baggage propagator clean up #6266

lucaspimentel · 2024-11-08T23:19:15Z

Summary of changes

Refactored how we generated the baggage header value in W3CBaggagePropagator to be faster and minimize heap allocations.

Reason for change

Performance

Implementation details

1. `string` and `StringBuilder` management

We now use a single StringBuilder (from StringBuilderCache) to build the entire baggage header. Previously the code was trying to acquire a builder to encode keys and values while already holding another builder for the header itself. With this change, we no longer allocate multiple StringBuilder instances that don't get reused, and we no longer allocation individual strings for:

keys that need encoding
values that need encoding
every key/value pair (the string was built first to measure its length in UTF-8 bytes, but we don't need that anymore)

2. checking earlier if string needs encoding

Before this PR, we always converted every key and value into UTF-8 bytes before checking if any bytes needed percent-encoding (aka url-encoding). When encoding was not needed (the common case), we could avoid the conversion and use the original string directly.

Now we check if the string's characters need encoding before converting them to bytes. When encoding is not needed (the common case), we skip the conversion from string to UTF-8 bytes.

Benchmarks

With allowed characters only (no percent-encoding required)

Method	Runtime	ItemCount	Mean	Error	StdDev	Ratio	Gen0	Allocated	Alloc Ratio
Before	.NET 6.0	1	91.63 ns	0.309 ns	0.289 ns	1.00	0.0050	64 B	1.00
After	.NET 6.0	1	49.84 ns	0.162 ns	0.152 ns	0.54	0.0025	32 B	0.50

Before	.NET Framework 4.6.1	1	253.30 ns	1.867 ns	1.655 ns	1.00	0.0215	136 B	1.00
After	.NET Framework 4.6.1	1	101.09 ns	0.368 ns	0.326 ns	0.40	0.0063	40 B	0.29

Before	.NET 6.0	5	343.57 ns	1.953 ns	1.731 ns	1.00	0.0215	272 B	1.00
After	.NET 6.0	5	139.36 ns	2.696 ns	3.104 ns	0.41	0.0062	80 B	0.29

Before	.NET Framework 4.6.1	5	1,303.19 ns	6.138 ns	5.742 ns	1.00	0.1602	1019 B	1.00
After	.NET Framework 4.6.1	5	328.44 ns	0.956 ns	0.848 ns	0.25	0.0138	88 B	0.09

With characters that require percent-encoding (whitespace and Unicode)

Method	Runtime	ItemCount	Mean	Error	StdDev	Ratio	Gen0	Allocated	Alloc Ratio
Before	.NET 6.0	1	229.4 ns	2.07 ns	1.73 ns	1.00	0.0815	1024 B	1.00
After	.NET 6.0	1	105.5 ns	0.76 ns	0.71 ns	0.46	0.0063	80 B	0.08

Before	.NET Framework 4.6.1	1	990.2 ns	2.04 ns	1.60 ns	1.00	0.4215	2656 B	1.00
After	.NET Framework 4.6.1	1	279.9 ns	1.46 ns	1.37 ns	0.28	0.0124	80 B	0.03

Before	.NET 6.0	5	989.9 ns	6.25 ns	5.22 ns	1.00	0.1469	1848 B	1.00
After	.NET 6.0	5	438.5 ns	4.18 ns	3.91 ns	0.44	0.0234	296 B	0.16

Before	.NET Framework 4.6.1	5	4,989.7 ns	50.79 ns	45.02 ns	1.00	1.6556	10447 B	1.00
After	.NET Framework 4.6.1	5	1,243.1 ns	7.70 ns	7.20 ns	0.25	0.0458	297 B	0.03

Test coverage

All W3CBaggagePropagatorTests are passing.

