Fix OleTxTests.Recovery CI timeout by bounding remote process wall-clock time by Copilot · Pull Request #125813 · dotnet/runtime

Copilot · 2026-03-20T00:24:14Z

Description

OleTxTests.Recovery repeatedly times out the CI work item (the entire System.Transactions.Local.Tests work item is killed). The CI error gives no detail beyond the timeout — no stack traces or indication of where the hang occurs.

The most likely scenario is that MSDTC is unresponsive on certain CI machines, causing the child process spawned by RemoteExecutor.Invoke to hang (either stuck in MSDTC COM interop, or sitting in its intentional Thread.Sleep waiting for a commit callback that never arrives). Since RemoteInvokeOptions.TimeOut defaults to Timeout.Infinite, RemoteInvokeHandle.Dispose() blocks indefinitely waiting for the child to exit, and the test never completes.

It's also possible the main test thread hangs in tx.Commit() or one of the WaitOne(Timeout) calls. This change would not help in that case — only the child process is bounded. If the timeout continues to repro after this change, the main thread is the more likely culprit and would need its own timeout or the test should be skipped on affected CI legs.

Change:

Add TimeOut = 120_000 to RemoteInvokeOptions so RemoteInvokeHandle.Dispose() will force-kill the child process after 2 minutes (with a diagnostic dump at the halfway point) instead of blocking indefinitely. This bounds the test's wall-clock time and should prevent it from consuming the entire work item timeout budget.

Changes

src/libraries/System.Transactions.Local/tests/OleTxTests.cs

Testing

This change targets a flaky/hanging test; no new test logic is added. The fix will be validated by the reduction in CI work item timeouts on Windows runs. The test requires MSDTC and cannot be run locally in most environments (the ConditionalFact skips it when MSDTC is unavailable).

FixesSystem.Transactions.Tests.OleTxTests.Recovery timeout #122621

…all-clock time Co-authored-by: danmoseley <6385855+danmoseley@users.noreply.github.com>

dotnet-policy-service · 2026-03-20T00:41:27Z

Tagging subscribers to this area: @SamMonoRT, @dotnet/efteam
See info in area-owners.md if you want to be subscribed.

Copilot

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.

Copilot

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated 2 comments.

…out on async methods) Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

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

Copilot

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.

Copilot

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated no new comments.

The Recovery test consistently times out (20+ min) under stress modes (fullpgo, jitstress2_jitstressregs) — confirmed from Helix logs across multiple hits. PR #125813 added a 120s child-process timeout but the main thread still hangs waiting for MSDTC under slow runtimes. This is a libraries-level test exercising MSDTC/OLE transaction recovery; running it under JIT stress provides no additional signal. Skip it on all non-regular CoreCLR test modes. Fixes#126304 Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Copilot · 2026-03-20T00:24:14Z

Description

OleTxTests.Recovery repeatedly times out the CI work item (the entire System.Transactions.Local.Tests work item is killed). The CI error gives no detail beyond the timeout — no stack traces or indication of where the hang occurs.

The most likely scenario is that MSDTC is unresponsive on certain CI machines, causing the child process spawned by RemoteExecutor.Invoke to hang (either stuck in MSDTC COM interop, or sitting in its intentional Thread.Sleep waiting for a commit callback that never arrives). Since RemoteInvokeOptions.TimeOut defaults to Timeout.Infinite, RemoteInvokeHandle.Dispose() blocks indefinitely waiting for the child to exit, and the test never completes.

It's also possible the main test thread hangs in tx.Commit() or one of the WaitOne(Timeout) calls. This change would not help in that case — only the child process is bounded. If the timeout continues to repro after this change, the main thread is the more likely culprit and would need its own timeout or the test should be skipped on affected CI legs.

Change:

Add TimeOut = 120_000 to RemoteInvokeOptions so RemoteInvokeHandle.Dispose() will force-kill the child process after 2 minutes (with a diagnostic dump at the halfway point) instead of blocking indefinitely. This bounds the test's wall-clock time and should prevent it from consuming the entire work item timeout budget.

