Users can create invalid ClusterCSIDriver instances: apiVersion: operator.openshift.io/v1 kind: ClusterCSIDriver metadata: name: manila.csi.openstack.org spec: driverConfig: driverName: ebs.csi.aws.com logLevel: Normal managementState: Managed operatorLogLevel: Normal Note that the CR name does not correspond to `spec.driverConfig.driverName`. CSO will always create valid instances, however, we should block users from breaking them via validation.
We'll going to fix it by validation in the operators, not by webhook. I.e. the operator sets a condition (Degraded?) that the CR is not valid and refuse to operate on it.
It would be even better to remove ClusterCSIDriver.Spec.DriverConfig, as it does not bring anything useful (yet). We may always introduce it later.
Attempt to remove ClusterCSIDriver.Spec.DriverConfig: https://github.com/openshift/api/pull/718
The fixed is not included in 4.6.0-0.nightly-2020-08-18-165040, waiting for a new version.
verified with: 4.6.0-0.nightly-2020-08-20-234448
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory (OpenShift Container Platform 4.6 GA Images), and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2020:4196