NVIDIA / NVIDIA/cccl

[FEA]: [DevEx] Parse, dedup, summarize, and format a report of CI errors per PR run

Open
#5,757 1 comment 0 reactions 1 assignee Claimed by @alliepiper View on GitHub
devex infra
Dominant language
C++
Stars
2.5k
Forks
486
Avg merge
2d 6h
Merged PRs (30d)
295

Description

### Is this a duplicate?

- [x] I confirmed there appear to be no [duplicate issues](https://github.com/NVIDIA/cccl/issues) for this request and that I agree to the [Code of Conduct](CODE_OF_CONDUCT.md)

### Area

Infrastructure

### Is your feature request related to a problem? Please describe.

It would be very useful to do some automated processing of CI results that condense and surface errors and failures.

### Describe the solution you'd like

1. Capture config/build/test logs from CI jobs as artifacts in the build jobs. Use the existing job results artifact machinery.
2. In the workflow-results action, parse error/failure messages from these logs when a job fails. Capture filename/line and error message, associate with job information.
3. Present a report that shows:

Sort project/location level alphabetically. by project name and location.
Sort error messages within a location alphabetically.
Sort job log links by job name alphabetically.

```md

🚨 Failure log

📄 [ProjectName:] path/to/some/file.ext:[lineno]

⚠️ [truncated error message]

[full error message]

- [Job Name](GHA Job Log URL)
- ... (All jobs with this location + error)

....other error messages at this location

...other project/location failures.

```

For failures without recognized locations / error messages, sort these by project and print lists of job log links.

Add this to the step summary for the verify-results job.

### Describe alternatives you've considered

_No response_

### Additional context

_No response_

Contributor guide

Open the contributing guide

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.