德鲁伊 oltp oltp_内存中OLTP –更快变得更简单！

德鲁伊 oltp oltp

In-memory OLTP is a revolutionary tool introduced on SQL Server 2014. On SQL Server 2016 this will be even better, with a broader surface of supported tools. Check this article to know the news.

内存中的OLTP是SQL Server 2014上引入的一种革命性工具。在SQL Server 2016上，它会变得更好，并具有广泛的受支持工具。查看本文以了解新闻。

During the Microsoft Ignite conference, held in the beginning of May, we got the nice news regarding the new SQL Server 2016. By the way, Microsoft already released the SQL Server 2016 CTP2, which is awesome for SQL Server addicted people, like me :). You can find more details here.

在5月初举行的Microsoft Ignite会议期间，我们获得了有关新SQL Server 2016的好消息。顺便说一句，Microsoft已经发布了SQL Server 2016 CTP2，对于像我这样SQL Server上瘾者来说真是棒极了：）。您可以在此处找到更多详细信息。

One of the features that I am eager to put my hands on, is the improved In-Memory OLTP. Still in the Microsoft Ignite, Sunil Argawal and Kevin Farlee delivered a session where the main new improvements of the In-Memory OLTP in SQL Server 2016 were announced. In this article I’m going to pass each one of those improvements, however if you prefer, here is the link for the full Ignite session.

我渴望使用的功能之一是改进的In-Memory OLTP。仍然在Microsoft Ignite中，Sunil Argawal和Kevin Farlee举行了会议，宣布了SQL Server 2016中的内存中OLTP的主要新改进。在本文中，我将逐一介绍这些改进，但是，如果您愿意，这里是完整的Ignite会话的链接。

SQL Server 2014上的内存中OLTP ( In-memory OLTP on SQL Server 2014 )

In the past few months I’ve been delivering sessions, and even pre-con, about In-Memory OLTP and most of the people are very excited with the final results, but in other hands, concerned because of the current limitations.

在过去的几个月中，我一直在进行有关内存中OLTP的会议，甚至进行预演，大多数人对最终结果感到非常兴奋，但另一方面，由于当前的局限性，我对此感到担忧。

The true, is that this is a revolutionary feature of SQL Server, and we are just in the seeing the beginning of a very complete engine.

的确，这是SQL Server的一项革命性功能，而我们只是在看到一个非常完整的引擎的开始。

Nowadays, it is not easy to fully-migrate existing applications to In-Memory OLTP, all because of its limitations. Just to refresh our minds, here are some of the main ones:

如今，将现有应用程序完全迁移到内存中OLTP并不容易，这都是其局限性。只是为了刷新我们的思想，以下是一些主要思想：

Total size for durable tables: 250 GB.耐用桌的总大小：250 GB。
Removal of In-Memory FG not allowed.不允许删除内存中的FG。
No table changes.表格没有变化。
- Including add new indexes.包括添加新索引。
- This is a problem when you set bad value for the BUCET_COUNT.当您为BUCET_COUNT设置错误的值时，这是一个问题。
Row size limit is 8060 bytes.行大小限制为8060字节。
- Transact-SQL Constructs Not Supported by In-Memory OLTP内存中OLTP不支持的Transact-SQL构造

* To have better understanding about most of the limitations, I recommend you to watch this good compilation, made by Brent Ozar: sp_Blitz Result: Hekaton Tables In Use.

*为了更好地了解大多数限制，我建议您观看由Brent Ozar编写的出色编译： sp_Blitz结果：Hekaton Tables Use 。

Even with the existing limitations, we can find scenarios where the In-Memory OLTP is very useful and there are big companies/applications using this in production with success!

即使存在现有限制，我们也可以找到内存中OLTP非常有用的场景，并且有大公司/应用程序在生产中成功使用它！

For database architects, the In-Memory OLTP feature came as an option to improve performance of critical environments. Bellow you have some examples of its utility:

对于数据库架构师而言，内存中OLTP功能是提高关键环境性能的一种选择。在下面，您可以找到其实用程序的一些示例：

- Better together with Columnstore indexes.与Columnstore索引一起使用效果更好。
- E.g: Website sessions table.例如：网站会话表。
TempDB alternative.TempDB替代。
- For highly concurrent system.对于高度并发的系统。
Performance Critical OLTP systems.性能至关重要的OLTP系统。

SQL Server 2016的内存中OLTP ( In-memory OLTP on SQL Server 2016 )

There are few articles about In-Memory OLTP out there, so I think that we can go ahead and see the improvements made on SQL Server 2016!

很少有关于内存中OLTP的文章，所以我认为我们可以继续进行并查看SQL Server 2016上所做的改进！

The Microsoft team did a really good job and the evolution from the previous version to this new one is notable! I’m anxious to have the SQL Server 2016 officially released in order to have this trick up my sleeve