[Test Optimization] perf: logging/timing helpers by tonyredondo · Pull Request #8070 · DataDog/dd-trace-dotnet

tonyredondo · 2026-01-15T16:51:18Z

Summary of changes

Add FixedSizeArrayPool and logger overloads to reduce per-log allocations.
Add RefStopwatch/StopwatchHelpers/CodeDuration and apply timing to TraceClock/TracerManagerFactory/Instrumentation and CI processors.
Lower init logs to Debug for CI protocol writer/sender and processors.

Reason for change

Reduce allocations and improve timing diagnostics in hot paths.

Implementation details

New utility helpers in Datadog.Trace.Util for allocation-free timing and pooling.
Logging changes to avoid params array allocations in common log paths.
Timing/log-level tweaks in CI protocol writer/sender and trace processors.

Test coverage

CI passes then all changes are good.

Other details

Stacked PRs (current marked):

pr-commenter · 2026-01-15T17:50:23Z

Benchmarks

Benchmark execution time: 2026-01-30 12:13:33

Comparing candidate commit 90b2923 in PR branch tony/testoptimization-performance-work-pr1 with baseline commit 3d020db in branch master.

Found 3 performance improvements and 14 performance regressions! Performance is the same for 160 metrics, 15 unstable metrics.

scenario:Benchmarks.Trace.ActivityBenchmark.StartStopWithChild net472

🟥 throughput [-6314.778op/s; -5796.540op/s] or [-7.085%; -6.504%]

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

🟩 throughput [+4752.715op/s; +6232.273op/s] or [+5.178%; +6.790%]

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

🟥 execution_time [+108.169ms; +108.888ms] or [+98.219%; +98.871%]

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

🟥 execution_time [+27.082ms; +27.195ms] or [+15.275%; +15.338%]

scenario:Benchmarks.Trace.CIVisibilityProtocolWriterBenchmark.WriteAndFlushEnrichedTraces net472

🟩 execution_time [-20.612ms; -14.769ms] or [-9.567%; -6.855%]

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

🟥 execution_time [+101.419µs; +107.474µs] or [+7.135%; +7.560%]
🟥 throughput [-49.477op/s; -46.811op/s] or [-7.033%; -6.654%]

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

🟥 throughput [-38376.427op/s; -31108.806op/s] or [-8.951%; -7.256%]

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

🟥 execution_time [+12.758ms; +17.320ms] or [+6.472%; +8.786%]

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

🟥 throughput [-511.737op/s; -309.780op/s] or [-23.043%; -13.949%]

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

🟥 throughput [-53024.493op/s; -40037.771op/s] or [-9.904%; -7.479%]

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

🟥 execution_time [+18.494ms; +19.309ms] or [+20.255%; +21.147%]

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

🟥 execution_time [+12.735ms; +16.204ms] or [+6.390%; +8.131%]
🟥 throughput [-97292.614op/s; -79890.252op/s] or [-8.818%; -7.241%]

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

🟥 execution_time [+16.183ms; +21.281ms] or [+8.102%; +10.655%]

scenario:Benchmarks.Trace.TraceAnnotationsBenchmark.RunOnMethodBegin net472

🟥 throughput [-92234.851op/s; -89630.979op/s] or [-12.805%; -12.443%]

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

🟩 execution_time [-21.575ms; -17.679ms] or [-9.973%; -8.172%]

dd-trace-dotnet-ci-bot · 2026-01-15T23:13:52Z

Execution-Time Benchmarks Report ⏱️

Execution-time results for samples comparing This PR (8070) and master.

⚠️ Potential regressions detected

FakeDbCommand

Metric	Master (Mean ± 95% CI)	Current (Mean ± 95% CI)	Change	Status
.NET Framework 4.8 - Bailout
duration	72.43 ± (72.33 - 72.57) ms	77.76 ± (77.54 - 77.92) ms	+7.4%	❌⬆️

