operator
operator copied to clipboard
canonical
Pure Python framework for writing Juju charms
Okay. I turned on branch-protection rules for any `*-maintenance` branches. I saved the last cherry-pick for 1.5.1 to put in this PR in order to test/check that CI works right...
The conspicuous absence of leader-settings-changed event emission from the harness was noted recently by @ca-scribner et al. Notably, the harness already supports changing the leadership status of the charm being...
Stderr output emitted during an action hook execution gets injected in an action's results by juju. - documented this in the ActionEvent docstring - added a check that users don't...
Charm authors can mark some of the relations as optional using 'optional' attribute in relations section (provides/requires/peers) in metadata.yaml [1], however the parameter is not saved in RelationMeta class. The...
I am writing a charm for a stateful application which has clustering built-in. For the complete application removal case I need to avoid removing members from the actual cluster while...
This PR adds a convenience method to harness for simulating a charm upgrade ([example](https://github.com/canonical/prometheus-k8s-operator/blob/7ae9e8fe139323463bc40e68d91d96cde29c0caf/tests/unit/test_remote_write.py#L271)). Manually emitting `upgrade_charm` could probably be enough to cover most of the need, but having all...
Is this the/an approach (just growing our charm testing list with "stable/good" charms) we want to take to help avoid regressions and breakage downstream when we do releases? I feel...
This is a theoretical problem that I haven't actually tested out. But it seems like it is possible for a hook observation call - for example `self.framework.observe(...)` to be run...
More specifically, before every hook execution. Point is: when a 'real' charm runs, it is reinitialized afresh. While if you do fire eventA and eventB on the harness, the harness...