node logs dumps are not saved any more

Summary

Since recent merge of !2774 (merged) node logs are not dumped any more, we can observe following log in admission controller:

2024-10-16T10:37:06Z	INFO	webhooks.resource.validate	validation/validation.go:120	validation failed	{"gvk": "/v1, Kind=Pod", "gvr": {"group":"","version":"v1","resource":"pods"}, "namespace": "sylva-system", "name": "node-debug-mgmt-1498167256-kubeadm-capm3-virt-management-md-0", "operation": "CREATE", "uid": "c236e089-ee41-4f6a-9810-41fb96c2f0ff", "user": {"username":"kubernetes-admin","groups":["system:masters","system:authenticated"]}, "roles": [], "clusterroles": ["cluster-admin", "system:basic-user", "system:discovery", "system:public-info-viewer"], "resource.gvk": "/v1, Kind=Pod", "kind": "Pod", "URLParams": "/disallow-latest-and-main-tag", "action": "validate", "resource": "sylva-system/Pod/node-debug-mgmt-1498167256-kubeadm-capm3-virt-management-md-0", "operation": "CREATE", "gvk": "/v1, Kind=Pod", "action": "Enforce", "policy": "disallow-latest-and-main-tag", "failed rules": ["require-image-tag"]}

EDIT: I concluded way too fast and suspected recent kyverno changes to prevent this pod from starting, but this pods has nothing to do with node logs, as they are directly served from the host.

Edited Oct 17, 2024 by Francois Eleouet
Assignee Loading
Time tracking Loading