Full Metrics Comparison

FakeDbCommand

Metric	Master (Mean ± 95% CI)	Current (Mean ± 95% CI)	Change	Status
.NET Framework 4.8 - Baseline
duration	68.49 ± (68.52 - 68.84) ms	73.16 ± (73.21 - 73.56) ms	+6.8%	✅⬆️
.NET Framework 4.8 - Bailout
duration	72.43 ± (72.33 - 72.57) ms	77.76 ± (77.54 - 77.92) ms	+7.4%	❌⬆️
.NET Framework 4.8 - CallTarget+Inlining+NGEN
duration	1023.25 ± (1029.36 - 1039.28) ms	1053.82 ± (1055.12 - 1062.71) ms	+3.0%	✅⬆️
.NET Core 3.1 - Baseline
process.internal_duration_ms	21.86 ± (21.83 - 21.88) ms	22.71 ± (22.66 - 22.75) ms	+3.9%	✅⬆️
process.time_to_main_ms	78.76 ± (78.61 - 78.92) ms	86.37 ± (86.17 - 86.57) ms	+9.7%	✅⬆️
runtime.dotnet.exceptions.count	0 ± (0 - 0)	0 ± (0 - 0)	+0.0%	✅
runtime.dotnet.mem.committed	10.90 ± (10.90 - 10.91) MB	10.92 ± (10.91 - 10.92) MB	+0.1%	✅⬆️
runtime.dotnet.threads.count	12 ± (12 - 12)	12 ± (12 - 12)	+0.0%	✅
.NET Core 3.1 - Bailout
process.internal_duration_ms	21.75 ± (21.73 - 21.77) ms	22.59 ± (22.54 - 22.63) ms	+3.8%	✅⬆️
process.time_to_main_ms	80.02 ± (79.91 - 80.12) ms	87.34 ± (87.16 - 87.52) ms	+9.2%	✅⬆️
runtime.dotnet.exceptions.count	0 ± (0 - 0)	0 ± (0 - 0)	+0.0%	✅
runtime.dotnet.mem.committed	10.94 ± (10.94 - 10.94) MB	10.95 ± (10.95 - 10.96) MB	+0.1%	✅⬆️
runtime.dotnet.threads.count	13 ± (13 - 13)	13 ± (13 - 13)	+0.0%	✅
.NET Core 3.1 - CallTarget+Inlining+NGEN
process.internal_duration_ms	241.72 ± (238.00 - 245.43) ms	238.84 ± (234.92 - 242.77) ms	-1.2%	✅
process.time_to_main_ms	474.07 ± (473.38 - 474.76) ms	505.28 ± (504.39 - 506.18) ms	+6.6%	✅⬆️
runtime.dotnet.exceptions.count	0 ± (0 - 0)	0 ± (0 - 0)	+0.0%	✅
runtime.dotnet.mem.committed	48.56 ± (48.54 - 48.59) MB	48.58 ± (48.56 - 48.60) MB	+0.0%	✅⬆️
runtime.dotnet.threads.count	28 ± (28 - 28)	28 ± (28 - 28)	-0.8%	✅
.NET 6 - Baseline
process.internal_duration_ms	20.65 ± (20.62 - 20.69) ms	21.56 ± (21.52 - 21.60) ms	+4.4%	✅⬆️
process.time_to_main_ms	68.36 ± (68.23 - 68.49) ms	74.83 ± (74.65 - 75.00) ms	+9.5%	✅⬆️
runtime.dotnet.exceptions.count	0 ± (0 - 0)	0 ± (0 - 0)	+0.0%	✅
runtime.dotnet.mem.committed	10.62 ± (10.61 - 10.62) MB	10.64 ± (10.64 - 10.64) MB	+0.2%	✅⬆️
runtime.dotnet.threads.count	10 ± (10 - 10)	10 ± (10 - 10)	+0.0%	✅
.NET 6 - Bailout
process.internal_duration_ms	20.52 ± (20.49 - 20.54) ms	21.44 ± (21.39 - 21.49) ms	+4.5%	✅⬆️
process.time_to_main_ms	69.27 ± (69.21 - 69.33) ms	75.20 ± (75.02 - 75.38) ms	+8.6%	✅⬆️
runtime.dotnet.exceptions.count	0 ± (0 - 0)	0 ± (0 - 0)	+0.0%	✅
runtime.dotnet.mem.committed	10.67 ± (10.67 - 10.67) MB	10.75 ± (10.75 - 10.76) MB	+0.8%	✅⬆️
runtime.dotnet.threads.count	11 ± (11 - 11)	11 ± (11 - 11)	+0.0%	✅
.NET 6 - CallTarget+Inlining+NGEN
process.internal_duration_ms	250.13 ± (249.04 - 251.22) ms	238.31 ± (234.35 - 242.26) ms	-4.7%	✅
process.time_to_main_ms	451.92 ± (451.43 - 452.42) ms	478.00 ± (477.30 - 478.70) ms	+5.8%	✅⬆️
runtime.dotnet.exceptions.count	0 ± (0 - 0)	0 ± (0 - 0)	+0.0%	✅
runtime.dotnet.mem.committed	49.29 ± (49.26 - 49.32) MB	49.30 ± (49.27 - 49.32) MB	+0.0%	✅⬆️
runtime.dotnet.threads.count	28 ± (28 - 28)	28 ± (28 - 28)	-0.1%	✅
.NET 8 - Baseline
process.internal_duration_ms	18.70 ± (18.67 - 18.72) ms	19.66 ± (19.62 - 19.69) ms	+5.1%	✅⬆️
process.time_to_main_ms	67.05 ± (66.95 - 67.16) ms	73.73 ± (73.56 - 73.90) ms	+10.0%	✅⬆️
runtime.dotnet.exceptions.count	0 ± (0 - 0)	0 ± (0 - 0)	+0.0%	✅
runtime.dotnet.mem.committed	7.69 ± (7.68 - 7.69) MB	7.69 ± (7.68 - 7.70) MB	+0.0%	✅⬆️
runtime.dotnet.threads.count	10 ± (10 - 10)	10 ± (10 - 10)	+0.0%	✅
.NET 8 - Bailout
process.internal_duration_ms	18.74 ± (18.71 - 18.76) ms	19.72 ± (19.68 - 19.77) ms	+5.3%	✅⬆️
process.time_to_main_ms	68.32 ± (68.25 - 68.38) ms	75.39 ± (75.22 - 75.57) ms	+10.4%	✅⬆️
runtime.dotnet.exceptions.count	0 ± (0 - 0)	0 ± (0 - 0)	+0.0%	✅
runtime.dotnet.mem.committed	7.73 ± (7.72 - 7.74) MB	7.73 ± (7.73 - 7.74) MB	+0.1%	✅⬆️
runtime.dotnet.threads.count	11 ± (11 - 11)	11 ± (11 - 11)	+0.0%	✅
.NET 8 - CallTarget+Inlining+NGEN
process.internal_duration_ms	178.01 ± (177.28 - 178.75) ms	187.74 ± (187.01 - 188.47) ms	+5.5%	✅⬆️
process.time_to_main_ms	435.35 ± (434.71 - 436.00) ms	459.31 ± (458.52 - 460.09) ms	+5.5%	✅⬆️
runtime.dotnet.exceptions.count	0 ± (0 - 0)	0 ± (0 - 0)	+0.0%	✅
runtime.dotnet.mem.committed	36.69 ± (36.66 - 36.72) MB	36.78 ± (36.75 - 36.81) MB	+0.3%	✅⬆️
runtime.dotnet.threads.count	27 ± (27 - 27)	27 ± (27 - 27)	-0.4%	✅

HttpMessageHandler

Metric	Master (Mean ± 95% CI)	Current (Mean ± 95% CI)	Change	Status
.NET Framework 4.8 - Baseline
duration	193.95 ± (194.21 - 195.30) ms	194.04 ± (193.91 - 194.73) ms	+0.0%	✅⬆️
.NET Framework 4.8 - Bailout
duration	197.41 ± (197.38 - 197.93) ms	198.06 ± (198.03 - 198.81) ms	+0.3%	✅⬆️
.NET Framework 4.8 - CallTarget+Inlining+NGEN
duration	1137.50 ± (1138.97 - 1147.41) ms	1149.20 ± (1154.47 - 1164.46) ms	+1.0%	✅⬆️
.NET Core 3.1 - Baseline
process.internal_duration_ms	188.25 ± (187.91 - 188.59) ms	188.52 ± (188.15 - 188.89) ms	+0.1%	✅⬆️
process.time_to_main_ms	81.73 ± (81.44 - 82.02) ms	81.91 ± (81.69 - 82.13) ms	+0.2%	✅⬆️
runtime.dotnet.exceptions.count	3 ± (3 - 3)	3 ± (3 - 3)	+0.0%	✅
runtime.dotnet.mem.committed	16.15 ± (16.12 - 16.18) MB	16.09 ± (16.07 - 16.11) MB	-0.4%	✅
runtime.dotnet.threads.count	20 ± (19 - 20)	20 ± (19 - 20)	+0.1%	✅⬆️
.NET Core 3.1 - Bailout
process.internal_duration_ms	187.65 ± (187.35 - 187.95) ms	187.64 ± (187.36 - 187.93) ms	-0.0%	✅
process.time_to_main_ms	82.80 ± (82.66 - 82.94) ms	83.08 ± (82.94 - 83.23) ms	+0.3%	✅⬆️
runtime.dotnet.exceptions.count	3 ± (3 - 3)	3 ± (3 - 3)	+0.0%	✅
runtime.dotnet.mem.committed	16.20 ± (16.17 - 16.23) MB	16.13 ± (16.11 - 16.16) MB	-0.4%	✅
runtime.dotnet.threads.count	21 ± (20 - 21)	21 ± (20 - 21)	+0.1%	✅⬆️
.NET Core 3.1 - CallTarget+Inlining+NGEN
process.internal_duration_ms	431.37 ± (428.25 - 434.49) ms	435.02 ± (432.69 - 437.35) ms	+0.8%	✅⬆️
process.time_to_main_ms	480.75 ± (480.06 - 481.43) ms	483.38 ± (482.65 - 484.12) ms	+0.5%	✅⬆️
runtime.dotnet.exceptions.count	3 ± (3 - 3)	3 ± (3 - 3)	+0.0%	✅
runtime.dotnet.mem.committed	58.93 ± (58.82 - 59.05) MB	59.01 ± (58.91 - 59.11) MB	+0.1%	✅⬆️
runtime.dotnet.threads.count	29 ± (29 - 29)	29 ± (29 - 29)	+0.0%	✅⬆️
.NET 6 - Baseline
process.internal_duration_ms	192.54 ± (192.22 - 192.86) ms	193.50 ± (193.07 - 193.93) ms	+0.5%	✅⬆️
process.time_to_main_ms	70.78 ± (70.58 - 70.97) ms	71.14 ± (70.88 - 71.39) ms	+0.5%	✅⬆️
runtime.dotnet.exceptions.count	4 ± (4 - 4)	4 ± (4 - 4)	+0.0%	✅
runtime.dotnet.mem.committed	16.34 ± (16.31 - 16.37) MB	16.33 ± (16.27 - 16.40) MB	-0.0%	✅
runtime.dotnet.threads.count	19 ± (19 - 19)	19 ± (19 - 19)	+0.3%	✅⬆️
