You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Makes it really easy to implement handling for GDPR requests
Cons:
Complicates codebase if we were to keep namespace scoped users as well
Other:
Discourages use of custom resources directly by applications because applications can't assume users are cluster or namespace scoped
Paves way to NamespaceClaim, to automatically clean up namespaces/sandboxes created for the user
Paves way to ServiceMailboxClaim, so Wildduck operator can pull mailbox claims from different namespaces and set up type: service users for them in first come, first serve manner
Cluster scoped users make sense especially when Kube API is configured to authenticate with the same OIDC provider
Makes it possible to include ownerReferences in an OIDC claim so applications that create resources on users behalf have easy way to add the reference, most notably that includes the Kubernetes uid of the user
Pros:
ownerReferencesand deletion of user results in cleanup of their resourcesCons:
Other:
NamespaceClaim, to automatically clean up namespaces/sandboxes created for the userServiceMailboxClaim, so Wildduck operator can pull mailbox claims from different namespaces and set uptype: serviceusers for them in first come, first serve mannerownerReferencesin an OIDC claim so applications that create resources on users behalf have easy way to add the reference, most notably that includes the Kubernetesuidof the user