Repository navigation
illink - Task host node exited prematurely #92951
Description
Build Information
Build: https://dev.azure.com/dnceng-public/public/_build/results?buildId=427703
Build error leg or test failing: Build / CoreCLR Product Build windows x64 checked / Build managed product components and packages
Pull request: #90110
Error Message
Fill the error message using step by step known issues guidance.
{
"ErrorMessage": "",
"ErrorPattern": "illink\\.targets\\(.*\\): error MSB4217: Task host node exited prematurely\\. Diagnostic information may be found in files in the temporary files directory named MSBuild_\\*\\.failure\\.txt",
"BuildRetry": true,
"ExcludeConsoleLog": false
}Known issue validation
Build: 🔎 https://dev.azure.com/dnceng-public/public/_build/results?buildId=427703
Error message validated: illink\.targets\(.*\): error MSB4217: Task host node exited prematurely\. Diagnostic information may be found in files in the temporary files directory named MSBuild_\*\.failure\.txt
Result validation: ✅ Known issue matched with the provided build.
Validation performed at: 10/5/2023 9:42:24 PM UTC
Report
Summary
| 24-Hour Hit Count | 7-Day Hit Count | 1-Month Count |
|---|---|---|
| 0 | 0 | 0 |
Activity
[-]Task host node exited prematurely[/-][+]illink - Task host node exited prematurely[/+]MichalStrehovsky commentedon Oct 5, 2023
Possibly related to #92290.
ViktorHofer commentedon Oct 5, 2023
#92290 is about child nodes exiting prematurely. This one is about the TaskHostFactory feature that we now use in illink.targets which has a slightly different error message.
@ericstj I can't get the errorPattern to match. Any idea whom to contact to fix that?
MichalStrehovsky commentedon Oct 5, 2023
I read both as "msbuild weirdly crashes". We should keep both bugs open but whoever investigates should be aware of possible duplicated effort because both are weird msbuild crashes that showed up around the same time.
ViktorHofer commentedon Oct 5, 2023
Agreed. I already reached out to @rainersigwald offline. He is aware but needs more information to debug the msbuild child node and task host issues.
ericstj commentedon Oct 5, 2023
@riarenas can you help us figure out why build analysis isn't linking up this issue with build?
You can see the failure it's trying to match in attempt 1 of https://dev.azure.com/dnceng-public/public/_build/results?buildId=425842&view=logs&j=39a07da7-60a1-56aa-621d-b74991bb2f3d&s=6884a131-87da-5381-61f3-d7acc3b91d76&t=577d5284-495b-5810-e593-0c14b147fbab&l=89
ericstj commentedon Oct 5, 2023
I think for these build failures it would be good to try to match previous attempts, since in these cases engineers will often rerun the build -- since it's not OK to merge with a build failure. I don't see it as a high pri fix though, but one to have on the backlog.
danmoseley commentedon Oct 6, 2023
Can we somehow redirect, or copy after the fact, files merging this pattern into either the log or into the artifacts folder that will suck them into helix so we can see the stack?
ViktorHofer commentedon Oct 20, 2023
32 hits so far (since so far), like about 60 per month
cc @rainersigwald @rokonec
rainersigwald commentedon Oct 20, 2023
Yes, the environment variable
MSBUILDDEBUGPATHcontrols where we try to put those last-chance exception handler stack dumps. However, since we try to read them from the entrypoint node when the process that crashes is a worker node, I don't think that'd help in this case.jeffschwMSFT commentedon Feb 16, 2024
removing blocking-clean-ci as it has not been hit in 30 days
MichalStrehovsky commentedon Feb 20, 2024
I think we can close this. It has not been hit since October, not just the last 30 days. I don't see anything actionable.