.NET 6 - Bailout
process.internal_duration_ms	191.96 ± (191.62 - 192.31) ms	192.38 ± (192.00 - 192.75) ms	+0.2%	✅⬆️
process.time_to_main_ms	71.94 ± (71.80 - 72.09) ms	72.24 ± (72.09 - 72.39) ms	+0.4%	✅⬆️
runtime.dotnet.exceptions.count	4 ± (4 - 4)	4 ± (4 - 4)	+0.0%	✅
runtime.dotnet.mem.committed	16.28 ± (16.16 - 16.40) MB	16.34 ± (16.24 - 16.43) MB	+0.4%	✅⬆️
runtime.dotnet.threads.count	20 ± (20 - 20)	20 ± (20 - 20)	+0.5%	✅⬆️
.NET 6 - CallTarget+Inlining+NGEN
process.internal_duration_ms	441.98 ± (438.85 - 445.10) ms	450.82 ± (448.32 - 453.31) ms	+2.0%	✅⬆️
process.time_to_main_ms	458.69 ± (458.22 - 459.15) ms	460.18 ± (459.66 - 460.71) ms	+0.3%	✅⬆️
runtime.dotnet.exceptions.count	4 ± (4 - 4)	4 ± (4 - 4)	+0.0%	✅
runtime.dotnet.mem.committed	59.06 ± (58.93 - 59.19) MB	59.06 ± (58.95 - 59.18) MB	+0.0%	✅⬆️
runtime.dotnet.threads.count	29 ± (29 - 29)	29 ± (29 - 29)	+0.0%	✅⬆️
.NET 8 - Baseline
process.internal_duration_ms	190.42 ± (190.04 - 190.81) ms	191.16 ± (190.76 - 191.55) ms	+0.4%	✅⬆️
process.time_to_main_ms	70.12 ± (69.92 - 70.31) ms	70.47 ± (70.27 - 70.67) ms	+0.5%	✅⬆️
runtime.dotnet.exceptions.count	4 ± (4 - 4)	4 ± (4 - 4)	+0.0%	✅
runtime.dotnet.mem.committed	11.75 ± (11.72 - 11.78) MB	11.75 ± (11.73 - 11.78) MB	+0.0%	✅⬆️
runtime.dotnet.threads.count	18 ± (18 - 18)	18 ± (18 - 18)	-0.0%	✅
.NET 8 - Bailout
process.internal_duration_ms	189.30 ± (189.05 - 189.56) ms	190.62 ± (190.26 - 190.97) ms	+0.7%	✅⬆️
process.time_to_main_ms	70.96 ± (70.86 - 71.06) ms	71.69 ± (71.52 - 71.85) ms	+1.0%	✅⬆️
runtime.dotnet.exceptions.count	4 ± (4 - 4)	4 ± (4 - 4)	+0.0%	✅
runtime.dotnet.mem.committed	11.82 ± (11.79 - 11.85) MB	11.79 ± (11.76 - 11.82) MB	-0.2%	✅
runtime.dotnet.threads.count	19 ± (19 - 19)	19 ± (19 - 19)	+0.5%	✅⬆️
.NET 8 - CallTarget+Inlining+NGEN
process.internal_duration_ms	367.13 ± (365.75 - 368.50) ms	366.63 ± (365.10 - 368.16) ms	-0.1%	✅
process.time_to_main_ms	442.90 ± (441.99 - 443.81) ms	447.10 ± (446.28 - 447.91) ms	+0.9%	✅⬆️
runtime.dotnet.exceptions.count	4 ± (4 - 4)	4 ± (4 - 4)	+0.0%	✅
runtime.dotnet.mem.committed	48.37 ± (48.33 - 48.40) MB	48.32 ± (48.29 - 48.36) MB	-0.1%	✅
runtime.dotnet.threads.count	29 ± (29 - 29)	29 ± (29 - 29)	+0.6%	✅⬆️

