[Bug] Regression of Bug 381

[andrewm-aero]

Describe the bug
Default rolling deployment strategy uses maxSurge and maxUnavailable of 25, not 25%, leading to downtime with any fewer than 26 replicas. This appears to be a regression of #381.

Version
docker.io/bitnami/grafana-operator:4.7.1-debian-11-r0

To Reproduce
Steps to reproduce the behavior:

1. Deploy a Grafana resource with 25 or fewer replicas
2. Wait for deployment to become healthy
3. Modify the Grafana resource in a way that triggers an update to the pods
4. Note how all pods are terminated immediately, resulting in downtime

Expected behavior
At least one pod to remain alive at all times.

Suspect component/Location where the bug might be occurring
Here

Screenshots
N/A

Runtime (please complete the following information):

N/A (but can be provided if needed)

Additional context
N/A

[pb82]

It looks like this has either never been fixed on master or was fixed and then changed back: https://github.com/grafana-operator/grafana-operator/blob/master/controllers/model/grafanaDeployment.go#L152

We're still setting 25 (as a number) instead of 25%