Changes

src/libraries/System.Transactions.Local/tests/OleTxTests.cs

Testing

This change targets a flaky/hanging test; no new test logic is added. The fix will be validated by the reduction in CI work item timeouts on Windows runs. The test requires MSDTC and cannot be run locally in most environments (the ConditionalFact skips it when MSDTC is unavailable).

FixesSystem.Transactions.Tests.OleTxTests.Recovery timeout #122621

…all-clock time Co-authored-by: danmoseley <6385855+danmoseley@users.noreply.github.com>

dotnet-policy-service · 2026-03-20T00:41:27Z

Tagging subscribers to this area: @SamMonoRT, @dotnet/efteam
See info in area-owners.md if you want to be subscribed.

Copilot

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.

Copilot

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated 2 comments.

…out on async methods) Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

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

Copilot

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.

Copilot

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated no new comments.

The Recovery test consistently times out (20+ min) under stress modes (fullpgo, jitstress2_jitstressregs) — confirmed from Helix logs across multiple hits. PR #125813 added a 120s child-process timeout but the main thread still hangs waiting for MSDTC under slow runtimes. This is a libraries-level test exercising MSDTC/OLE transaction recovery; running it under JIT stress provides no additional signal. Skip it on all non-regular CoreCLR test modes. Fixes#126304 Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Copilot · 2026-03-20T00:24:14Z

Description

OleTxTests.Recovery repeatedly times out the CI work item (the entire System.Transactions.Local.Tests work item is killed). The CI error gives no detail beyond the timeout — no stack traces or indication of where the hang occurs.

The most likely scenario is that MSDTC is unresponsive on certain CI machines, causing the child process spawned by RemoteExecutor.Invoke to hang (either stuck in MSDTC COM interop, or sitting in its intentional Thread.Sleep waiting for a commit callback that never arrives). Since RemoteInvokeOptions.TimeOut defaults to Timeout.Infinite, RemoteInvokeHandle.Dispose() blocks indefinitely waiting for the child to exit, and the test never completes.

It's also possible the main test thread hangs in tx.Commit() or one of the WaitOne(Timeout) calls. This change would not help in that case — only the child process is bounded. If the timeout continues to repro after this change, the main thread is the more likely culprit and would need its own timeout or the test should be skipped on affected CI legs.

Change:

Add TimeOut = 120_000 to RemoteInvokeOptions so RemoteInvokeHandle.Dispose() will force-kill the child process after 2 minutes (with a diagnostic dump at the halfway point) instead of blocking indefinitely. This bounds the test's wall-clock time and should prevent it from consuming the entire work item timeout budget.

Changes

src/libraries/System.Transactions.Local/tests/OleTxTests.cs

Testing

This change targets a flaky/hanging test; no new test logic is added. The fix will be validated by the reduction in CI work item timeouts on Windows runs. The test requires MSDTC and cannot be run locally in most environments (the ConditionalFact skips it when MSDTC is unavailable).

FixesSystem.Transactions.Tests.OleTxTests.Recovery timeout #122621

…all-clock time Co-authored-by: danmoseley <6385855+danmoseley@users.noreply.github.com>

dotnet-policy-service · 2026-03-20T00:41:27Z

Tagging subscribers to this area: @SamMonoRT, @dotnet/efteam
See info in area-owners.md if you want to be subscribed.

Copilot

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.

Copilot

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated 2 comments.

…out on async methods) Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

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

Copilot

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.

Copilot

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated no new comments.

The Recovery test consistently times out (20+ min) under stress modes (fullpgo, jitstress2_jitstressregs) — confirmed from Helix logs across multiple hits. PR #125813 added a 120s child-process timeout but the main thread still hangs waiting for MSDTC under slow runtimes. This is a libraries-level test exercising MSDTC/OLE transaction recovery; running it under JIT stress provides no additional signal. Skip it on all non-regular CoreCLR test modes. Fixes#126304 Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Copilot · 2026-03-20T00:24:14Z

Description

OleTxTests.Recovery repeatedly times out the CI work item (the entire System.Transactions.Local.Tests work item is killed). The CI error gives no detail beyond the timeout — no stack traces or indication of where the hang occurs.

