README
¶
claimedresourcesnamespaces
When using claimed resources, it can be situations where resource does not yet exists in the consumer cluster OR
we are operating in cluster scoped mode, and hence no resource namespace is created. In scenarios like this,
APIServiceNamespace object is never creates, and hence backend will not create appropriate RBAC for the consumer to be able to
access the resources.
It will watch same GVR resources, as claimed resources (hence the name of the controller), and will create APIServiceNamespace objects,
if object is created in the consumer cluster.
We can't add this logic into claimedresource controller, as it never gets to healthy state until provider has right rbac configured and
consumer can access the resources and hence start the informers. Alternative to this is create APIServiceNamespace objects outside reconile
loop, but that would be against the controller pattern.
Documentation
¶
Index ¶
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
func NewController ¶
func NewController( gvr schema.GroupVersionResource, claim kubebindv1alpha2.PermissionClaim, apiServiceExport *kubebindv1alpha2.APIServiceExport, providerNamespace string, providerConfig *rest.Config, consumerConfig *rest.Config, consumerDynamicInformer informers.GenericInformer, serviceNamespaceInformer dynamic.Informer[bindlisters.APIServiceNamespaceLister], claimedResourcesEnqueueChan chan<- string, ) (*controller, error)
NewController returns a new controller ensuring APIServiceNamespaces exist for claimed objects.
Types ¶
This section is empty.
Source Files