Other details

Follow-up from #6158.

datadog-ddstaging · 2024-11-08T23:41:32Z

Datadog Report

Branch report: lpimentel/baggage-cleanup
Commit report: 7932f45
Test service: dd-trace-dotnet

❌ 1 Failed (1 Known Flaky), 367888 Passed, 2082 Skipped, 14h 10m 26.9s Total Time

❌ Failed Tests (1)

WorkingWithContinuousProfiler - Datadog.Trace.Tools.dd_dotnet.ArtifactTests.Checks.ProcessBasicChecksTests - ❄️ Known flaky - Details

Expand for error

 
 
 With StandardOutput:
 Running checks on process 5192 Process name: Samples.Console  ---- STARTING TRACER SETUP CHECKS ----- Target process is running with .NET Framework 1. Checking Modules Needed so the Tracer Loads:  [SUCCESS]: The native library version 3.6.0.0 is loaded into the process.  [WARNING]: Found multiple instances of Datadog.Trace.dll in the target process. Detected versions: 3.6.0.0  2. Checking DD_DOTNET_TRACER_HOME and related configuration value:  [SUCCESS]: DD_DOTNET_TRACER_HOME is set to 'D:\a\1\s\shared\bin\monitoring-home' and the directory was found correctly. 3. Checking COR_PROFILER_PATH and related configuration value:  [SUCCESS]: The environment variable COR_PROFILER_PATH is set to the correct value of D:\a\1\s\shared\bin\monitoring-home\win-x64\Datadog.Trace.ClrProfiler.Native.dll . 4. Checking COR_PROFILER and related configuration value:  [SUCCESS]: The environment variable COR_PROFILER is set to the correct value of {846F5F1C-F9AE-4B07-969E-05C26BC060D8}. 5. Checking COR_ENABLE_PROFILING and related configuration value:  [SUCCESS]: The environment variable COR_ENABLE_PROFILING is set to the correct value of 1.  ---- CONFIGURATION CHECKS ----- 1. Checking if tracing is disabled using DD_TRACE_ENABLED.  [INFO]: DD_TRACE_ENABLED is not set, the default value is true. 2. Checking if profiling is enabled using DD_PROFILING_ENABLED.  [SUCCESS]: DD_PROFILING_ENABLED is set.  ---- DATADOG AGENT CHECKS ----- Detected agent url: http://127.0.0.1:53768/. Note: this url may be incorrect if you configured the application through a configuration file. Connecting to Agent at endpoint http://127.0.0.1:53768/ using HTTP  [WARNING]: Could not detect the agent version. It may be running with a version older than 7.27.0.  [SUCCESS]: No issue found with the target process. 
 With ErrorOutput:
 
 With ExitCode:
 0acer version 3.6.0.0 is loaded into the process."}.

andrewlock · 2024-11-08T23:50:43Z

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 (6266) - mean (70ms)  : 67, 72
     .   : milestone, 70,
    master - mean (70ms)  : 68, 72
     .   : milestone, 70,

    section CallTarget+Inlining+NGEN
    This PR (6266) - mean (1,114ms)  : 1094, 1134
     .   : milestone, 1114,
    master - mean (1,114ms)  : 1097, 1130
     .   : milestone, 1114,

gantt
    title Execution time (ms) FakeDbCommand (.NET Core 3.1) 
    dateFormat  X
    axisFormat %s
    todayMarker off
    section Baseline
    This PR (6266) - mean (108ms)  : 106, 110
     .   : milestone, 108,
    master - mean (108ms)  : 105, 110
     .   : milestone, 108,

    section CallTarget+Inlining+NGEN
    This PR (6266) - mean (771ms)  : 758, 783
     .   : milestone, 771,
    master - mean (775ms)  : 762, 788
     .   : milestone, 775,

gantt
    title Execution time (ms) FakeDbCommand (.NET 6) 
    dateFormat  X
    axisFormat %s
    todayMarker off
    section Baseline
    This PR (6266) - mean (92ms)  : 90, 94
     .   : milestone, 92,
    master - mean (92ms)  : 90, 94
     .   : milestone, 92,

    section CallTarget+Inlining+NGEN
    This PR (6266) - mean (731ms)  : 717, 744
     .   : milestone, 731,
    master - mean (732ms)  : 712, 751
     .   : milestone, 732,

gantt
    title Execution time (ms) HttpMessageHandler (.NET Framework 4.6.2) 
    dateFormat  X
    axisFormat %s
    todayMarker off
    section Baseline
    This PR (6266) - mean (190ms)  : 187, 193
     .   : milestone, 190,
    master - mean (190ms)  : 186, 193
     .   : milestone, 190,

    section CallTarget+Inlining+NGEN
    This PR (6266) - mean (1,229ms)  : 1206, 1253
     .   : milestone, 1229,
    master - mean (1,229ms)  : 1202, 1256
     .   : milestone, 1229,

