Troubleshoot the OpenStack Control Plane
-
Check the pods status
oc get pods -n openstack
-
If you see any of the pods staying in
pending
state for logn check pod’s logs.Example:oc logs openstack-galera-0
-
For further investigation you may describe the pod in question to see any hint about the problem there.
Example:oc describe pod openstack-galera-0
In the output of the abobe command, scroll down at the bottom to check the events generated by the pod.
-
You may alos try attaching to the pod’s shell for further investigation on any pod related issue.
Example:oc rsh openstack-galera-0
-
Check running jobs for openstack control plane deployment.
oc get jobs
Sample outputNAME COMPLETIONS DURATION AGE glance-db-create 1/1 6s 66s glance-db-sync 1/1 22s 56s keystone-bootstrap 0/1 1s 1s keystone-db-create 1/1 6s 37s keystone-db-sync 1/1 16s 27s neutron-db-create 1/1 4s 38s
-
To track the progress or troubleshoot, you may check the logs for the job which is not completed.
Example:oc logs job/keystone-bootstrap
You may need to re-deploy the control plane if you identify any problem with your configuration.
-
Delete the current deployment
-
Make desired changed in the deployment configuration file.
-
Make sure all the resources of the deployment are cleaned up.
oc get openstackcontrolplane -n openstack oc get all -n openstack oc get pods -n openstack oc get pvc -n openstack
-
Delete any of the resource left back after deletion of controlplane configuration.
-
Re-run deployment by applying configuration
-
Check again for the status of openstack control plane deployment.
oc get openstackcontrolplane -n openstack
-
Repeat the query until MESSAGE is Setup complete
Sample OutputNAME STATUS MESSAGE openstack-galera-network-isolation True Setup complete