Repository navigation
Conversation
Resolves dotnet#56815 (comment) The initial fix was made in INTERNAL on dotnet/dotnet 11.0.100 for GA. https://login.microsoftonline.com/72f988bf-86f1-41af-91ab-2d7cd011db47/oauth2/authorize?client_id=499b84ac-1321-427f-aa17-267ca6975798&site_id=501454&response_mode=form_post&response_type=code+id_token&redirect_uri=https%3A%2F%2Fspsprodcus3.vssps.visualstudio.com%2F_signedin&nonce=6e955c3b-2d8e-4b63-a9e6-27ed4d3332c6&state=realm%3Ddev.azure.com%26reply_to%3Dhttps%253A%252F%252Fdev.azure.com%252Fdnceng%252Finternal%252F_git%252Fdotnet-dotnet%252Fpullrequest%252F67911%26ht%3D2%26protocol%3Dentra.web.3%26hid%3Db55de4ed-4b5a-4215-a8e4-0a0a5f71e7d8%26nonce%3D6e955c3b-2d8e-4b63-a9e6-27ed4d3332c6%26lltid%3D221510f4-2b26-40d5-988a-08b05ea44407&resource=499b84ac-1321-427f-aa17-267ca6975798&cid=6e955c3b-2d8e-4b63-a9e6-27ed4d3332c6&wsucxt=1&instance_aware=true Scenario Pre-patch Post-patch --version, default NativeAOT path 2.11 s 0.21 s No-op build, managed CLI + MSBuild logger 8.76 s 4.52 s Build telemetry overhead versus opt-out 5.05 s 0.91 s The tracerProvider often get blocked on pending HTTP requests from recent Azure Monitor Exporter changes. It is supposed to be a minimal shutdown delay. Pending requests often take 3-5 seconds to complete, which results in extreme exit delay. The MS Build logger telemetry client calls ForceFlush, which is why the additional change is needed. The defaultTimeout remains as 5 seconds, because it informs network timeout for HTTP requests in options.Retry.NetworkTimeout - since telemetry POST completion time is often around 3-5 seconds, setting this lower would drop telemetry. -- The DIFFERENCE between this PR and what went into GA: - Fixed a test only in dotnet/sdk - I went from 10 to 100 ms because 10ms may not give enough time to write blob storage on slower disks for short lived commands When merging, prefer these changes.
|
Azure Pipelines: Successfully started running 2 pipeline(s). 1 pipeline(s) were filtered out due to trigger conditions. There may be pipelines that require an authorized user to comment /azp run to run. |
Member
Author
|
Already approved in servicing |
baronfel
approved these changes
Oct 30, 2026
Member
Author
|
Once this is approved I will backport to 11.0.2xx or monitor flow to make sure it gets in. |
Member
Author
|
Ah, ty 😉 |
Member
Author
|
/backport to release/11.0.2xx |
Contributor
|
Started backporting to |
Contributor
|
@nagilson backporting to |
nagilson
enabled auto-merge
October 30, 2026 21:30
Contributor
There was a problem hiding this comment.
🟡 Changes recommended
The local MSBuild flush timeout lacks regression coverage because existing tests do not exercise the non-CI, non-OTLP path and would not catch a reversion to Timeout.Infinite.
1 open issue
What changed in this PR
This PR addresses telemetry-related CLI exit delays reported in #56815 by bounding local trace persistence waits, including the MSBuild logger’s flush path.
Changes:
- Applies a 100-ms local trace budget to successful shutdown and nonterminal flushes.
- Adds exporter persistence tests and updates the shutdown timeout assertion.
- Documents the bounded persistence behavior while retaining existing CI and failed-command budgets.
| File | Description |
|---|---|
test/Microsoft.DotNet.Cli.Telemetry.Tests/AzureExporterLifecycleTests.cs |
Adds shutdown and force-flush persistence checks using a 100-ms budget. |
test/dotnet.Tests/TelemetryTests/TelemetryClientTests.cs |
Updates the expected successful local shutdown timeout. |
src/Cli/dotnet/Telemetry/TelemetryClient.cs |
Bounds local trace shutdown and force-flush waits. |
documentation/project-docs/telemetry.md |
Documents the local persistence budgets and MSBuild flush behavior. |
Review effort: Balanced
Comment on lines
+294
to
+296
| int timeout = s_isCIEnvironment || s_enableOtlpExporter | ||
| ? s_shutdownTimeoutMs | ||
| : LocalTracePersistenceTimeoutMilliseconds; |
This branch has not been deployed
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 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.
Resolves #56815 (comment)
The initial fix was made in INTERNAL on dotnet/dotnet 11.0.100 for GA.
https://dev.azure.com/dnceng/internal/_git/dotnet-dotnet/pullrequest/67911?_a=files
The tracerProvider often get blocked on pending HTTP requests from recent Azure Monitor Exporter changes. It is supposed to be a minimal shutdown delay. Pending requests often take 3-5 seconds to complete, which results in extreme exit delay.
The MS Build logger telemetry client calls ForceFlush, which is why the additional change is needed. The defaultTimeout remains as 5 seconds, because it informs network timeout for HTTP requests in options.Retry.NetworkTimeout - since telemetry POST completion time is often around 3-5 seconds, setting this lower would drop telemetry.
--
The DIFFERENCE between this PR and what went into GA:
When merging, prefer these changes.