gantt
    title Execution time (ms) HttpMessageHandler (.NET Core 3.1) 
    dateFormat  X
    axisFormat %s
    todayMarker off
    section Baseline
    This PR (6266) - mean (276ms)  : 270, 281
     .   : milestone, 276,
    master - mean (277ms)  : 272, 282
     .   : milestone, 277,

    section CallTarget+Inlining+NGEN
    This PR (6266) - mean (948ms)  : 932, 964
     .   : milestone, 948,
    master - mean (954ms)  : 937, 970
     .   : milestone, 954,

gantt
    title Execution time (ms) HttpMessageHandler (.NET 6) 
    dateFormat  X
    axisFormat %s
    todayMarker off
    section Baseline
    This PR (6266) - mean (266ms)  : 262, 270
     .   : milestone, 266,
    master - mean (263ms)  : 259, 268
     .   : milestone, 263,

    section CallTarget+Inlining+NGEN
    This PR (6266) - mean (939ms)  : 918, 959
     .   : milestone, 939,
    master - mean (934ms)  : 914, 954
     .   : milestone, 934,

andrewlock · 2024-11-09T00:47:17Z

Benchmarks Report for tracer 🐌

Benchmarks for #6266 compared to master:

All benchmarks have the same speed
1 benchmarks have fewer 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	8.05μs	43.6ns	235ns	0.0158	0.00792	0	5.61 KB
master	`StartStopWithChild`	netcoreapp3.1	10.1μs	54.9ns	296ns	0.021	0.0105	0	5.8 KB
master	`StartStopWithChild`	net472	16.4μs	43.9ns	170ns	1.03	0.303	0.0901	6.22 KB
#6266	`StartStopWithChild`	net6.0	8.05μs	45.7ns	317ns	0.0157	0.00784	0	5.61 KB
#6266	`StartStopWithChild`	netcoreapp3.1	10.2μs	57.1ns	365ns	0.0147	0.00492	0	5.8 KB
#6266	`StartStopWithChild`	net472	16.6μs	59.8ns	232ns	1.03	0.309	0.0917	6.21 KB

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

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Allocated
master	`WriteAndFlushEnrichedTraces`	net6.0	486μs	222ns	799ns	0	2.7 KB
master	`WriteAndFlushEnrichedTraces`	netcoreapp3.1	646μs	247ns	955ns	0	2.7 KB
master	`WriteAndFlushEnrichedTraces`	net472	843μs	580ns	2.25μs	0.419	3.3 KB
#6266	`WriteAndFlushEnrichedTraces`	net6.0	468μs	306ns	1.18μs	0	2.7 KB
#6266	`WriteAndFlushEnrichedTraces`	netcoreapp3.1	634μs	259ns	1μs	0	2.7 KB
#6266	`WriteAndFlushEnrichedTraces`	net472	843μs	748ns	2.8μs	0.419	3.3 KB

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

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Allocated
master	`SendRequest`	net6.0	198μs	1.16μs	10.9μs	0.2	18.73 KB
master	`SendRequest`	netcoreapp3.1	229μs	1.42μs	14μs	0.206	20.89 KB
master	`SendRequest`	net472	8.3E‑05ns	6.89E‑05ns	0.000248ns	0	0 b
#6266	`SendRequest`	net6.0	198μs	1.14μs	9.65μs	0.193	18.73 KB
#6266	`SendRequest`	netcoreapp3.1	216μs	1.17μs	6.73μs	0.213	20.89 KB
#6266	`SendRequest`	net472	0.00241ns	0.00045ns	0.00162ns	0	0 b

Benchmarks.Trace.CIVisibilityProtocolWriterBenchmark - Same speed ✔️ Fewer allocations 🎉

Fewer allocations 🎉 in #6266

Benchmark	Base Allocated	Diff Allocated	Change	Change %
Benchmarks.Trace.CIVisibilityProtocolWriterBenchmark.WriteAndFlushEnrichedTraces‑net6.0	41.82 KB	41.46 KB	-362 B	-0.87%

