feat(batch): support BATCH_PARALLELISM

[fuyufjh]

Should the batch query jobs i.e. SELECT follow this parallelism?

Do you mean default parallelism of the non-scan stages? Good catch. 🤔

Yes, exactly.

Maybe we need another BATCH_PARALLELISM for this. 😄

Originally posted by @BugenZhao in #7370 (comment)

Currently, the parallelism of each stage in distributed batch query is determined as follows:

1. Scan stage has the same number of parallelism as the number of partitions of table/mv.
2. Root stage's parallelism is always one.
3. Intermediate stage's parallelism is the same as the number of worker nodes.

Sometimes users may want to increase the number of parallelism stages, and we can have a session variable batch_parallelism for this.

[ZENOTME]

closed by #8552