New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Make the kola test workflow reusable #374
Conversation
this initial attempt runs right after the "packages" jobs and downloads the resulting artifacts. The QEMU image is unzipped then the kola test is running using the ci-automation scripts. Signed-off-by: Mathieu Tortuyaux <mtortuyaux@microsoft.com>
af0b1e1
to
e7fa8e4
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Looks good, thanks!
For cherry-picking the changes in ci-automation/*
should be done in a separate commit.
I'll look into the kola VM profile RAM situation - why was 8 GB swap chosen?
A little over what 3 VMs of 2.5 GB memory would consume. It gives a chance for tests to finish before an out of memory condition is reached. Hitting swap will slow down newly spawned tests to allow already spawned tests to finish and free up memory. Without swap, some VMs will be killed when memory is exhausted. Will split the PR in 2 commits. |
* Add SKIP_COPY_TO_BINCACHE environment variable that will skip uploading test results to bincache. This is useful if we want to upload test results as artifacts on github. * make QEMU_IMAGE_NAME configurable Signed-off-by: Gabriel Adrian Samfira <gsamfira@cloudbasesolutions.com>
e7fa8e4
to
e7fe3a4
Compare
* Adds a reusable workflow that can run tests * Adapts the ci.yaml to use reusable workflow * Creates a new workflow that helps trigger tests using an arbitrary workflow run. Signed-off-by: Gabriel Adrian Samfira <gsamfira@cloudbasesolutions.com>
e7fe3a4
to
879e5a2
Compare
Remove arch and parallel options. Hardcode them for now.
Signed-off-by: Gabriel Adrian Samfira <gsamfira@cloudbasesolutions.com>
ea3948a
to
667a6cc
Compare
Signed-off-by: Gabriel Adrian Samfira <gsamfira@cloudbasesolutions.com>
* arm64 tests now run in LXD containers instead of VMs * parallelism increased for arm64 * some setup steps are skipped on arm64 Signed-off-by: Gabriel Adrian Samfira <gsamfira@cloudbasesolutions.com>
667a6cc
to
340fd39
Compare
A fix for the failed test is prepared in flatcar/bootengine#47 I just realize that we even should backport the .github changes (and before that the previous PR), otherwise the a PR against an older scripts branch like |
Yes. The workflows would need to be backported to branches where you push PRs to as well as far as I remember. |
Make the kola test workflow reusable
Make the kola test workflow reusable
Make the kola test workflow reusable
Make the kola test workflow reusable
He, I was a bit late with picking #356 and it already started the workflow builds for all branches :D |
no worries. They can easily be canceled in the "Actions" tab of the repository. :D |
….20210914 net-vpn/wireguard-tools: Move to portage-stable; Sync with Gentoo; Updates to wireguard-tools-1.0.20210914
In this PR, the artifact download action is from my own fork, because it includes a fix that streams the artifact directly to disk. The upstream action downloads to memory and also un-archives the artifact in memory, which leads to huge memory usage in our case. I created a PR for this with the upstream project here:
dawidd6/action-download-artifact#175
Signed-off-by: Gabriel Adrian Samfira gsamfira@cloudbasesolutions.com