chore: Revisit Nightly CI Scheduled Run Frequency #1122
Labels
ci
Issues pertaining to CI / Pipelines / Testing
Comments
5 tasks
mjnagel
added a commit
that referenced
this issue
Dec 17, 2024
## Description Switches nightly CI to weekly, but also adds `milestoned` trigger to run this CI pre-release. To date we have seen minimal issues with uds-core in these environments that have actually been caused by core issues. Reducing this CI to weekly + before release will still ensure that we can catch issues, but also optimize our time/cost. I also added a modification to the cluster wait script to wait on jobs. Some of our recent CI has failed due to metallb not being installed completely yet (it is installed via a rancher `helmchart` CR, which creates a job). Example failure this should resolve: https://github.com/defenseunicorns/uds-core/actions/runs/12327456734/job/34409486515 ## Related Issue Fixes #1122 ## Type of change - [ ] Bug fix (non-breaking change which fixes an issue) - [ ] New feature (non-breaking change which adds functionality) - [x] Other (security config, docs update, etc) ## Checklist before merging - [x] Test, docs, adr added or updated as needed - [x] [Contributor Guide](https://github.com/defenseunicorns/uds-template-capability/blob/main/CONTRIBUTING.md) followed
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Describe what should be investigated or refactored
Currently, our nightly CI runs 3 jobs per 3 different k8s distros on a nightly basis. These workflows take anywhere from 30-80min to run and incur costs. Failures to date have been related to infrastructure specific misconfigurations or failures related to the runner or external dependencies (timeouts, rate limits, etc). Given the consistent failures unrelated to uds-core, there has not been a ton of value added from nightly CI runs. @mjnagel and I suggest pivoting to running 2/3x weekly and before releases as an alternative.
Links to any relevant code
Nightly CI workflows under
.github/workflows.Additional context
It's expensive. We should optimize our CI strategy to ensure we're measuring the correct results. Relates to #777, #784, and #1121.
The text was updated successfully, but these errors were encountered: