[FEATURE] Skip some servers if having multi servers for one partition

### Code of Conduct

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

### Search before asking

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

### Describe the feature

If the partition reassign is activated, the single partition's data will be stored in the multi servers, that will make the client read side to retrieve these data from multi servers.

When the spark AQE is enabled and this job stage is optimized by the AQE rules, the task will skip the partial data from the mapId. For the LOAD_BALANCE mode of partition split, some mapIds data doesn't exist in some servers, we could skip it to speed up reading.

### Motivation

_No response_

### Describe the solution

_No response_

### Additional context

_No response_

### Are you willing to submit PR?

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

Start by tracing the client read path for partition reassignment and the LOAD_BALANCE mode, then inspect how Spark AQE affects mapId reads. Done means the client can skip servers where the requested mapId data is absent while still retrieving all available partition data correctly.

Written by the indexing model from the issue text.