Skip to content

ci: ignore juju/4 timeouts - #1998

Merged
dimaqq merged 2 commits into
canonical:mainfrom
dimaqq:ci-continue-on-error
Aug 20, 2025
Merged

dimaqq merged 2 commits into
canonical:mainfrom
dimaqq:ci-continue-on-error

Conversation

@dimaqq

@dimaqq dimaqq commented Aug 19, 2025

Copy link
Copy Markdown
Contributor

Ignore Juju 4/beta timeouts on microk8s, as those seem to happen all the time.

@dimaqq

dimaqq commented Aug 19, 2025

Copy link
Copy Markdown
Contributor Author

matrix:
juju-channel: ['2.9/stable', '3/stable', '4/beta']
preset: ['machine', 'microk8s']
continue-on-error: ${{ matrix.juju-channel == '4/beta' && matrix.preset == 'microk8s' }}

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Am I understanding this correctly? Since we use fail-fast: false, this doesn't change whether other jobs run, but it will show Juju4+microk8s as successful even if there are errors.

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think it means that Juju4 + microk8s will still be visible as a failure (good!), but the workflow doesn't care whether that particular job was a success or failure -- the workflow can still get a green tick even if Juju4 + microk8s fails.

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Do we still get an email about the failure? I do think that's nice, so that each month we can check to see what's up - if it was just an expected issue with Juju 4 that's fine (although worth knowing for when we talk about 4 with charmers, although if there are indeed no more betas then perhaps this is all moot), but if it fails for some other reason we'd want to look into that?

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I don't know about the email.
Individual job is marked as failed.
The total job is marked as passed.

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Ok, happy to approve and figure it out next month if needed :)

@dimaqq
dimaqq merged commit 892c44a into canonical:main Aug 20, 2025
@dimaqq
dimaqq deleted the ci-continue-on-error branch August 20, 2025 02:52
tonyandrewmeyer added a commit that referenced this pull request Oct 22, 2026
Backport of #1998, which was missed from the original cherry-pick list.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
tonyandrewmeyer added a commit that referenced this pull request Oct 22, 2026
Backport of the move from pytest-operator and python-libjuju to
Jubilant, so that the 2.23 maintenance branch no longer depends on
either.

Cherry-picked from main, in order:
 * #1958 (smoke tests to jubilant-backports)
 * #1963 (drop the `juju` dependency and add Juju 4 to the smoke matrix)
 * #1987 (httpbin example charm)
 * #1988 (zero-to-hero tutorial and the k8s example charms)
 * #2000 and #2004 (how-to guides and the testing explanation)
 * #1998 (ignore Juju 4 on microk8s timeouts in the smoke matrix)
 * #2190 (Juju 4/stable rather than 4/beta in the smoke matrix)

Adjustments for the 2.x branch:

- The Python 3.8 and 3.9 environment markers that #1958 removed from
`pyproject.toml` are kept, as is the 20.04 base in the smoke test
parametrisation.
- The smoke tox env injects `ops-2.*.whl` rather than `ops-3.*.whl`.
- `uv.lock` was regenerated rather than merged, which drops
pytest-operator, python-libjuju, pytest-asyncio and their transitive
dependencies.
- The Jubilant intersphinx entry is in `docs/custom_conf.py`, since this
branch still has the older starter-pack layout.
- The tutorial's "review the final code" links stay on the
juju-sdk-tutorial-k8s repository, which is what the other chapters on
this branch link to.

---------

Co-authored-by: Dima Tisnek <dima.tisnek@canonical.com>
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants