[Feature] It is recommended not to start multiple fe or multiple be with one statefulset

### Search before asking

- [x] I had searched in the [issues](https://github.com/apache/doris-operator/issues?q=is%3Aissue) and found no similar issues.

### Description

Problem description:
It is relatively simple to start a pod with statefulset, but it is also very inflexible. Statefulset always starts rolling updates from the last pod, and one of the pods cannot be stopped separately. When one of the fe or be has a problem, it cannot be stopped separately.

Expected results:
It is recommended to change to one deployment to maintain one fe or be. If there are three fes, start three single-copy deployments. If there are 6 bes, start 6 single-copy deployments. This is more flexible and can control each fe or be separately, which is very suitable for some scenarios that require manual processing. Of course, it is also possible to change to a single-copy statefulset for fqdn. In combination with webhook control, the number of statefulset copies cannot be modified.
The rook-ceph operator also uses this method to maintain the ceph cluster.

### Use case

_No response_

### Related issues

_No response_

### Are you willing to submit PR?

- [x] Yes I am willing to submit a PR!

### Code of Conduct

- [x] I agree to follow this project's [Code of Conduct](https://www.apache.org/foundation/policies/conduct)