The most likely scenario is that MSDTC is unresponsive on certain CI machines, causing the child process spawned by RemoteExecutor.Invoke to hang (either stuck in MSDTC COM interop, or sitting in its intentional Thread.Sleep waiting for a commit callback that never arrives). Since RemoteInvokeOptions.TimeOut defaults to Timeout.Infinite, RemoteInvokeHandle.Dispose() blocks indefinitely waiting for the child to exit, and the test never completes.

It's also possible the main test thread hangs in tx.Commit() or one of the WaitOne(Timeout) calls. This change would not help in that case — only the child process is bounded. If the timeout continues to repro after this change, the main thread is the more likely culprit and would need its own timeout or the test should be skipped on affected CI legs.

Change:

Add TimeOut = 120_000 to RemoteInvokeOptions so RemoteInvokeHandle.Dispose() will force-kill the child process after 2 minutes (with a diagnostic dump at the halfway point) instead of blocking indefinitely. This bounds the test's wall-clock time and should prevent it from consuming the entire work item timeout budget.

Changes

src/libraries/System.Transactions.Local/tests/OleTxTests.cs

Testing

This change targets a flaky/hanging test; no new test logic is added. The fix will be validated by the reduction in CI work item timeouts on Windows runs. The test requires MSDTC and cannot be run locally in most environments (the ConditionalFact skips it when MSDTC is unavailable).

FixesSystem.Transactions.Tests.OleTxTests.Recovery timeout #122621

…all-clock time Co-authored-by: danmoseley <6385855+danmoseley@users.noreply.github.com>

dotnet-policy-service · 2026-03-20T00:41:27Z

Tagging subscribers to this area: @SamMonoRT, @dotnet/efteam
See info in area-owners.md if you want to be subscribed.

Copilot

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.

Copilot

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated 2 comments.

…out on async methods) Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

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

Copilot

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.

Copilot

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated no new comments.

The Recovery test consistently times out (20+ min) under stress modes (fullpgo, jitstress2_jitstressregs) — confirmed from Helix logs across multiple hits. PR #125813 added a 120s child-process timeout but the main thread still hangs waiting for MSDTC under slow runtimes. This is a libraries-level test exercising MSDTC/OLE transaction recovery; running it under JIT stress provides no additional signal. Skip it on all non-regular CoreCLR test modes. Fixes#126304 Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Copilot · 2026-03-20T00:24:14Z

Description

OleTxTests.Recovery repeatedly times out the CI work item (the entire System.Transactions.Local.Tests work item is killed). The CI error gives no detail beyond the timeout — no stack traces or indication of where the hang occurs.

The most likely scenario is that MSDTC is unresponsive on certain CI machines, causing the child process spawned by RemoteExecutor.Invoke to hang (either stuck in MSDTC COM interop, or sitting in its intentional Thread.Sleep waiting for a commit callback that never arrives). Since RemoteInvokeOptions.TimeOut defaults to Timeout.Infinite, RemoteInvokeHandle.Dispose() blocks indefinitely waiting for the child to exit, and the test never completes.

It's also possible the main test thread hangs in tx.Commit() or one of the WaitOne(Timeout) calls. This change would not help in that case — only the child process is bounded. If the timeout continues to repro after this change, the main thread is the more likely culprit and would need its own timeout or the test should be skipped on affected CI legs.

Change:

Add TimeOut = 120_000 to RemoteInvokeOptions so RemoteInvokeHandle.Dispose() will force-kill the child process after 2 minutes (with a diagnostic dump at the halfway point) instead of blocking indefinitely. This bounds the test's wall-clock time and should prevent it from consuming the entire work item timeout budget.

Changes

src/libraries/System.Transactions.Local/tests/OleTxTests.cs

Testing

This change targets a flaky/hanging test; no new test logic is added. The fix will be validated by the reduction in CI work item timeouts on Windows runs. The test requires MSDTC and cannot be run locally in most environments (the ConditionalFact skips it when MSDTC is unavailable).

