Closed vandonr closed 1 week ago
Branch report: vandonr/chiseled
Commit report: 7f869b9
Test service: dd-trace-dotnet
:white_check_mark: 0 Failed, 340205 Passed, 1652 Skipped, 14h 17m 21.84s Total Time :hourglass: 1 Performance Regression
Baseline
- Samples.HttpMessageHandler.windows.net60.json.scenarios
272.14ms (+6.44ms, +2%) - DetailsExecution-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:
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 (5731) - mean (73ms) : 64, 83
. : milestone, 73,
master - mean (73ms) : 66, 81
. : milestone, 73,
section CallTarget+Inlining+NGEN
This PR (5731) - mean (987ms) : 971, 1002
. : milestone, 987,
master - mean (987ms) : 966, 1007
. : milestone, 987,
gantt
title Execution time (ms) FakeDbCommand (.NET Core 3.1)
dateFormat X
axisFormat %s
todayMarker off
section Baseline
This PR (5731) - mean (109ms) : 106, 112
. : milestone, 109,
master - mean (110ms) : 107, 113
. : milestone, 110,
section CallTarget+Inlining+NGEN
This PR (5731) - mean (692ms) : 671, 714
. : milestone, 692,
master - mean (697ms) : 677, 718
. : milestone, 697,
gantt
title Execution time (ms) FakeDbCommand (.NET 6)
dateFormat X
axisFormat %s
todayMarker off
section Baseline
This PR (5731) - mean (93ms) : 90, 96
. : milestone, 93,
master - mean (93ms) : 89, 96
. : milestone, 93,
section CallTarget+Inlining+NGEN
This PR (5731) - mean (655ms) : 632, 679
. : milestone, 655,
master - mean (652ms) : 630, 675
. : milestone, 652,
gantt
title Execution time (ms) HttpMessageHandler (.NET Framework 4.6.2)
dateFormat X
axisFormat %s
todayMarker off
section Baseline
This PR (5731) - mean (192ms) : 188, 196
. : milestone, 192,
master - mean (190ms) : 187, 194
. : milestone, 190,
section CallTarget+Inlining+NGEN
This PR (5731) - mean (1,091ms) : 1067, 1115
. : milestone, 1091,
master - mean (1,080ms) : 1055, 1104
. : milestone, 1080,
gantt
title Execution time (ms) HttpMessageHandler (.NET Core 3.1)
dateFormat X
axisFormat %s
todayMarker off
section Baseline
This PR (5731) - mean (279ms) : 271, 286
. : milestone, 279,
master - mean (275ms) : 271, 280
. : milestone, 275,
section CallTarget+Inlining+NGEN
This PR (5731) - mean (873ms) : 853, 893
. : milestone, 873,
master - mean (869ms) : 846, 892
. : milestone, 869,
gantt
title Execution time (ms) HttpMessageHandler (.NET 6)
dateFormat X
axisFormat %s
todayMarker off
section Baseline
This PR (5731) - mean (272ms) : 268, 276
. : milestone, 272,
master - mean (264ms) : 260, 269
. : milestone, 264,
section CallTarget+Inlining+NGEN
This PR (5731) - mean (871ms) : 825, 917
. : milestone, 871,
master - mean (851ms) : 826, 875
. : milestone, 851,
Benchmarks for #5731 compared to master:
The following thresholds were used for comparing the benchmark speeds:
Allocation changes below 0.5% are ignored.
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 (5731) (11.739M) : 0, 11739337
master (11.680M) : 0, 11680352
benchmarks/2.9.0 (11.545M) : 0, 11545343
section Automatic
This PR (5731) (7.881M) : 0, 7880906
master (7.820M) : 0, 7819638
benchmarks/2.9.0 (8.126M) : 0, 8125785
section Trace stats
master (8.158M) : 0, 8157732
section Manual
This PR (5731) (10.044M) : 0, 10044313
master (9.925M) : 0, 9924578
section Manual + Automatic
This PR (5731) (7.400M) : 0, 7399648
master (7.482M) : 0, 7482449
section Version Conflict
master (6.635M) : 0, 6634968
gantt
title Throughput Linux arm64 (Total requests)
dateFormat X
axisFormat %s
section Baseline
This PR (5731) (9.462M) : 0, 9462166
master (9.294M) : 0, 9294338
benchmarks/2.9.0 (9.619M) : 0, 9619209
section Automatic
This PR (5731) (6.541M) : 0, 6541395
master (6.468M) : 0, 6468177
section Trace stats
master (6.754M) : 0, 6754461
section Manual
This PR (5731) (8.336M) : 0, 8336159
master (8.162M) : 0, 8161768
section Manual + Automatic
This PR (5731) (6.259M) : 0, 6259087
master (6.262M) : 0, 6262414
section Version Conflict
master (5.565M) : 0, 5565453
gantt
title Throughput Windows x64 (Total requests)
dateFormat X
axisFormat %s
section Baseline
This PR (5731) (10.027M) : 0, 10026717
master (10.199M) : 0, 10198615
benchmarks/2.9.0 (10.207M) : 0, 10207218
section Automatic
This PR (5731) (7.068M) : 0, 7068167
master (7.255M) : 0, 7255144
benchmarks/2.9.0 (7.563M) : 0, 7562593
section Trace stats
master (7.633M) : 0, 7633330
section Manual
This PR (5731) (8.830M) : 0, 8830332
master (9.048M) : 0, 9047753
section Manual + Automatic
This PR (5731) (6.905M) : 0, 6905005
master (6.964M) : 0, 6963644
section Version Conflict
master (6.423M) : 0, 6422917
Summary of changes
The docker setup on a chiseled base image was using a placeholder file to create an empty directory, this is actually not needed, we can copy empty directories with the COPY command (what was missing from the commented code was trailing slashes). I also changed the
chown
argument, because I don't know where 64198 was coming from. It's important that the user$APP_UID
owns the logs folder, otherwise they cannot write to it (somehow,chmod=777
alone didn't work when I tested it). ~In the existing setup before this PR, I'm not sure any logs were written.~ -> I checked and logs were actually produced, but I prefer the new setup :)