deploy strimzi via helm without cluster-level permissions

[andreaprslp]

Hi all, we are trying to deploy strimzi via helm in an environment where we cannot change permissions at the cluster level.
In a few words in our case strimzi only needs to manage one cluster kafka and kafkaconnect within the same namespace.
In the helm/templates dir there are many files with the same names and it is not very clear to us which of them we need to use/edit.
Are there any example files for such a configuration?
Can you please help us?

[im-konge]

Hi, you can specify the Namespaces that CO should watch using the watchNamespace field in the values.yaml
More info is in the README - https://github.com/strimzi/strimzi-kafka-operator/tree/main/helm-charts/helm3/strimzi-kafka-operator#configuration

[andreaprslp]

is it necessary to specify the namespace even if the CO is within the same namesapce?

[im-konge]

By default it watches just the one Namespace, so no, if you don't want any other Namespace to watch, it should be enough to just install it in the particular Namespace and CO will watch just the one.

[scholzj]

To start, you should probably explain what is cluster-wide permission. Please keep in mind that some of the key features for production ready Kafka (such as rack awareness for example) require cluster-level permissions such as listing nodes. Also, given the nature of CRDs and ClusterRoles, you always need to create resources outside of the namespace you plan to use.

[andreaprslp]

the environment in which I want to deploy strimzi and kafka is an enterprise environment, and I do not have permissions to release objects that control the node list or rack status. I only need a strimzi installation to manage a kafka

[scholzj]

I only need a strimzi installation to manage a kafka

But these things are a very important part of the Kafka management.