FixesSystem.Transactions.Tests.OleTxTests.Recovery timeout #122621

…all-clock time Co-authored-by: danmoseley <6385855+danmoseley@users.noreply.github.com>

dotnet-policy-service · 2026-03-20T00:41:27Z

Tagging subscribers to this area: @SamMonoRT, @dotnet/efteam
See info in area-owners.md if you want to be subscribed.

Copilot

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.

Copilot

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated 2 comments.

…out on async methods) Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

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

Copilot

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.

Copilot

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated no new comments.

The Recovery test consistently times out (20+ min) under stress modes (fullpgo, jitstress2_jitstressregs) — confirmed from Helix logs across multiple hits. PR #125813 added a 120s child-process timeout but the main thread still hangs waiting for MSDTC under slow runtimes. This is a libraries-level test exercising MSDTC/OLE transaction recovery; running it under JIT stress provides no additional signal. Skip it on all non-regular CoreCLR test modes. Fixes#126304 Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Copilot · 2026-03-20T00:24:14Z

Description

OleTxTests.Recovery repeatedly times out the CI work item (the entire System.Transactions.Local.Tests work item is killed). The CI error gives no detail beyond the timeout — no stack traces or indication of where the hang occurs.

The most likely scenario is that MSDTC is unresponsive on certain CI machines, causing the child process spawned by RemoteExecutor.Invoke to hang (either stuck in MSDTC COM interop, or sitting in its intentional Thread.Sleep waiting for a commit callback that never arrives). Since RemoteInvokeOptions.TimeOut defaults to Timeout.Infinite, RemoteInvokeHandle.Dispose() blocks indefinitely waiting for the child to exit, and the test never completes.

It's also possible the main test thread hangs in tx.Commit() or one of the WaitOne(Timeout) calls. This change would not help in that case — only the child process is bounded. If the timeout continues to repro after this change, the main thread is the more likely culprit and would need its own timeout or the test should be skipped on affected CI legs.

Change:

Add TimeOut = 120_000 to RemoteInvokeOptions so RemoteInvokeHandle.Dispose() will force-kill the child process after 2 minutes (with a diagnostic dump at the halfway point) instead of blocking indefinitely. This bounds the test's wall-clock time and should prevent it from consuming the entire work item timeout budget.

Changes

src/libraries/System.Transactions.Local/tests/OleTxTests.cs

Testing

This change targets a flaky/hanging test; no new test logic is added. The fix will be validated by the reduction in CI work item timeouts on Windows runs. The test requires MSDTC and cannot be run locally in most environments (the ConditionalFact skips it when MSDTC is unavailable).

FixesSystem.Transactions.Tests.OleTxTests.Recovery timeout #122621

…all-clock time Co-authored-by: danmoseley <6385855+danmoseley@users.noreply.github.com>

dotnet-policy-service · 2026-03-20T00:41:27Z

Tagging subscribers to this area: @SamMonoRT, @dotnet/efteam
See info in area-owners.md if you want to be subscribed.

Copilot

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.

Copilot

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated 2 comments.

…out on async methods) Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

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

Copilot

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.

Copilot

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated no new comments.

The Recovery test consistently times out (20+ min) under stress modes (fullpgo, jitstress2_jitstressregs) — confirmed from Helix logs across multiple hits. PR #125813 added a 120s child-process timeout but the main thread still hangs waiting for MSDTC under slow runtimes. This is a libraries-level test exercising MSDTC/OLE transaction recovery; running it under JIT stress provides no additional signal. Skip it on all non-regular CoreCLR test modes. Fixes#126304 Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Copilot · 2026-03-20T00:24:14Z

Description

OleTxTests.Recovery repeatedly times out the CI work item (the entire System.Transactions.Local.Tests work item is killed). The CI error gives no detail beyond the timeout — no stack traces or indication of where the hang occurs.

