.net - Can different RDBMS have impact on ORM performance test results?

Question

I made some ORM framework performance tests on MSSQL (EF,NH in comparrison with pure ADO.NET approach). EF, NH and ADO.NET should use same ADO.NET provider for DB. My question is, if its possible that EF could have some advantage over NH on MSSQL.

Therefore I'm asking if it is possible that using different DB (MySql for example), could have relevant impact on ORM performance test results.

(by relevant I mean that ORM X has improved its performance by x% and ORM Y by y% where x != y :) )

Answer 1

绝对地。例如，MySQL 没有散列连接和合并连接。某些查询模式在那里会受到很大影响。

RDBMS 使用无关紧要的概念对我来说很奇怪。不同的数据库具有非常不统一的性能特征。性能随工作负载而变化很大。

如果您想从您的测试中尽可能多地排除 RDBMS，您必须仔细确保您不会意外测试任何 RDBMS 特定的性能特性。这很难。

Answer 2

当然。即使是同一个 RDBMS 上的不同设置也会导致不同的测试结果。