In RHOSO18 deployment guide, several service's configuration has the following comment, (# backend needs to be configured to activate the service).
~~~
cinder:
apiOverride:
route: {}
template:
databaseInstance: openstack
secret: osp-secret
cinderAPI:
replicas: 3
override:
service:
internal:
metadata:
annotations:
metallb.universe.tf/address-pool: internalapi
metallb.universe.tf/allow-shared-ip: internalapi
metallb.universe.tf/loadBalancerIPs: 172.17.0.80
spec:
type: LoadBalancer
cinderScheduler:
replicas: 1
cinderBackup:
networkAttachments:

• storage
  replicas: 0 # backend needs to be configured to activate the service
  cinderVolumes:
  volume1:
  networkAttachments:
• storage
  replicas: 0 # backend needs to be configured to activate the service
  ~~~

Glance has the same comment.
~~~
glance:
apiOverrides:
default:
route: {}
template:
databaseInstance: openstack
storage:
storageRequest: 10G
secret: osp-secret
keystoneEndpoint: default
glanceAPIs:
default:
replicas: 0 # backend needs to be configured to activate the service
override:
service:
internal:
metadata:
annotations:
metallb.universe.tf/address-pool: internalapi
metallb.universe.tf/allow-shared-ip: internalapi
metallb.universe.tf/loadBalancerIPs: 172.17.0.80
spec:
type: LoadBalancer
networkAttachments:

• storage
  ~~~

I suppose those comments are safeguard for provisioning services with insufficient configurations.

On the other hand, if the user can provide the sufficient configuration, how many replicas are recommended?

For cinder backup will be 3+ and cinder volume is 1 per container.
For glance will be 3.

We should have a recommended replica number in the comments.