Deployments are not removed when the CR's are removed about elasticsearch-operator HOT 3 CLOSED

AFAIK cascade deletes have to be issued explicitly.
The operator sets owner references on its managed objects, so the following command should delete all es objects in foreground fashion.

oc delete elasticsearch my-es --cascade=true

Also, did you try to wait a bit after deleting the CR?
K8s' garbage collector might kick in after some time and recycle orphaned objects. But I'm not sure about this.

@ewolinetz what do you think?

from elasticsearch-operator.

I don't believe we need to use --cascade=true, I've just deleted the parent object which is the ClusterLogging CR object.

@jcantrill K8s does the clean up, I would verify that you have an OwnerRef in each of the deployments. It should point to the Elasticsearch CR object for that specific cluster (the OwnerRef of that Elasticsearch object would be the ClusterLogging CR so long as the CLO created it)

from elasticsearch-operator.

Closing as I don't see this happening any longer

from elasticsearch-operator.