Comparison explanation

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 highlighted 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).

Duration charts

FakeDbCommand (.NET Framework 4.8)

gantt
    title Execution time (ms) FakeDbCommand (.NET Framework 4.8)
    dateFormat  x
    axisFormat %Q
    todayMarker off
    section Baseline
    This PR (8070) - mean (73ms)  : 71, 76
    master - mean (69ms)  : 67, 71

    section Bailout
    This PR (8070) - mean (78ms)  : crit, 76, 80
    master - mean (72ms)  : 71, 74

    section CallTarget+Inlining+NGEN
    This PR (8070) - mean (1,059ms)  : 1004, 1113
    master - mean (1,034ms)  : 962, 1106

FakeDbCommand (.NET Core 3.1)

gantt
    title Execution time (ms) FakeDbCommand (.NET Core 3.1)
    dateFormat  x
    axisFormat %Q
    todayMarker off
    section Baseline
    This PR (8070) - mean (116ms)  : 113, 120
    master - mean (106ms)  : 104, 108

    section Bailout
    This PR (8070) - mean (117ms)  : crit, 115, 119
    master - mean (107ms)  : 106, 108

    section CallTarget+Inlining+NGEN
    This PR (8070) - mean (784ms)  : 726, 841
    master - mean (741ms)  : 683, 799

FakeDbCommand (.NET 6)

gantt
    title Execution time (ms) FakeDbCommand (.NET 6)
    dateFormat  x
    axisFormat %Q
    todayMarker off
    section Baseline
    This PR (8070) - mean (103ms)  : 99, 106
    master - mean (94ms)  : 91, 96

    section Bailout
    This PR (8070) - mean (103ms)  : crit, 101, 106
    master - mean (94ms)  : 93, 95

    section CallTarget+Inlining+NGEN
    This PR (8070) - mean (747ms)  : 666, 828
    master - mean (728ms)  : 703, 753

FakeDbCommand (.NET 8)

gantt
    title Execution time (ms) FakeDbCommand (.NET 8)
    dateFormat  x
    axisFormat %Q
    todayMarker off
    section Baseline
    This PR (8070) - mean (101ms)  : 98, 104
    master - mean (92ms)  : 89, 95

    section Bailout
    This PR (8070) - mean (103ms)  : crit, 100, 106
    master - mean (93ms)  : 92, 94

    section CallTarget+Inlining+NGEN
    This PR (8070) - mean (686ms)  : crit, 662, 710
    master - mean (644ms)  : 628, 660

