Replace AWS ARM64 graviton images with Azure Managed DevOps pools #6787
+42
−46
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
andrewlock
added
the
area:builds
Datadog ReportBranch report: ✅ 0 Failed, 247691 Passed, 2400 Skipped, 20h 47m 19.62s Total Time |
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:
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 (6787) - mean (76ms) : 71, 81
. : milestone, 76,
master - mean (69ms) : 66, 73
. : milestone, 69,
section CallTarget+Inlining+NGEN
This PR (6787) - mean (1,059ms) : 1029, 1088
. : milestone, 1059,
master - mean (1,008ms) : 983, 1033
. : milestone, 1008,
gantt
title Execution time (ms) FakeDbCommand (.NET Core 3.1)
dateFormat X
axisFormat %s
todayMarker off
section Baseline
This PR (6787) - mean (112ms) : 108, 116
. : milestone, 112,
master - mean (102ms) : 100, 104
. : milestone, 102,
section CallTarget+Inlining+NGEN
This PR (6787) - mean (729ms) : crit, 708, 749
. : crit, milestone, 729,
master - mean (688ms) : 670, 706
. : milestone, 688,
gantt
title Execution time (ms) FakeDbCommand (.NET 6)
dateFormat X
axisFormat %s
todayMarker off
section Baseline
This PR (6787) - mean (98ms) : 94, 103
. : milestone, 98,
master - mean (89ms) : 87, 91
. : milestone, 89,
section CallTarget+Inlining+NGEN
This PR (6787) - mean (687ms) : crit, 668, 707
. : crit, milestone, 687,
master - mean (642ms) : 626, 658
. : milestone, 642,
gantt
title Execution time (ms) HttpMessageHandler (.NET Framework 4.6.2)
dateFormat X
axisFormat %s
todayMarker off
section Baseline
This PR (6787) - mean (190ms) : 186, 195
. : milestone, 190,
master - mean (190ms) : 186, 194
. : milestone, 190,
section CallTarget+Inlining+NGEN
This PR (6787) - mean (1,103ms) : 1073, 1134
. : milestone, 1103,
master - mean (1,109ms) : 1079, 1140
. : milestone, 1109,
gantt
title Execution time (ms) HttpMessageHandler (.NET Core 3.1)
dateFormat X
axisFormat %s
todayMarker off
section Baseline
This PR (6787) - mean (269ms) : 264, 274
. : milestone, 269,
master - mean (269ms) : 265, 273
. : milestone, 269,
section CallTarget+Inlining+NGEN
This PR (6787) - mean (873ms) : 843, 903
. : milestone, 873,
master - mean (880ms) : 847, 914
. : milestone, 880,
gantt
title Execution time (ms) HttpMessageHandler (.NET 6)
dateFormat X
axisFormat %s
todayMarker off
section Baseline
This PR (6787) - mean (261ms) : 257, 265
. : milestone, 261,
master - mean (263ms) : 259, 267
. : milestone, 263,
section CallTarget+Inlining+NGEN
This PR (6787) - mean (857ms) : 817, 896
. : milestone, 857,
master - mean (858ms) : 822, 895
. : milestone, 858,
|
Benchmarks Report for tracer 🐌Benchmarks for #6787 compared to master:
The following thresholds were used for comparing the benchmark speeds:
Allocation changes below 0.5% are ignored. Benchmark detailsBenchmarks.Trace.ActivityBenchmark - Same speed ✔️ Same allocations ✔️Raw results
Benchmarks.Trace.AgentWriterBenchmark - Same speed ✔️ Same allocations ✔️Raw results
Benchmarks.Trace.AspNetCoreBenchmark - Same speed ✔️ Same allocations ✔️Raw results
Benchmarks.Trace.CIVisibilityProtocolWriterBenchmark - Same speed ✔️ Same allocations ✔️Raw results
Benchmarks.Trace.DbCommandBenchmark - Same speed ✔️ Same allocations ✔️Raw results
Benchmarks.Trace.ElasticsearchBenchmark - Same speed ✔️ Same allocations ✔️Raw results
Benchmarks.Trace.GraphQLBenchmark - Same speed ✔️ Same allocations ✔️Raw results
Benchmarks.Trace.HttpClientBenchmark - Same speed ✔️ Same allocations ✔️Raw results
Benchmarks.Trace.ILoggerBenchmark - Same speed ✔️ Same allocations ✔️Raw results
Benchmarks.Trace.Log4netBenchmark - Same speed ✔️ Same allocations ✔️Raw results
Benchmarks.Trace.NLogBenchmark - Same speed ✔️ Same allocations ✔️Raw results
Benchmarks.Trace.RedisBenchmark - Same speed ✔️ Same allocations ✔️Raw results
Benchmarks.Trace.SerilogBenchmark - Same speed ✔️ Same allocations ✔️Raw results
Benchmarks.Trace.SpanBenchmark - Faster 🎉 Same allocations ✔️
|
