kopf VS Celery-Kubernetes-Operator

If you're looking for more examples beyond what was shown and referenced above, I recommend exploring other popular tools that make use Python Kubernetes client, such kopf - the library for creating Kubernetes operators. I also find it very useful to take a look at tests of the library itself, as it showcases its intended usage such this client test suite.

Depending on how you want to configure your selecting logic, it can be solved by mutating admission webhooks for the pods. For example, in Kopf, the simplest approach would be:

Hello. Kopf (a framework to write Kubernetes operators in Python) 1.31 is released and has finally got admission webhooks — https://github.com/nolar/kopf/releases/tag/1.31.0. I would appreciate some feedback from experienced operator developers on how easy or hard it is to write webhooks now, and what is missing and makes it inconvenient. The docs: https://kopf.readthedocs.io/en/stable/admission/ For a brief preview, it looks like this:

Correct, using generic resources with Client.watch should work. In general to create more complex operators I would recommend to check out kopf.

The take-out from that course is: don't deploy anything stateful on Kubernetes in production, period. Even disregarding that, don't deploy anything stateful that doesn't come in a form of an operator. For celery, https://github.com/celery/Celery-Kubernetes-Operator is a WIP, so obviously not suitable for anything.