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Gen 1	Gen 2	Allocated
master	`WriteAndFlushEnrichedTraces`	net6.0	602μs	3.43μs	30.9μs	0.558	0	0	41.82 KB
master	`WriteAndFlushEnrichedTraces`	netcoreapp3.1	676μs	3.26μs	12.6μs	0.349	0	0	41.78 KB
master	`WriteAndFlushEnrichedTraces`	net472	867μs	3.74μs	14.5μs	8.25	2.6	0.434	53.27 KB
#6266	`WriteAndFlushEnrichedTraces`	net6.0	554μs	2.45μs	9.17μs	0.553	0	0	41.46 KB
#6266	`WriteAndFlushEnrichedTraces`	netcoreapp3.1	670μs	2.04μs	7.9μs	0.338	0	0	41.83 KB
#6266	`WriteAndFlushEnrichedTraces`	net472	888μs	2.44μs	9.11μs	8.5	2.55	0.425	53.32 KB

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

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Gen 1	Allocated
master	`ExecuteNonQuery`	net6.0	1.23μs	1.52ns	5.87ns	0.0145	0	1.02 KB
master	`ExecuteNonQuery`	netcoreapp3.1	1.77μs	0.875ns	3.27ns	0.0133	0	1.02 KB
master	`ExecuteNonQuery`	net472	2.11μs	1.33ns	4.97ns	0.156	0.00106	987 B
#6266	`ExecuteNonQuery`	net6.0	1.3μs	1.04ns	4.04ns	0.0141	0	1.02 KB
#6266	`ExecuteNonQuery`	netcoreapp3.1	1.78μs	1.55ns	5.99ns	0.0141	0	1.02 KB
#6266	`ExecuteNonQuery`	net472	2.09μs	1.76ns	6.34ns	0.157	0.00104	987 B

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

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Allocated
master	`CallElasticsearch`	net6.0	1.23μs	0.403ns	1.51ns	0.0136	976 B
master	`CallElasticsearch`	netcoreapp3.1	1.55μs	0.864ns	3.35ns	0.0131	976 B
master	`CallElasticsearch`	net472	2.48μs	2.6ns	10.1ns	0.157	995 B
master	`CallElasticsearchAsync`	net6.0	1.24μs	0.543ns	2.1ns	0.013	952 B
master	`CallElasticsearchAsync`	netcoreapp3.1	1.6μs	0.683ns	2.55ns	0.0137	1.02 KB
master	`CallElasticsearchAsync`	net472	2.61μs	2.81ns	10.9ns	0.167	1.05 KB
#6266	`CallElasticsearch`	net6.0	1.16μs	0.916ns	3.3ns	0.0134	976 B
#6266	`CallElasticsearch`	netcoreapp3.1	1.52μs	0.423ns	1.53ns	0.0128	976 B
#6266	`CallElasticsearch`	net472	2.5μs	0.997ns	3.86ns	0.158	995 B
#6266	`CallElasticsearchAsync`	net6.0	1.32μs	0.707ns	2.74ns	0.0132	952 B
#6266	`CallElasticsearchAsync`	netcoreapp3.1	1.61μs	0.724ns	2.8ns	0.0145	1.02 KB
#6266	`CallElasticsearchAsync`	net472	2.52μs	0.978ns	3.79ns	0.167	1.05 KB

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

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Allocated
master	`ExecuteAsync`	net6.0	1.29μs	0.685ns	2.56ns	0.0135	952 B
master	`ExecuteAsync`	netcoreapp3.1	1.67μs	1.73ns	6.47ns	0.0125	952 B
master	`ExecuteAsync`	net472	1.84μs	0.699ns	2.61ns	0.145	915 B
#6266	`ExecuteAsync`	net6.0	1.29μs	0.575ns	2.15ns	0.013	952 B
#6266	`ExecuteAsync`	netcoreapp3.1	1.68μs	1.04ns	4.02ns	0.0125	952 B
#6266	`ExecuteAsync`	net472	1.7μs	0.323ns	1.21ns	0.145	915 B

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

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Allocated
master	`SendAsync`	net6.0	4.46μs	1.22ns	4.57ns	0.0313	2.31 KB
master	`SendAsync`	netcoreapp3.1	5.2μs	1.63ns	6.3ns	0.039	2.85 KB
master	`SendAsync`	net472	7.52μs	2.64ns	10.2ns	0.495	3.12 KB
#6266	`SendAsync`	net6.0	4.45μs	3.51ns	13.1ns	0.0309	2.31 KB
#6266	`SendAsync`	netcoreapp3.1	5.1μs	2.27ns	8.2ns	0.0383	2.85 KB
#6266	`SendAsync`	net472	7.35μs	1.5ns	5.4ns	0.493	3.12 KB

