Closed archana203 closed 1 year ago
I see that cinderAPI
is defined twice, I will fix it now, but I believe this is utimately caused by a recently introduced bug that should be fixed by https://github.com/openstack-k8s-operators/cinder-operator/pull/250
For the time being a workaround is to move the customServiceConfig
from the API to be global:
cinder:
enabled: true
template:
databaseInstance: openstack
secret: osp-secret
customServiceConfig: |
[DEFAULT]
default_volume_type=tripleo
cinderAPI:
externalEndpoints:
- endpoint: internal
ipAddressPool: internalapi
loadBalancerIPs:
- 172.17.0.80
replicas: 1
cinderScheduler:
replicas: 1
...
We have merged PR#250 in the cinder-operator, so this bug should have been fixed.
During cinder adoption we are passing default_volume_type=tripleo in customServiceConfig section of CinderApi service but the configuration is not reflected in cinder service pods and volume creation failed with following message :-
[ospng@dell-r640-089 osp2]$ openstack volume create --image cirros --bootable --size 1 disk_new The request cannot be fulfilled as the default volume type DEFAULT cannot be found. (HTTP 500) (Request-ID: req-621e986d-7220-474b-aaa2-5282a49f866c) command terminated with exit code 1