Sergey Vasilyev

Interesting. At the moment, I would assume that it is intended so that kubernetes does not inject the namespace before the pod is really created. And asmission-to-creation happens before the...

Hello. Sorry for the delayed review — I was a bit busy with work & life in the past months. I generally like the idea that daemons & timers can...

Hello. This is the default behaviour of the deletion handlers, as of any other handlers if `kopf.TemporaryError` is raised. The exact delay can be passed as `delay=...` to the exception....

@kvendingoldo No updates so far. As I understood, this already can be done — see https://github.com/nolar/kopf/issues/289#issuecomment-732685543 — it is just not documented explicitly. Is anything missing? (I do not use...

Hello. Thanks for the proposal! Conceptually, I'm totally "for" that feature. The API/DSL needs some thorough discussion though — before putting it into the public promises of the framework. Conditions...

Yes, that is a known issue. Sadly, it is difficult to fix. Several (2 or 3) attempts were made in the past, all failed (https://github.com/zalando-incubator/kopf/issues/279#issuecomment-570163497). My comment from there: >...

Might it be explained by this note (https://kopf.readthedocs.io/en/stable/filters/#field-filters)? > Due to a special nature of the update handlers (`@on.update`, `@on.field`), described in a note below, this filtering semantics is extended...

Also, can you run the same experiment with the operator being deployed to the cluster (where it would be as fast as possible talking to the API)? I wonder if...

Hm. That is as if the status.succeeded is either not yielded asan “extra_field” (highly unlikely), or the extra fields are not cherry-picked in the diffbase.build() — which would also be...

Then it looks like a bug in the logic. I’ll be able to check it only on this weekend (if I will be reconnected to the internet by the provider...