HttpMessageHandler (.NET Framework 4.8)

gantt
    title Execution time (ms) HttpMessageHandler (.NET Framework 4.8)
    dateFormat  x
    axisFormat %Q
    todayMarker off
    section Baseline
    This PR (8070) - mean (194ms)  : 190, 199
    master - mean (195ms)  : 189, 201

    section Bailout
    This PR (8070) - mean (198ms)  : 195, 202
    master - mean (198ms)  : 195, 200

    section CallTarget+Inlining+NGEN
    This PR (8070) - mean (1,159ms)  : 1083, 1236
    master - mean (1,143ms)  : 1082, 1204

HttpMessageHandler (.NET Core 3.1)

gantt
    title Execution time (ms) HttpMessageHandler (.NET Core 3.1)
    dateFormat  x
    axisFormat %Q
    todayMarker off
    section Baseline
    This PR (8070) - mean (279ms)  : 273, 284
    master - mean (278ms)  : 272, 284

    section Bailout
    This PR (8070) - mean (279ms)  : 276, 283
    master - mean (279ms)  : 275, 282

    section CallTarget+Inlining+NGEN
    This PR (8070) - mean (945ms)  : 893, 998
    master - mean (942ms)  : 894, 990

HttpMessageHandler (.NET 6)

gantt
    title Execution time (ms) HttpMessageHandler (.NET 6)
    dateFormat  x
    axisFormat %Q
    todayMarker off
    section Baseline
    This PR (8070) - mean (273ms)  : 266, 280
    master - mean (272ms)  : 266, 278

    section Bailout
    This PR (8070) - mean (273ms)  : 268, 277
    master - mean (272ms)  : 267, 277

    section CallTarget+Inlining+NGEN
    This PR (8070) - mean (940ms)  : 904, 976
    master - mean (928ms)  : 874, 982

HttpMessageHandler (.NET 8)

gantt
    title Execution time (ms) HttpMessageHandler (.NET 8)
    dateFormat  x
    axisFormat %Q
    todayMarker off
    section Baseline
    This PR (8070) - mean (272ms)  : 267, 277
    master - mean (270ms)  : 264, 276

    section Bailout
    This PR (8070) - mean (272ms)  : 268, 276
    master - mean (270ms)  : 267, 273

    section CallTarget+Inlining+NGEN
    This PR (8070) - mean (844ms)  : 826, 863
    master - mean (841ms)  : 825, 857

tonyredondo · 2026-01-16T10:55:17Z

tracer/src/Datadog.Trace/Logging/Internal/DatadogLogging.cs

        {
            // Tells us which types are loaded, when, and how often.
-            SharedLogger.Debug("Logger retrieved for: {AssemblyQualifiedName}", classType.AssemblyQualifiedName);
+            if (SharedLogger.IsEnabled(LogEventLevel.Debug))


RuntimeType.AssemblyQualifiedName always allocates a new string and do some internals QCalls, is fast but we can avoid it.

tonyredondo · 2026-01-16T10:56:54Z

tracer/src/Datadog.Trace/Logging/Internal/DatadogSerilogLogger.cs

+            using var array = FixedSizeArrayPool<object?>.OneItemPool.Get();
+            array.Array[0] = property;


previously we allocate an array for every single log line, here the idea is to have a small pool for it and reuse it.

tonyredondo · 2026-01-16T10:58:04Z

tracer/src/Datadog.Trace/TraceClock.cs

    {
        [MethodImpl(MethodImplOptions.AggressiveInlining)]
-        get => _utcStart.Add(Elapsed);
+        get => _utcStart.AddTicks(StopwatchHelpers.GetElapsedTicks(Stopwatch.GetTimestamp() - _timestamp));


we really don't need to create a TimeSpan struct for this by calling Elapsed, we just count the ticks and add them.

is this measurably faster?

not in recent frameworks, in olders frameworks the codegen is just a little better.

tonyredondo · 2026-01-16T11:00:16Z

tracer/src/Datadog.Trace/Util/CodeDuration.cs

+
+namespace Datadog.Trace.Util;
+
+internal struct CodeDuration : IDisposable


This is the version for use in an async/await method because we cannot use ref struct there. So creating a CodeDuration in that case will allocate in the heap (struct boxing).

Below there's the ref struct version to avoid any heap allocation in a normal sync method.

