SET NOCOUNT ON 真的会产生如此大的性能差异吗 [英] Does SET NOCOUNT ON really make that much of a performance difference

问题描述

在这篇文章中，作者建议存在与SET NOCOUNT ON 并且通过从网络中消除这种额外的开销，它可以大大提高您的数据库和应用程序的整体性能"

In this article, the author suggests that there is material overhead associated with SET NOCOUNT ON and that "By removing this extra overhead from the network it can greatly improve overall performance for your database and application"

作者引用了 2000 年到 2005 年默认存储过程模板的变化，并暗示微软甚至意识到了这个问题"，这促使了这个模板的变化.

The author references a change in the default stored procedure template from 2000 to 2005 and suggests that "Microsoft even realized the issue " which prompted the change in this template.

是否有人有确凿的证据支持或反驳通过设置 NOCOUNT ON 所声称的性能提升.

Does somebody have hard evidence that either supports or refutes the claimed performance gain with setting NOCOUNT ON.

推荐答案

我的问题中的最后一个链接:SET NOCOUNT ON usage" 指的是 关于它的文章.

The last link in my question here: "SET NOCOUNT ON usage" refers to an article on it.

考虑到它是多么微不足道，为什么不把它留在里面并停止客户端处理另一个结果集?

Given how trivial it is, why not leave it in and stop the client processing another resultset?

如果没有 SET NOCOUNT ON，nHibernate 也会崩溃(在问题中提到)

And without SET NOCOUNT ON, nHibernate can break too (mentioned in question)

这篇关于SET NOCOUNT ON 真的会产生如此大的性能差异吗的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！