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

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Allocated
master	`EnrichedLog`	net6.0	1.48μs	1ns	3.76ns	0.0231	1.64 KB
master	`EnrichedLog`	netcoreapp3.1	2.32μs	1.88ns	7.05ns	0.022	1.64 KB
master	`EnrichedLog`	net472	2.64μs	2ns	7.74ns	0.249	1.57 KB
#6266	`EnrichedLog`	net6.0	1.49μs	1.5ns	5.62ns	0.0227	1.64 KB
#6266	`EnrichedLog`	netcoreapp3.1	2.11μs	1.9ns	7.34ns	0.022	1.64 KB
#6266	`EnrichedLog`	net472	2.63μs	0.82ns	2.96ns	0.249	1.57 KB

Benchmarks.Trace.Log4netBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Gen 1	Allocated
master	`EnrichedLog`	net6.0	119μs	95.3ns	369ns	0.0594	0	4.28 KB
master	`EnrichedLog`	netcoreapp3.1	122μs	175ns	632ns	0	0	4.28 KB
master	`EnrichedLog`	net472	154μs	177ns	685ns	0.688	0.229	4.46 KB
#6266	`EnrichedLog`	net6.0	120μs	295ns	1.14μs	0.0601	0	4.28 KB
#6266	`EnrichedLog`	netcoreapp3.1	121μs	219ns	850ns	0	0	4.28 KB
#6266	`EnrichedLog`	net472	154μs	220ns	853ns	0.688	0.229	4.46 KB

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

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Allocated
master	`EnrichedLog`	net6.0	3.04μs	13.8ns	55.2ns	0.0302	2.2 KB
master	`EnrichedLog`	netcoreapp3.1	4.23μs	3.92ns	15.2ns	0.0295	2.2 KB
master	`EnrichedLog`	net472	4.91μs	1.62ns	6.27ns	0.32	2.02 KB
#6266	`EnrichedLog`	net6.0	2.88μs	0.827ns	3.09ns	0.0301	2.2 KB
#6266	`EnrichedLog`	netcoreapp3.1	4.14μs	3.48ns	13.5ns	0.0289	2.2 KB
#6266	`EnrichedLog`	net472	4.9μs	0.943ns	3.53ns	0.32	2.02 KB

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

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Allocated
master	`SendReceive`	net6.0	1.45μs	0.771ns	2.99ns	0.0159	1.14 KB
master	`SendReceive`	netcoreapp3.1	1.78μs	0.775ns	3ns	0.0151	1.14 KB
master	`SendReceive`	net472	2.17μs	4.16ns	16.1ns	0.183	1.16 KB
#6266	`SendReceive`	net6.0	1.36μs	1.05ns	4.06ns	0.0163	1.14 KB
#6266	`SendReceive`	netcoreapp3.1	1.81μs	1.62ns	6.29ns	0.0153	1.14 KB
#6266	`SendReceive`	net472	2.18μs	2.62ns	10.2ns	0.183	1.16 KB

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

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Allocated
master	`EnrichedLog`	net6.0	2.8μs	1.58ns	5.9ns	0.0224	1.6 KB
master	`EnrichedLog`	netcoreapp3.1	3.9μs	2.96ns	11.5ns	0.0214	1.65 KB
master	`EnrichedLog`	net472	4.39μs	2.57ns	9.96ns	0.323	2.04 KB
#6266	`EnrichedLog`	net6.0	2.73μs	0.846ns	3.28ns	0.0221	1.6 KB
#6266	`EnrichedLog`	netcoreapp3.1	3.76μs	1.81ns	7.01ns	0.0207	1.65 KB
#6266	`EnrichedLog`	net472	4.39μs	1.31ns	4.89ns	0.323	2.04 KB