Would it make sense to add a short comment in code about when to use CodeDuration vs CodeDurationRef?
My concern is that if someone uses the struct version in a non async methos and in a non-local way (copied by value) he could end up calling Dispose() on multiple copies and get an incorrect CodeDurationBase.Count.

done in 5cb68e6

tonyredondo · 2026-01-16T11:01:24Z

tracer/src/Datadog.Trace/Util/CodeDuration.cs

+    }
+}
+
+internal ref struct CodeDurationRef


ref struct version to avoid an allocation in the heap.

NIT: wondering if the code inside the methods / ctors could have been shared between the two via some static helpers , for maintenance

tonyredondo · 2026-01-16T11:04:51Z

tracer/src/Datadog.Trace/Util/FixedSizeArrayPool.cs

+#endif
+
+#if NETCOREAPP3_0_OR_GREATER
+        if (RuntimeHelpers.IsReferenceOrContainsReferences<T>())


we clear the array if we know that T is a reference type (so we don't hold any reference from the GC)

tracer/src/Datadog.Trace/Util/RefStopwatch.cs

tracer/src/Datadog.Trace/Util/FixedSizeArrayPool.cs

dudikeleti

Great work

dudikeleti · 2026-01-20T18:21:07Z

tracer/src/Datadog.Trace/Util/CodeDuration.cs

+
+namespace Datadog.Trace.Util;
+
+internal struct CodeDuration : IDisposable


Would it make sense to add a short comment in code about when to use CodeDuration vs CodeDurationRef?
My concern is that if someone uses the struct version in a non async methos and in a non-local way (copied by value) he could end up calling Dispose() on multiple copies and get an incorrect CodeDurationBase.Count.

anna-git · 2026-01-29T15:53:40Z

tracer/src/Datadog.Trace/Util/CodeDuration.cs

+    }
+}
+
+internal ref struct CodeDurationRef


NIT: wondering if the code inside the methods / ctors could have been shared between the two via some static helpers , for maintenance

## Summary of changes - Introduce Test Optimization RunId and EnsureRunId with cache folder lifecycle. - Add `_DD_INTERNAL_TOPT_RUNID` to config mapping/docs and regenerate keys. - Update CI integration tests to set/print RunId for deterministic runs. - Adjust test helper thread naming for clearer CI debugging. ## Reason for change Provide a stable RunId before caching work that depends on it and keep configuration in sync. ## Implementation details - Add RunId/EnsureRunId to `ITestOptimization`/`TestOptimization` with environment propagation. - Update configuration mapping and generated CI visibility keys. - Update CI integration tests for RunId coverage. ## Test coverage CI passes then all changes are good. ## Other details - Stacked PRs (current marked): - PR1 #8070 - **CURRENT**: PR2 #8071 - PR3 #8072 - PR4 #8073

## Summary of changes - Add git command caching and telemetry improvements for CI visibility. - Update GitInfo discovery and CI environment logging. - Update impacted tests for new git discovery behavior. JIRA: SDTEST-3226 ## Reason for change Reduce git command overhead and improve visibility into CI git metadata collection. ## Implementation details - Disk cache keyed by RunId in GitCommandHelper with safe.directory handling. - JSON annotations in ProcessHelpers for cache serialization. - Adjusted CI runner git discovery logic and tests. ## Test coverage CI passes then all changes are good. ## Other details - Stacked PRs (current marked): - PR1 #8070 - PR2 #8071 - **CURRENT**: PR3 #8072 - PR4 #8073

## Summary of changes - Add cached/file Test Optimization clients and update client wrappers. - Reduce feature init overhead and adjust background initialization. - Improve runner logging and CI workspace cache usage. ## Reason for change Reduce repeated HTTP calls and improve initialization performance for CI visibility features. ## Implementation details - New cached/file client implementations for disk/memory caching. - Feature creation updated to reduce dependencies on client calls. - CodeDuration instrumentation added to CI initialization paths. ## Test coverage CI passes then all changes are good. ## Other details - Stacked PRs (current marked): - PR1 #8070 - PR2 #8071 - PR3 #8072 - **CURRENT**: PR4 #8073

tonyredondo changed the title ~~perf: logging/timing helpers~~ [Test Optimization] perf: logging/timing helpers Jan 15, 2026

This was referenced Jan 15, 2026

[Test Optimization] ci: add run id + config keys #8071

Merged

[Test Optimization] ci: git caching + env improvements #8072

Merged

[Test Optimization] ci: client caching + feature tweaks #8073

Merged

github-actions bot added the area:ci-visibility label Jan 15, 2026

tonyredondo force-pushed the tony/testoptimization-performance-work-pr1 branch from ad6c32d to f0ccdd2 Compare January 15, 2026 22:06

This comment has been minimized.

