-
-
Notifications
You must be signed in to change notification settings - Fork 166
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Resources filtered out in event handlers still get annotated by Kopf #374
Comments
Same problem here with This unfortunately renders kopf almost unusable for us as it means that kopf demands 'patch' access to basically all resources even if they are only watched and RBAC could be restricted otherwise (e.g. to get, list, watch). Would be very glad if this could be addressed soon. |
Reading through all this again the RBAC related problems are probably caused by me using kopf wrongly. |
@asteven Don't worry, this is not the noise. Indeed, if the object is filtered out, Kopf should be completely blind to it, including the annotations — i.e. not to put them. This behaviour is especially important with, for example, namespace filtering: if only the namespaces "teamname-*-pr-featurename" are configured for filter via the This should be easily fixable, I hope. I'll take a look now. |
@nolar Unfortunately, this is still a problem for me. I have a
The handler looks like this:
Any idea what I might be doing wrong? |
Long story short
Resources that are indicated to be of no interest to an event handler through a filter (like label=) still get annotated with Kopf data like
/last-handled-configuration
, across the entire cluster.Description
I have an Operator that watches Pod events with specific label values, so the event handlers use the label= filter. However, whenever the Operator runs all pods in all namespaces get updated with annotations for persisted progress and diffbase data, instead of just the filtered pods.
Coupled with issue #372, this could be disruptive to a cluster, since it's possible that unrelated operators can no longer add annotations to the pods they're managing, because the resource size grew too big with data from unrelated operators.
The code snippet to reproduce the issue
The exact command to reproduce the issue
The full output of the command that failed
Environment
Python packages installed
The text was updated successfully, but these errors were encountered: