Unable to upgrade operator

[JuanFCabrera]

when using the upgrade command i get the following. kubectl apply -f https://github.com/crowdstrike/falcon-operator/releases/latest/download/falcon-operator.yaml namespace/falcon-operator unchanged customresourcedefinition.apiextensions.k8s.io/falconcontainers.falcon.crowdstrike.com unchanged customresourcedefinition.apiextensions.k8s.io/falconnodesensors.falcon.crowdstrike.com configured serviceaccount/falcon-operator-controller-manager unchanged role.rbac.authorization.k8s.io/falcon-operator-leader-election-role unchanged clusterrole.rbac.authorization.k8s.io/falcon-operator-admission-controller-role unchanged clusterrole.rbac.authorization.k8s.io/falcon-operator-container-role unchanged clusterrole.rbac.authorization.k8s.io/falcon-operator-manager-role unchanged clusterrole.rbac.authorization.k8s.io/falcon-operator-node-sensor-role configured rolebinding.rbac.authorization.k8s.io/falcon-operator-leader-election-rolebinding unchanged clusterrolebinding.rbac.authorization.k8s.io/falcon-operator-manager-rolebinding unchanged configmap/falcon-operator-manager-config unchanged deployment.apps/falcon-operator-controller-manager configured The CustomResourceDefinition "falconadmissions.falcon.crowdstrike.com" is invalid: spec.scope: Invalid value: "Cluster": field is immutable

installation of the operator was performed with kubectl apply -f https://github.com/crowdstrike/falcon-operator/releases/latest/download/falcon-operator.yaml

installation of the node sensor uses:

apiVersion: falcon.crowdstrike.com/v1alpha1 kind: FalconNodeSensor metadata: name: falcon-node-sensor spec: falcon: tags:

• daemonset trace: none falcon_api: client_id: $CLIENT_ID client_secret: $CLIENT_SECRET cloud_region: autodiscover node: tolerations:
• key: "VMtype" operator: "Equal" value: "Inf1" effect: "NoSchedule"

I have found that uninstalling the operator and node sensor, then reinstalling works fine and installs latest version.

[redhatrises]

Hello,

Can you please provide your k8s cluster distro? If you are trying to deploy via OpenShift (which the error message leads me to believe), you should be deploying directly from operatorhub in OpenShift.

[JuanFCabrera]

i am deploying in Tanzu Kubernetes, (tanzu managed EKS) we used the generic kubernetes instructions

[redhatrises]

Got it. Thanks. Currently, upgrades are unsupported in the 0.9.x series. This will change with the impending 1.0 release. That said, you might try deleting the CRD falconadmissions.falcon.crowdstrike.com and trying again.

[JuanFCabrera]

does this CRD have a finalizer? i was able to delete 2/3 but the last one just hangs.

[JuanFCabrera]

one last question, after 1.0 release will the upgrade continue to a manual deployment or will it use latest version anytime a pod is created/redeployed.

[redhatrises]

Is there already an existing deployment for the CRD that just hangs?

[JuanFCabrera]

yes, there is an existing deployment where it hangs

[redhatrises]

What CRD is hanging?

[JuanFCabrera]

falconnodesensors.falcon.crowdstrike.com is hanging. normally i have 3 of these CRD. two were deleted but the last one hangs

[redhatrises]

@JuanFCabrera when uninstalling, make sure to delete all the deployed CRs before deleting everything else. Try to wait for the falconnodesensor cleanup job to complete. If it doesn't, deleting the finalizer should work, and then you would delete the renaming contents.

[redhatrises]

Closing as there isn't further movement on this, please re-open (it would be better to open a CrowdStrike support case) if necessary.