Question

I have algorythm where I need to join two tables , first table has 40 millions of rows and second one has 300 000. Both tables were created with same ORDER BY so while joining I use SETTINGS join_algorithm = full_sorting_merge , max_bytes_before_external_sort = 0; It looks like following:

SELECT A.* , B.*  FROM db1.table1 A  -- 40 millions of rows
left join (select * from db1.table2) B  -- 300 000 rows
on A.col1 = B.col1
SETTINGS join_algorithm =  full_sorting_merge , max_bytes_before_external_sort = 0;

现在,我需要增加表db1.table3,以便与表1合并。

SELECT A.* , B.*  FROM db1.table1 A  -- 40 millions of rows
left join (select * from db1.table2) B  -- 300 000 rows
on A.col1 = B.col1
left join (select * from db1.table3) C  -- 400 000 rows
on A.col1 = C.col1

我应说明加入algrothm的情况? 每次加入之后? 或者只是在问询结束时? 如果在问答结束时我只具体说明哪一种SETlingS在两个加入时是否得到推进? 是否有能力为每一个加入询问的人指明不同的农林? 辅助器

Answer 1

唯一可能的做法是指定一个<代码>。缩略语逐项条款,因此,如果你多次加入询问,同样的环境将传播给双方。这意味着,目前无法为某个问题确定不同的并入算法。

如果各位愿意利用不同的环境参加,那么,你们需要与各自的环境分开查询。

Answer 2

for 300k rows in right table you could use engine=Join 以及 joinGet look

https://clickhouse.com/docs/en/engines/special/join

以及

https://clickhouse.com/docs/en/sql-vis/Functions/other-Functions#joinget”rel=“nofollow noreferer”>https://clickhouse.com/docs/en/sql-vis/Functions/other-Functions#joinget

友情链接