| Benchmark | base/diff | Base Median (ns) | Diff Median (ns) | Modality |
|---|---|---|---|---|
| Benchmarks.Trace.SpanBenchmark.StartFinishSpan‑net6.0 | 1.190 | 468.60 | 393.93 |
Raw results
| Branch | Method | Toolchain | Mean | StdError | StdDev | Gen 0 | Gen 1 | Gen 2 | Allocated |
|---|---|---|---|---|---|---|---|---|---|
| master | StartFinishSpan |
net6.0 | 469ns | 0.166ns | 0.643ns | 0.0082 | 0 | 0 | 576 B |
| master | StartFinishSpan |
netcoreapp3.1 | 581ns | 0.902ns | 3.49ns | 0.00774 | 0 | 0 | 576 B |
| master | StartFinishSpan |
net472 | 598ns | 0.148ns | 0.573ns | 0.0917 | 0 | 0 | 578 B |
| master | StartFinishScope |
net6.0 | 495ns | 0.178ns | 0.69ns | 0.00985 | 0 | 0 | 696 B |
| master | StartFinishScope |
netcoreapp3.1 | 723ns | 0.293ns | 1.13ns | 0.00935 | 0 | 0 | 696 B |
| master | StartFinishScope |
net472 | 852ns | 0.339ns | 1.27ns | 0.104 | 0 | 0 | 658 B |
| #6787 | StartFinishSpan |
net6.0 | 394ns | 0.106ns | 0.396ns | 0.00801 | 0 | 0 | 576 B |
| #6787 | StartFinishSpan |
netcoreapp3.1 | 589ns | 0.387ns | 1.45ns | 0.00766 | 0 | 0 | 576 B |
| #6787 | StartFinishSpan |
net472 | 662ns | 0.389ns | 1.51ns | 0.0915 | 0 | 0 | 578 B |
| #6787 | StartFinishScope |
net6.0 | 494ns | 0.124ns | 0.479ns | 0.00976 | 0 | 0 | 696 B |
| #6787 | StartFinishScope |
netcoreapp3.1 | 735ns | 0.241ns | 0.902ns | 0.00925 | 0 | 0 | 696 B |
| #6787 | StartFinishScope |
net472 | 854ns | 0.639ns | 2.48ns | 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 | 676ns | 0.192ns | 0.745ns | 0.00986 | 0 | 0 | 696 B |
| master | RunOnMethodBegin |
netcoreapp3.1 | 957ns | 0.315ns | 1.22ns | 0.00911 | 0 | 0 | 696 B |
| master | RunOnMethodBegin |
net472 | 1.12μs | 0.226ns | 0.876ns | 0.104 | 0 | 0 | 658 B |
| #6787 | RunOnMethodBegin |
net6.0 | 666ns | 0.256ns | 0.99ns | 0.00979 | 0 | 0 | 696 B |
| #6787 | RunOnMethodBegin |
netcoreapp3.1 | 1.03μs | 0.557ns | 2.16ns | 0.00914 | 0 | 0 | 696 B |
| #6787 | RunOnMethodBegin |
net472 | 1.1μs | 0.254ns | 0.984ns | 0.104 | 0 | 0 | 658 B |
andrewlock
force-pushed
the
andrew/ci/managed-devops-pools
branch
from
b6059b2 to
18f9104
Compare
This reverts commit 18f9104.
…ad on linux, to avoid permissions issues
andrewlock
force-pushed
the
andrew/ci/managed-devops-pools
branch
from
9f5e76c to
ed15394
Compare
kevingosse
approved these changes
Mar 21, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Summary of changes
Replaces our previous ARM64 auto-scaling graviton agents with Azure Managed DevOps pool images
Reason for change
The AWS sandbox is being retired, which we means we would have to rebuild the somewhat hacky "VMSS emulator" that we built there, because Microsoft have refused to support ARM64 VMSS pools with Azure DevOps
This will hopefully bring a number of additional benefits, in that it supports more dynamic (and greater) scaling, allows dynamic scaling up and down (e.g. scale all the images off at night) and still allows us to use custom images.
Implementation details
Followed the documentation to perform all the preliminary configuration to enable Managed DevOps pools. Adding a new pool needs to be done from Azure, but should be relatively simple.
It still requires that create new VM images in certain circumstances, but now the process of creating an image is identical for x64 and arm64.
Currently the x64 Linux and Windows images are still using VMSS instead of Managed DevOps pools, but we can consider migrating those across in the future.
Test coverage
Ran some manual tests, and once the initial configuration was done, worked like a charm with one exception - the Docker tests were failing to upload the logs, because localstack was locking the directory. I don't think we actually need to report this temp file, so removed the binding for simplicity.
However we also realised that the same permissions issue was preventing crashdump uploads too, so fixed that "everywhere" by creating a small template to do the permissions fix.