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

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Allocated
master	`StartFinishSpan`	net6.0	406ns	0.284ns	1.1ns	0.00803	576 B
master	`StartFinishSpan`	netcoreapp3.1	640ns	0.846ns	3.28ns	0.00779	576 B
master	`StartFinishSpan`	net472	733ns	0.76ns	2.63ns	0.0916	578 B
master	`StartFinishScope`	net6.0	492ns	0.575ns	2.07ns	0.00976	696 B
master	`StartFinishScope`	netcoreapp3.1	736ns	0.596ns	2.31ns	0.00911	696 B
master	`StartFinishScope`	net472	873ns	1.52ns	5.48ns	0.104	658 B
#6266	`StartFinishSpan`	net6.0	400ns	0.141ns	0.509ns	0.00801	576 B
#6266	`StartFinishSpan`	netcoreapp3.1	585ns	0.628ns	2.43ns	0.00755	576 B
#6266	`StartFinishSpan`	net472	712ns	0.204ns	0.79ns	0.0916	578 B
#6266	`StartFinishScope`	net6.0	488ns	0.231ns	0.894ns	0.00985	696 B
#6266	`StartFinishScope`	netcoreapp3.1	753ns	1.42ns	5.33ns	0.00939	696 B
#6266	`StartFinishScope`	net472	907ns	0.355ns	1.38ns	0.104	658 B

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

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Allocated
master	`RunOnMethodBegin`	net6.0	600ns	0.444ns	1.72ns	0.00965	696 B
master	`RunOnMethodBegin`	netcoreapp3.1	954ns	1.8ns	6.98ns	0.00911	696 B
master	`RunOnMethodBegin`	net472	1.12μs	0.714ns	2.76ns	0.104	658 B
#6266	`RunOnMethodBegin`	net6.0	596ns	0.312ns	1.21ns	0.00987	696 B
#6266	`RunOnMethodBegin`	netcoreapp3.1	932ns	0.283ns	1.06ns	0.00943	696 B
#6266	`RunOnMethodBegin`	net472	1.16μs	0.429ns	1.66ns	0.104	658 B

andrewlock · 2024-11-09T01:48:27Z

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 (6266) (10.976M)   : 0, 10976437
    master (11.114M)   : 0, 11114066
    benchmarks/2.9.0 (11.033M)   : 0, 11032866

    section Automatic
    This PR (6266) (7.092M)   : 0, 7091776
    master (7.233M)   : 0, 7232575
    benchmarks/2.9.0 (7.786M)   : 0, 7785853

    section Trace stats
    master (7.493M)   : 0, 7493020

    section Manual
    master (10.881M)   : 0, 10881078

    section Manual + Automatic
    This PR (6266) (6.491M)   : 0, 6491000
    master (6.542M)   : 0, 6542224

    section DD_TRACE_ENABLED=0
    master (10.339M)   : 0, 10339304

gantt
    title Throughput Linux arm64 (Total requests) 
    dateFormat  X
    axisFormat %s
    section Baseline
    This PR (6266) (9.505M)   : 0, 9505238
    master (9.350M)   : 0, 9349696
    benchmarks/2.9.0 (9.495M)   : 0, 9494821

    section Automatic
    This PR (6266) (6.364M)   : 0, 6363801
    master (6.299M)   : 0, 6299169

    section Trace stats
    master (6.519M)   : 0, 6519254

    section Manual
    master (9.686M)   : 0, 9685912

    section Manual + Automatic
    This PR (6266) (5.964M)   : 0, 5964289
    master (6.024M)   : 0, 6023599

    section DD_TRACE_ENABLED=0
    master (8.947M)   : 0, 8946801

gantt
    title Throughput Windows x64 (Total requests) 
    dateFormat  X
    axisFormat %s
    section Baseline
    This PR (6266) (9.873M)   : 0, 9872584
    master (9.845M)   : 0, 9844694
    benchmarks/2.9.0 (10.020M)   : 0, 10019592

    section Automatic
    This PR (6266) (6.402M)   : 0, 6402153
    master (6.255M)   : 0, 6255151
    benchmarks/2.9.0 (7.255M)   : 0, 7255257

    section Trace stats
    master (6.988M)   : 0, 6988095

    section Manual
    master (10.029M)   : 0, 10029454

    section Manual + Automatic
    This PR (6266) (5.904M)   : 0, 5903947
    master (5.901M)   : 0, 5901061

    section DD_TRACE_ENABLED=0
    master (9.102M)   : 0, 9101760

andrewlock

Nice results! Didn't entirely follow the changes tbh, but this is a case of "if you have tests for it all, I'm happy" 😄