The most likely scenario is that MSDTC is unresponsive on certain CI machines, causing the child process spawned by RemoteExecutor.Invoke to hang (either stuck in MSDTC COM interop, or sitting in its intentional Thread.Sleep waiting for a commit callback that never arrives). Since RemoteInvokeOptions.TimeOut defaults to Timeout.Infinite, RemoteInvokeHandle.Dispose() blocks indefinitely waiting for the child to exit, and the test never completes.

It's also possible the main test thread hangs in tx.Commit() or one of the WaitOne(Timeout) calls. This change would not help in that case — only the child process is bounded. If the timeout continues to repro after this change, the main thread is the more likely culprit and would need its own timeout or the test should be skipped on affected CI legs.

Change:

Add TimeOut = 120_000 to RemoteInvokeOptions so RemoteInvokeHandle.Dispose() will force-kill the child process after 2 minutes (with a diagnostic dump at the halfway point) instead of blocking indefinitely. This bounds the test's wall-clock time and should prevent it from consuming the entire work item timeout budget.

Changes

src/libraries/System.Transactions.Local/tests/OleTxTests.cs

Testing

This change targets a flaky/hanging test; no new test logic is added. The fix will be validated by the reduction in CI work item timeouts on Windows runs. The test requires MSDTC and cannot be run locally in most environments (the ConditionalFact skips it when MSDTC is unavailable).

FixesSystem.Transactions.Tests.OleTxTests.Recovery timeout #122621

…all-clock time Co-authored-by: danmoseley <6385855+danmoseley@users.noreply.github.com>

dotnet-policy-service · 2026-03-20T00:41:27Z

Tagging subscribers to this area: @SamMonoRT, @dotnet/efteam
See info in area-owners.md if you want to be subscribed.

Copilot

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.

Copilot

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated 2 comments.

…out on async methods) Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

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

Copilot

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.

Copilot

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated no new comments.

The Recovery test consistently times out (20+ min) under stress modes (fullpgo, jitstress2_jitstressregs) — confirmed from Helix logs across multiple hits. PR #125813 added a 120s child-process timeout but the main thread still hangs waiting for MSDTC under slow runtimes. This is a libraries-level test exercising MSDTC/OLE transaction recovery; running it under JIT stress provides no additional signal. Skip it on all non-regular CoreCLR test modes. Fixes#126304 Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Copilot · 2026-03-20T00:24:14Z

Description

OleTxTests.Recovery repeatedly times out the CI work item (the entire System.Transactions.Local.Tests work item is killed). The CI error gives no detail beyond the timeout — no stack traces or indication of where the hang occurs.

The most likely scenario is that MSDTC is unresponsive on certain CI machines, causing the child process spawned by RemoteExecutor.Invoke to hang (either stuck in MSDTC COM interop, or sitting in its intentional Thread.Sleep waiting for a commit callback that never arrives). Since RemoteInvokeOptions.TimeOut defaults to Timeout.Infinite, RemoteInvokeHandle.Dispose() blocks indefinitely waiting for the child to exit, and the test never completes.

It's also possible the main test thread hangs in tx.Commit() or one of the WaitOne(Timeout) calls. This change would not help in that case — only the child process is bounded. If the timeout continues to repro after this change, the main thread is the more likely culprit and would need its own timeout or the test should be skipped on affected CI legs.

Change:

Add TimeOut = 120_000 to RemoteInvokeOptions so RemoteInvokeHandle.Dispose() will force-kill the child process after 2 minutes (with a diagnostic dump at the halfway point) instead of blocking indefinitely. This bounds the test's wall-clock time and should prevent it from consuming the entire work item timeout budget.

Changes

src/libraries/System.Transactions.Local/tests/OleTxTests.cs

Testing

This change targets a flaky/hanging test; no new test logic is added. The fix will be validated by the reduction in CI work item timeouts on Windows runs. The test requires MSDTC and cannot be run locally in most environments (the ConditionalFact skips it when MSDTC is unavailable).

FixesSystem.Transactions.Tests.OleTxTests.Recovery timeout #122621

…all-clock time Co-authored-by: danmoseley <6385855+danmoseley@users.noreply.github.com>

dotnet-policy-service · 2026-03-20T00:41:27Z

Tagging subscribers to this area: @SamMonoRT, @dotnet/efteam
See info in area-owners.md if you want to be subscribed.

