FazBrowse GitHub Viewer | Trending |
URL:
| Home
Tools: [Download Repo ZIP]   [Original HTTPS Page]

Use direct URLs for test resources by ericstj · Pull Request #7663 · dotnet/machinelearning · GitHub

Use direct URLs for test resources - #7663

Merged
ericstj merged 4 commits into
dotnet:mainfrom
ericstj:fix/direct-test-resource-urls
Jul 22, 2026
Merged

Use direct URLs for test resources#7663
ericstj merged 4 commits into
dotnet:mainfrom
ericstj:fix/direct-test-resource-urls

Conversation

ericstj commented Jul 16, 2026
edited
Loading

Copy link
Copy Markdown
Member

Summary

  • temporarily bypass deleted aka.ms/mlnet-resources aliases in tests
  • use direct public blob URLs for image and benchmark datasets
  • centralize bounded, retrying test downloads in TestDownloadUtils

Fixes #7662

Validation

  • rebuilt Microsoft.ML.TestFrameworkCommon, Microsoft.ML.AutoML.Tests, Microsoft.ML.TensorFlow.Tests, and Microsoft.ML.PerformanceTests
  • ran TensorFlowGettingSchemaMultipleTimes and AutoFeaturizer_image_test on net8.0

Temporarily bypass deleted aka.ms aliases for test datasets and centralize resilient direct downloads in the shared test framework.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Copilot-Session: 57cca22a-bb40-4b25-b169-1c9e46ae88e1
Copilot AI review requested due to automatic review settings July 16, 2026 20:13

Copilot AI left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

Choose a reason Spam Abuse Off Topic Outdated Duplicate Resolved Low Quality

Pull request overview

This PR updates ML.NET test infrastructure to bypass broken/deleted aka.ms/mlnet-resources aliases by switching tests/benchmarks to direct public URLs, and centralizes downloading logic into a shared helper to improve reliability.

Changes:

  • Added TestDownloadUtils to perform bounded, retrying downloads via a temp-file + atomic move pattern.
  • Updated TensorFlow and AutoML test datasets to download from direct mlpublicassets blob URLs.
  • Updated performance benchmarks to fetch datasets from either raw GitHub (repo test/data) or mlpublicassets benchmark blobs.
Show a summary per file
File Description
test/Microsoft.ML.TestFrameworkCommon/TestDownloadUtils.cs New shared download helper with retry + temp-file move semantics.
test/Microsoft.ML.TestFrameworkCommon/Microsoft.ML.TestFrameworkCommon.csproj Enables nullable annotations and links in RetryHelper for retries.
test/Microsoft.ML.TensorFlow.Tests/TensorflowTests.cs Replaces ResourceManagerUtils downloads with direct blob URLs + TestDownloadUtils.
test/Microsoft.ML.PerformanceTests/ImageClassificationBench.cs Switches dataset URL from aka.ms to direct blob URL.
test/Microsoft.ML.PerformanceTests/BenchmarkBase.cs Replaces resource manager downloads with direct URL downloads via TestDownloadUtils.
test/Microsoft.ML.AutoML.Tests/DatasetUtil.cs Switches dataset URL from aka.ms to direct blob URL.

Copilot's findings

  • Files reviewed: 6/6 changed files
  • Comments generated: 2

Validate response content length and remove the unused benchmark path local.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Copilot-Session: 57cca22a-bb40-4b25-b169-1c9e46ae88e1
ericstj enabled auto-merge (squash) July 16, 2026 21:11
ericstj closed this Jul 17, 2026
auto-merge was automatically disabled July 17, 2026 02:03

Pull request was closed

ericstj reopened this Jul 17, 2026
ericstj requested review from matouskozak and tarekgh July 17, 2026 02:04

matouskozak left a comment

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

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

Choose a reason Spam Abuse Off Topic Outdated Duplicate Resolved Low Quality

Thanks a lot for fixing this

Use a local-only helper for checked-in benchmark data and reserve network downloads for explicit external datasets.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Copilot-Session: 57cca22a-bb40-4b25-b169-1c9e46ae88e1

tarekgh left a comment

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

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

Choose a reason Spam Abuse Off Topic Outdated Duplicate Resolved Low Quality

LGTM, thanks @ericstj for taking care with this!

Handle OperationCanceledException from timed-out response body copies in addition to TaskCanceledException.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Copilot-Session: 57cca22a-bb40-4b25-b169-1c9e46ae88e1
ericstj enabled auto-merge (squash) July 17, 2026 16:16

Copy link
Copy Markdown
Member

@ericstj is this ready for merge? I think the failures are related to #7657

ericstj commented Jul 22, 2026

Copy link
Copy Markdown
Member Author

Yeah, but I can't merge it. Let me try the override to see if it works here.

ericstj commented Jul 22, 2026

Copy link
Copy Markdown
Member Author

/ba-g Failures are #7657

ericstj merged commit 2a06d59 into dotnet:main Jul 22, 2026
24 of 26 checks passed
ericstj deleted the fix/direct-test-resource-urls branch July 22, 2026 01:34
github-actions Bot locked and limited conversation to collaborators Aug 21, 2026
This file contains hidden or 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
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Many test failures with System.IO.InvalidDataException : End of Central Directory record could not be found.

4 participants


Back | FazBrowse Home | New Git URL