns) Modality
Benchmarks.Trace.ElasticsearchBenchmark.CallElasticsearchAsync‑net6.0 1.121 1,399.23 1,248.56

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master CallElasticsearch net6.0 1.18μs 0.683ns 2.56ns 0.0136 0 0 976 B
master CallElasticsearch netcoreapp3.1 1.47μs 0.638ns 2.39ns 0.0127 0 0 976 B
master CallElasticsearch net472 2.5μs 1.38ns 5.14ns 0.158 0 0 995 B
master CallElasticsearchAsync net6.0 1.41μs 4.89ns 19ns 0.0135 0 0 952 B
master CallElasticsearchAsync netcoreapp3.1 1.75μs 0.81ns 2.92ns 0.014 0 0 1.02 KB
master CallElasticsearchAsync net472 2.7μs 1.69ns 6.32ns 0.167 0 0 1.05 KB
#6356 CallElasticsearch net6.0 1.17μs 0.502ns 1.94ns 0.0134 0 0 976 B
#6356 CallElasticsearch netcoreapp3.1 1.55μs 0.654ns 2.45ns 0.013 0 0 976 B
#6356 CallElasticsearch net472 2.6μs 2.11ns 8.16ns 0.158 0 0 995 B
#6356 CallElasticsearchAsync net6.0 1.25μs 0.634ns 2.37ns 0.0131 0 0 952 B
#6356 CallElasticsearchAsync netcoreapp3.1 1.64μs 0.773ns 3ns 0.0139 0 0 1.02 KB
#6356 CallElasticsearchAsync net472 2.58μs 2.21ns 8.57ns 0.166 0 0 1.05 KB
Benchmarks.Trace.GraphQLBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master ExecuteAsync net6.0 1.28μs 0.408ns 1.53ns 0.0135 0 0 952 B
master ExecuteAsync netcoreapp3.1 1.6μs 0.646ns 2.5ns 0.0127 0 0 952 B
master ExecuteAsync net472 1.86μs 1.25ns 4.84ns 0.144 0 0 915 B
#6356 ExecuteAsync net6.0 1.26μs 0.577ns 2.16ns 0.0132 0 0 952 B
#6356 ExecuteAsync netcoreapp3.1 1.64μs 0.58ns 2.17ns 0.013 0 0 952 B
#6356 ExecuteAsync net472 1.9μs 0.456ns 1.71ns 0.145 0 0 915 B
Benchmarks.Trace.HttpClientBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master SendAsync net6.0 4.35μs 2.07ns 8.01ns 0.0327 0 0 2.31 KB
master SendAsync netcoreapp3.1 5.38μs 3.19ns 11.9ns 0.0375 0 0 2.85 KB
master SendAsync net472 7.33μs 1.78ns 6.9ns 0.494 0 0 3.12 KB
#6356 SendAsync net6.0 4.45μs 2.03ns 7.86ns 0.0311 0 0 2.31 KB
#6356 SendAsync netcoreapp3.1 5.27μs 4.28ns 16.6ns 0.0369 0 0 2.85 KB
#6356 SendAsync net472 7.3μs 2.44ns 9.47ns 0.493 0 0 3.12 KB
Benchmarks.Trace.ILoggerBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master EnrichedLog net6.0 1.57μs 0.551ns 1.99ns 0.0228 0 0 1.64 KB
master EnrichedLog netcoreapp3.1 2.31μs 1.37ns 4.94ns 0.0216 0 0 1.64 KB
master EnrichedLog net472 2.56μs 1.45ns 5.62ns 0.249 0 0 1.57 KB
#6356 EnrichedLog net6.0 1.49μs 0.796ns 3.08ns 0.0232 0 0 1.64 KB
#6356 EnrichedLog netcoreapp3.1 2.24μs 1.37ns 5.14ns 0.0222 0 0 1.64 KB
#6356 EnrichedLog net472 2.44μs 0.93ns 3.48ns 0.249 0 0 1.57 KB
Benchmarks.Trace.Log4netBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master EnrichedLog net6.0 120μs 114ns 440ns 0.0602 0 0 4.28 KB
master EnrichedLog netcoreapp3.1 122μs 156ns 584ns 0 0 0 4.28 KB
master EnrichedLog net472 151μs 267ns 1.03μs 0.673 0.224 0 4.46 KB
#6356 EnrichedLog net6.0 118μs 146ns 564ns 0 0 0 4.28 KB
#6356 EnrichedLog netcoreapp3.1 125μs 114ns 442ns 0 0 0 4.28 KB
#6356 EnrichedLog net472 152μs 59.8ns 224ns 0.683 0.228 0 4.46 KB
Benchmarks.Trace.NLogBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master EnrichedLog net6.0 3.12μs 0.905ns 3.5ns 0.0311 0 0 2.2 KB
master EnrichedLog netcoreapp3.1 4.38μs 1.46ns 5.45ns 0.0306 0 0 2.2 KB
master EnrichedLog net472 4.88μs 1.47ns 5.69ns 0.319 0 0 2.02 KB
#6356 EnrichedLog net6.0 2.99μs 0.92ns 3.44ns 0.0302 0 0 2.2 KB
#6356 EnrichedLog netcoreapp3.1 4.09μs 1.41ns 5.47ns 0.0286 0 0 2.2 KB
#6356 EnrichedLog net472 4.81μs 1.02ns 3.82ns 0.321 0 0 2.02 KB
Benchmarks.Trace.RedisBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master SendReceive net6.0 1.35μs 0.558ns 2.16ns 0.0162 0 0 1.14 KB
master SendReceive netcoreapp3.1 1.72μs 0.984ns 3.81ns 0.0147 0 0 1.14 KB
master SendReceive net472 2.1μs 1.87ns 7.25ns 0.183 0 0 1.16 KB
#6356 SendReceive net6.0 1.34μs 1.53ns 5.93ns 0.0162 0 0 1.14 KB
#6356 SendReceive netcoreapp3.1 1.73μs 0.487ns 1.82ns 0.0156 0 0 1.14 KB
#6356 SendReceive net472 2.15μs 1.64ns 6.34ns 0.183 0 0 1.16 KB
Benchmarks.Trace.SerilogBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master EnrichedLog net6.0 2.79μs 0.924ns 3.58ns 0.0224 0 0 1.6 KB
master EnrichedLog netcoreapp3.1 3.87μs 1.35ns 5.21ns 0.0213 0 0 1.65 KB
master EnrichedLog net472 4.36μs 1.97ns 7.38ns 0.323 0 0 2.04 KB
#6356 EnrichedLog net6.0 2.72μs 1.15ns 4.45ns 0.0218 0 0 1.6 KB
#6356 EnrichedLog netcoreapp3.1 3.92μs 0.863ns 3.23ns 0.0217 0 0 1.65 KB
#6356 EnrichedLog net472 4.35μs 4.84ns 18.7ns 0.322 0 0 2.04 KB
Benchmarks.Trace.SpanBenchmark - Slower ⚠️ Same allocations ✔️