tracer/src/Datadog.Trace/Propagators/W3CBaggagePropagator.cs

tracer/test/Datadog.Trace.Tests/Propagators/W3CBaggagePropagatorTests.cs

Co-authored-by: Andrew Lock <[email protected]>

…lable

lucaspimentel marked this pull request as ready for review November 11, 2024 17:45

lucaspimentel requested a review from a team as a code owner November 11, 2024 17:45

lucaspimentel added the type:performance Performance, speed, latency, resource usage (CPU, memory) label Nov 11, 2024

lucaspimentel requested a review from a team November 11, 2024 23:07

andrewlock approved these changes Nov 12, 2024

View reviewed changes

tracer/src/Datadog.Trace/Propagators/W3CBaggagePropagator.cs Outdated Show resolved Hide resolved

tracer/test/Datadog.Trace.Tests/Propagators/W3CBaggagePropagatorTests.cs Show resolved Hide resolved

zacharycmontoya approved these changes Nov 12, 2024

View reviewed changes

lucaspimentel force-pushed the lpimentel/baggage-cleanup branch from 466606c to 6a5004f Compare November 12, 2024 21:19

lucaspimentel and others added 13 commits November 13, 2024 11:32

refactor to use a single StringBuilder through out

26f10d2

check if any chars need encoding before converting to bytes

5a9dd53

use TryFormat on a stack buffer

70e68dd

Co-authored-by: Andrew Lock <[email protected]>

add StringBuilderExtensions.Append(ReadOnlySpan<char>) where not avai…

65b4beb

…lable

more optimizations: use HexString and stackalloc when formatting

12a7f28

clean up usings

ef9a165

add null annotation

c4cce29

rename test

860e5c6

change parameters type to ReadOnlySpan<byte>

89dd7bf

add local variable

3c16435

refactor out a CharRequiresEncoding() to reduce duplication

325479a

remove HashSets, split EncodeStringAndAppend

3b01d1a

fix tests, and add cases for all ASCII chars that require encoding

3286c90

lucaspimentel force-pushed the lpimentel/baggage-cleanup branch from 7932f45 to 3286c90 Compare November 13, 2024 16:32

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

baggage propagator clean up #6266

baggage propagator clean up #6266

lucaspimentel commented Nov 8, 2024 •

edited

Loading

datadog-ddstaging bot commented Nov 8, 2024 •

edited

Loading

andrewlock commented Nov 8, 2024 •

edited

Loading

andrewlock commented Nov 9, 2024 •

edited

Loading

Raw results

Raw results

Raw results

Fewer allocations 🎉 in #6266

Raw results

Raw results

Raw results

Raw results

Raw results

Raw results

Raw results

Raw results

Raw results

Raw results

Raw results

Raw results

andrewlock commented Nov 9, 2024 •

edited

Loading

andrewlock left a comment

baggage propagator clean up #6266

Are you sure you want to change the base?

baggage propagator clean up #6266

Conversation

lucaspimentel commented Nov 8, 2024 • edited Loading

Summary of changes

Reason for change

Implementation details

1. string and StringBuilder management

2. checking earlier if string needs encoding

Benchmarks

Test coverage

Other details

datadog-ddstaging bot commented Nov 8, 2024 • edited Loading

Datadog Report

❌ Failed Tests (1)

andrewlock commented Nov 8, 2024 • edited Loading

Execution-Time Benchmarks Report ⏱️

andrewlock commented Nov 9, 2024 • edited Loading

Benchmarks Report for tracer 🐌

Benchmark details

Raw results

Raw results

Raw results

Fewer allocations 🎉 in #6266

Raw results

Raw results

Raw results

Raw results

Raw results

Raw results

Raw results

Raw results

Raw results

Raw results

Raw results

Raw results

andrewlock commented Nov 9, 2024 • edited Loading

Throughput/Crank Report ⚡

andrewlock left a comment

Choose a reason for hiding this comment

lucaspimentel commented Nov 8, 2024 •

edited

Loading

1. `string` and `StringBuilder` management

datadog-ddstaging bot commented Nov 8, 2024 •

edited

Loading

andrewlock commented Nov 8, 2024 •

edited

Loading

andrewlock commented Nov 9, 2024 •

edited

Loading

andrewlock commented Nov 9, 2024 •

edited

Loading