When I deploy the CR with a valid label on node_selector and after the pod is running, I changed the label to a invalid one (not existent on any node). So, the pod is terminated (as expected) and then move the a pending state as there is not elegible nodes to let it run (as expected). But if I change the CR again to a valid label and apply it, I expect it to be terminated and started again in the new label that I provided but it does not happen. The pod keeps in the pending state. So, to get it running again I need to delete the CR and create again.
As stated by dev team, this issue may be fixed in https://github.com/artemiscloud/activemq-artemis-operator/pull/184
- is related to
-
ENTMQBR-5164 Add NodeSelector functionality with AMQ 7.x broker CR.
- Closed