### What problem does this PR solve?

This PR changes Redis to be a statefulset. In some situation when we
Redis pod gets rescheduled to another Node, it gets stuck in pending
state due to the PVC attached to another Kubernetes node.

### Type of change

- [ ] Bug Fix (non-breaking change which fixes an issue)
- [ ] New Feature (non-breaking change which adds functionality)
- [ ] Documentation Update
- [X] Refactoring
- [ ] Performance Improvement
- [ ] Other (please describe):

### What problem does this PR solve?

Fix Helm Ingress template; Trying to access a global variable within a
loop
Fix #6191

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)
- [ ] New Feature (non-breaking change which adds functionality)
- [ ] Documentation Update
- [ ] Refactoring
- [ ] Performance Improvement
- [ ] Other (please describe):

fix redis pvc in helm deployment

### What problem does this PR solve?

_Briefly describe what this PR aims to solve. Include background context
that will help reviewers understand the purpose of the PR._

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

### What problem does this PR solve?

Adds a new Kubernetes Service resource to the Helm chart which
specifically targets the RAGFlow API. This feature useful for cases
where you want to expose the RAGFlow HTTP API separately from the web
interface, for example if RAGFlow is running behind an authenticating
proxy it allows a route to bypass the proxy (e.g. by defining a separate
ingress resource which forwards to the separate API-only k8s service
added here) to provide RAGFlow API access. This is still secure since
API access is already authenticated by API keys inside RAGFlow itself.

### Type of change

- [X] New Feature (non-breaking change which adds functionality)

### What problem does this PR solve?

Changed elasticsearch image url to speed up image downloading. 

### Type of change

- [x] Refactoring

### What problem does this PR solve?

Address #4391 

### Type of change

- [x] Other (please describe):

### What problem does this PR solve?

The initial Helm chart implementation added in #3815 suffers from an
issue where the 5GB data volume for the SQL DB is filled up with
[binlog](https://dev.mysql.com/doc/refman/8.4/en/binary-log.html) files
after just a few days. Since the app uses a non-replicated SQL DB config
I think it makes sense to disable the binlog in the SQL DB container.
This is achieved by simply adding the required argument to the container
startup command.

### Type of change

- [X] Bug Fix (non-breaking change which fixes an issue)

### What problem does this PR solve?

Add's a Helm chart for deploying RAGFlow on Kubernetes. 

Closes #864. 

### Type of change

- [X] New Feature (non-breaking change which adds functionality)