[CH] Optimization for multi joins on the same keys

[lgbo-ustc]

Backend

CH (ClickHouse)

Bug description

We have met a case as following

select * from t1 left join t2 on t1.key = t2.key 
 left join t3 on t1.key = t3.key 
 left join t4 on t1.key = t4.key
 ....

This will put multi joins in one execution stage, it's easy to cause OOM, since each join operator consume a lot of memory. It's easy to alloc memory for each join properly.

A related issue #8003

Spark version

None

Spark configurations

No response

System information

No response

Relevant logs

No response

[lgbo-ustc]

We may have other ways to implete this but not joins