Slower ⚠️ in #6356

Benchmark diff/base Base Median (ns) Diff Median (ns) Modality
Benchmarks.Trace.SpanBenchmark.StartFinishScope‑net6.0 1.132 475.52 538.44

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master StartFinishSpan net6.0 447ns 0.985ns 3.81ns 0.00811 0 0 576 B
master StartFinishSpan netcoreapp3.1 581ns 1.33ns 4.98ns 0.00753 0 0 576 B
master StartFinishSpan net472 585ns 0.801ns 3.1ns 0.0917 0 0 578 B
master StartFinishScope net6.0 476ns 0.665ns 2.58ns 0.00966 0 0 696 B
master StartFinishScope netcoreapp3.1 729ns 1.21ns 4.38ns 0.00953 0 0 696 B
master StartFinishScope net472 860ns 2.31ns 8.94ns 0.104 0 0 658 B
#6356 StartFinishSpan net6.0 404ns 0.604ns 2.34ns 0.00802 0 0 576 B
#6356 StartFinishSpan netcoreapp3.1 624ns 1.6ns 6.2ns 0.00772 0 0 576 B
#6356 StartFinishSpan net472 619ns 0.817ns 3.16ns 0.0917 0 0 578 B
#6356 StartFinishScope net6.0 539ns 0.653ns 2.53ns 0.00964 0 0 696 B
#6356 StartFinishScope netcoreapp3.1 752ns 0.829ns 3.21ns 0.00924 0 0 696 B
#6356 StartFinishScope net472 885ns 2.15ns 8.32ns 0.104 0 0 658 B
Benchmarks.Trace.TraceAnnotationsBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master RunOnMethodBegin net6.0 602ns 1.3ns 5.03ns 0.00969 0 0 696 B
master RunOnMethodBegin netcoreapp3.1 895ns 2.22ns 8.32ns 0.00956 0 0 696 B
master RunOnMethodBegin net472 1.13μs 2.37ns 9.18ns 0.104 0 0 658 B
#6356 RunOnMethodBegin net6.0 671ns 1.25ns 4.83ns 0.00977 0 0 696 B
#6356 RunOnMethodBegin netcoreapp3.1 932ns 1.93ns 7.48ns 0.00931 0 0 696 B
#6356 RunOnMethodBegin net472 1.11μs 1.31ns 5.06ns 0.104 0 0 658 B

@andrewlock

andrewlock commented Nov 26, 2024

Copy link
Copy Markdown
Member Author

Throughput/Crank Report ⚡

Throughput results for AspNetCoreSimpleController comparing the following branches/commits:

Cases where throughput results for the PR are worse than latest master (5% drop or greater), results are shown in red.