Copilot

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.

Copilot

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated 2 comments.

…out on async methods) Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

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

Copilot

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.

Copilot

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated no new comments.

The Recovery test consistently times out (20+ min) under stress modes (fullpgo, jitstress2_jitstressregs) — confirmed from Helix logs across multiple hits. PR #125813 added a 120s child-process timeout but the main thread still hangs waiting for MSDTC under slow runtimes. This is a libraries-level test exercising MSDTC/OLE transaction recovery; running it under JIT stress provides no additional signal. Skip it on all non-regular CoreCLR test modes. Fixes#126304 Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Initial plan
29d2ea6

CopilotAI assigned Copilot and danmoseley Mar 20, 2026

Copilotstarted work on behalf of danmoseley March 20, 2026 00:24 View session

Fix OleTxTests.Recovery timeout by bounding remote process and test w…
3e0e74c
…all-clock time Co-authored-by: danmoseley <6385855+danmoseley@users.noreply.github.com>

CopilotAI changed the title ~~[WIP] Fix recovery timeout error in OleTx tests~~Fix OleTxTests.Recovery CI timeout by bounding remote process and test wall-clock timeMar 20, 2026

CopilotAI requested a review from danmoseley March 20, 2026 00:37

Copilotfinished work on behalf of danmoseley March 20, 2026 00:37

jkotas added the area-System.Transactions label Mar 20, 2026

danmoseley requested a review from Copilot March 20, 2026 00:50

CopilotAI reviewed Mar 20, 2026

danmoseley requested a review from Copilot March 20, 2026 01:16

CopilotAI reviewed Mar 20, 2026
View reviewed changes

Comment threadsrc/libraries/System.Transactions.Local/tests/OleTxTests.cs Outdated
Comment threadsrc/libraries/System.Transactions.Local/tests/OleTxTests.cs

Remove broken Timeout on ConditionalFact (xunit v2 only supports Time…
fc53538
…out on async methods) Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

danmoseley changed the title ~~Fix OleTxTests.Recovery CI timeout by bounding remote process and test wall-clock time~~Fix OleTxTests.Recovery CI timeout by bounding remote process wall-clock timeMar 20, 2026

Add comment explaining the TimeOut on RemoteInvokeOptions
195ffdc
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

danmoseley approved these changes Mar 20, 2026
View reviewed changes

danmoseley marked this pull request as ready for review March 20, 2026 02:19

CopilotAI review requested due to automatic review settings March 20, 2026 02:19

Copilotstarted reviewing on behalf of danmoseley March 20, 2026 02:20 View session

danmoseley requested a review from AndriySvyryd March 20, 2026 02:21

CopilotAI reviewed Mar 20, 2026

build-analysisBot mentioned this pull request Mar 20, 2026
RequestSendingResponseDisposed_ThrowsOnServer fails with Stream aborted by peer #125806
Closed

danmoseley requested a review from Copilot March 20, 2026 18:39

Copilotstarted reviewing on behalf of danmoseley March 20, 2026 18:40 View session

CopilotAI reviewed Mar 20, 2026
View reviewed changes

AndriySvyryd requested a review from roji March 20, 2026 20:47

AndriySvyryd approved these changes Mar 20, 2026
View reviewed changes

danmoseley merged commit 4521881 into mainMar 20, 2026
83 of 92 checks passed

danmoseley deleted the copilot/fix-recovery-timeout-error branch March 20, 2026 21:52

dotnet-maestroBot mentioned this pull request Mar 21, 2026
[main] Source code updates from dotnet/runtime dotnet/dotnet#5588
Merged

danmoseley mentioned this pull request Apr 8, 2026
System.Transactions.Tests.OleTxTests.Recovery fails with timeout #126304
Open

danmoseley mentioned this pull request Apr 8, 2026
Add watchdog instrumentation to OleTxTests.Recovery to capture hang diagnostics #126666
Merged

github-actionsBot locked and limited conversation to collaborators Apr 20, 2026