-
Bug
-
Resolution: Unresolved
-
Normal
-
None
-
None
-
None
I created a similar ticket already for Telemetry, it's actually the same request for all components, but just to illustrate my point:
[root@jumphost 05_control_plane]# oc get openstackcontrolplane NAME STATUS MESSAGE openstack-control-plane False OpenStackControlPlane DNS in progress [root@jumphost 05_control_plane]# oc get events LAST SEEN TYPE REASON OBJECT MESSAGE 6m22s Normal Killing pod/ceilometer-0 Stopping container ceilometer-central-agent 6m22s Normal Killing pod/ceilometer-0 Stopping container sg-core 6m22s Normal Killing pod/ceilometer-0 Stopping container ceilometer-notification-agent 6m22s Normal Killing pod/ceilometer-0 Stopping container proxy-httpd [root@jumphost 05_control_plane]#
# oc describe openstackcontrolplane | less
(...)
Dns:
Enabled: true
Template:
Container Image: registry.redhat.io/rhoso/openstack-neutron-server-rhel9@sha256:3c49822b33a4d9b05ee9946ca92923d6697c7c66787a02c69a8420ba2de94778
Dns Data Label Selector Value: dnsdata
Options:
Key: server
Values:
8.8.8.8
Key: server
Values:
8.8.4.4
Override:
Service:
Metadata:
Annotations:
metallb.universe.tf/address-pool: ctlplane
metallb.universe.tf/allow-shared-ip: ctlplane
metallb.universe.tf/loadBalancerIPs: 172.20.1.80
Spec:
Type: LoadBalancer
Replicas: 2
(...)
Message: OpenStackControlPlane DNS in progress
Reason: Requested
Severity: Info
Status: False
Type: Ready
Last Transition Time: 2024-09-08T16:34:34Z
(...)
I mentioned in the other ticket that it's an anti-pattern in OpenShift to log issues into pod errors only, and here's why:
I as a user of the OpenStack operator have no idea which one of the many operator-pods is actually responsible for driving forward the DNS progress. So I have to do something as absurd as:
# oc get pods -n openstack-operators -o name | while read p; do echo "=== $p ==="; oc logs -n openstack-operators $p | grep -i dns; done | less
.. in order to figure out that the the infra-operator-controller-manager is responsible for this. Please make sure to update the OpenStackControlPlane CR with the actual status and failures, and please emit events as well
Thanks!