Note that these results are based on a single point-in-time result for each branch. For full results, see one of the many, many dashboards!

gantt
    title Throughput Linux x64 (Total requests) 
    dateFormat  X
    axisFormat %s
    section Baseline
    This PR (6356) (11.250M)   : 0, 11249747
    master (10.987M)   : 0, 10986728
    benchmarks/2.9.0 (11.033M)   : 0, 11032866

    section Automatic
    This PR (6356) (7.149M)   : 0, 7149415
    master (7.175M)   : 0, 7174684
    benchmarks/2.9.0 (7.786M)   : 0, 7785853

    section Trace stats
    master (7.577M)   : 0, 7577344

    section Manual
    master (11.201M)   : 0, 11201309

    section Manual + Automatic
    This PR (6356) (6.740M)   : 0, 6739672
    master (6.769M)   : 0, 6769214

    section DD_TRACE_ENABLED=0
    master (10.257M)   : 0, 10257268
Loading 
gantt
    title Throughput Linux arm64 (Total requests) 
    dateFormat  X
    axisFormat %s
    section Baseline
    This PR (6356) (9.438M)   : 0, 9437808
    master (9.557M)   : 0, 9557319
    benchmarks/2.9.0 (9.495M)   : 0, 9494821

    section Automatic
    This PR (6356) (6.489M)   : 0, 6489097
    master (6.368M)   : 0, 6367805

    section Trace stats
    master (6.673M)   : 0, 6672869

    section Manual
    master (9.631M)   : 0, 9630961

    section Manual + Automatic
    This PR (6356) (5.993M)   : 0, 5993283
    master (5.951M)   : 0, 5951439

    section DD_TRACE_ENABLED=0
    master (8.900M)   : 0, 8899929
Loading 
gantt
    title Throughput Windows x64 (Total requests) 
    dateFormat  X
    axisFormat %s
    section Baseline
    This PR (6356) (10.497M)   : 0, 10497333
    master (9.199M)   : 0, 9198633
    benchmarks/2.9.0 (10.020M)   : 0, 10019592

    section Automatic
    This PR (6356) (6.844M)   : 0, 6843518
    master (6.083M)   : 0, 6082614
    benchmarks/2.9.0 (7.255M)   : 0, 7255257

    section Trace stats
    master (6.603M)   : 0, 6602631

    section Manual
    master (9.506M)   : 0, 9505886

    section Manual + Automatic
    This PR (6356) (6.326M)   : 0, 6325779
    master (5.789M)   : 0, 5788592

    section DD_TRACE_ENABLED=0
    master (8.972M)   : 0, 8971897
Loading

@andrewlock
andrewlock merged commit fb7f5ad into master Nov 27, 2024
@andrewlock
andrewlock deleted the andrew/dsm-thread-use branch November 27, 2024 11:54
@github-actions github-actions Bot added this to the vNext-v3 milestone Nov 27, 2024
veerbia pushed a commit that referenced this pull request Dec 16, 2024
@andrewlock @veerbia
[DSM] Switch from ManualResetEventSlim to AsyncManualResetEvent (#…
cf6e890 
…6356)

## Summary of changes

Use `AsyncManualResetEvent` instead of `ManualResetEventSlim` in the DSM
writer

## Reason for change

Using `ManualResetEventSlim` means we hog a ThreadPool thread for a
potentially long time. Using `async`/`await` releases it.

## Implementation details

We have already vendored a `AsyncManualResetEvent` based on [the AsyncEx
version](https://github.com/StephenCleary/AsyncEx/blob/master/src/Nito.AsyncEx.Coordination/AsyncManualResetEvent.cs)
so just switched to that

## Test coverage

Covered by existing, if it passes 👍
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@lucaspimentel lucaspimentel lucaspimentel approved these changes
@bouwkast bouwkast bouwkast approved these changes

Assignees

No one assigned

Labels

area:data-streams-monitoring area:tracer The core tracer library (Datadog.Trace, does not include OpenTracing, native code, or integrations) type:refactor

Projects

None yet

Milestone

3.7.0

Development

Successfully merging this pull request may close these issues.

3 participants

@andrewlock @lucaspimentel @bouwkast