软件开发Db2
db2 fetch first查询结果每次是否一致？
想问一下如果每天有100w数据，把 insert into 历史表 select * from 当天表 fetch first N rows only 和delete from 当前表 fetch fist N rows only 放在一个事务里每次从当前表转移10w数据到历史表循环转移10次，没加 order by，每次insert into 和...显示全部
查看其它 1 个回答leo_wyn的回答
来自官方文档的解释：
https://www.ibm.com/docs/en/db2-for-zos/12?topic=subselect-fetch-clause
Determining a predictable set of rows to retrieve requires the specification of an ORDER BY clause with sort keys that would uniquely identify the sort order of each row in the intermediate result table. If the intermediate result table includes duplicate sort keys for some rows, the order of these rows is not deterministic. If there is no ORDER BY clause, the intermediate result table is not in a deterministic order. If the order of the intermediate result table is not deterministic, the set of rows retrieved is unpredictable. If both the fetch-clause and the ORDER BY clause are specified, the fetch-clause is processed on the ordered data.