Closed hyoklee closed 4 months ago
I like the idea, but we need to standardize how we do things. How long does this run? We should incorporate into the callable workflows and consistent runs. Why report all pushes to CDash? CI will stop and the developer can see the failures, so no need for CDash. I could see a daily_build reporting to CDash - too many reports of the same can just clutter the CDash page and blur the impact and/or push important reports to another page. Need to discuss a plan.
Thinking more about the reporting to CDash. I think we should create a "Nightly" workflow that runs only once a day when there have been changes the previous day. Then we can have a standard set of configurations that would run across all the platforms and report to CDash. This would identify any issues with configuration failure and also it would be easy to highlight a platform failure.
A nightly build would be easy to add an argument to the existing workflows to indicate that builds should send reports to CDash.
It utilizes my.cdash.org for test result reporting.