Sign in to view

tonyredondo commented Jan 16, 2026

View reviewed changes

tracer/src/Datadog.Trace/Util/RefStopwatch.cs Show resolved Hide resolved

tonyredondo force-pushed the tony/testoptimization-performance-work-pr1 branch from 7a2fadb to c58aa40 Compare January 16, 2026 12:39

tonyredondo marked this pull request as ready for review January 16, 2026 12:40

tonyredondo requested review from a team as code owners January 16, 2026 12:40

anmarchenko approved these changes Jan 16, 2026

View reviewed changes

lucaspimentel reviewed Jan 16, 2026

View reviewed changes

tracer/src/Datadog.Trace/Util/FixedSizeArrayPool.cs Outdated Show resolved Hide resolved

lucaspimentel reviewed Jan 16, 2026

View reviewed changes

tracer/src/Datadog.Trace/Util/FixedSizeArrayPool.cs Outdated Show resolved Hide resolved

lucaspimentel reviewed Jan 16, 2026

View reviewed changes

tracer/src/Datadog.Trace/Util/FixedSizeArrayPool.cs Show resolved Hide resolved

dudikeleti approved these changes Jan 20, 2026

View reviewed changes

tonyredondo added 2 commits January 23, 2026 14:55

perf: logging/timing helpers

8ebc94e

improve the fixed size array pool

469f3d0

tonyredondo added 3 commits January 23, 2026 14:55

refactor improvement

58ccbef

changes from review

c64a617

Add comment in the RefStopwatch

3f3f143

tonyredondo force-pushed the tony/testoptimization-performance-work-pr1 branch from 4a649b1 to 3f3f143 Compare January 23, 2026 13:55

docs: clarify CodeDuration usage

5cb68e6

bouwkast approved these changes Jan 28, 2026

View reviewed changes

anna-git approved these changes Jan 29, 2026

View reviewed changes

Merge branch 'master' into tony/testoptimization-performance-work-pr1

90b2923

tonyredondo merged commit 30a543a into master Jan 30, 2026
145 checks passed

tonyredondo deleted the tony/testoptimization-performance-work-pr1 branch January 30, 2026 15:33

github-actions bot added this to the vNext-v3 milestone Jan 30, 2026

		using var array = FixedSizeArrayPool<object?>.OneItemPool.Get();
		array.Array[0] = property;


		namespace Datadog.Trace.Util;

		internal struct CodeDuration : IDisposable

Conversation

tonyredondo commented Jan 15, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary of changes

Reason for change

Implementation details

Test coverage

Other details

Uh oh!

pr-commenter bot commented Jan 15, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Benchmarks

scenario:Benchmarks.Trace.ActivityBenchmark.StartStopWithChild net472

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

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

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

scenario:Benchmarks.Trace.CIVisibilityProtocolWriterBenchmark.WriteAndFlushEnrichedTraces net472

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

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

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

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

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

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

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

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

scenario:Benchmarks.Trace.TraceAnnotationsBenchmark.RunOnMethodBegin net472

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

Uh oh!

dd-trace-dotnet-ci-bot bot commented Jan 15, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Execution-Time Benchmarks Report ⏱️

⚠️ Potential regressions detected

FakeDbCommand

FakeDbCommand

HttpMessageHandler

Uh oh!

This comment has been minimized.

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

tonyredondo Jan 16, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

dudikeleti Jan 20, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

dudikeleti left a comment

Choose a reason for hiding this comment

Uh oh!

dudikeleti Jan 20, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

tonyredondo commented Jan 15, 2026 •

edited

Loading

pr-commenter bot commented Jan 15, 2026 •

edited

Loading

dd-trace-dotnet-ci-bot bot commented Jan 15, 2026 •

edited

Loading

tonyredondo Jan 16, 2026 •

edited

Loading

dudikeleti Jan 20, 2026 •

edited

Loading

dudikeleti Jan 20, 2026 •

edited

Loading