sql server高可用

介绍 (Introduction)

On November 4th, 2019, during the Ignite conference at Orlando, Microsoft released the General Availability of its flagship product Microsoft SQL Server 2019. SQL 2019 provides various enhancements to its core database engine and offers integration with Big data (Apache Spark, Data Lake), Machine learning, Linux/container compatibility with Kubernetes.

在2019年11月4日于奥兰多举行的Ignite会议上，Microsoft发布了其旗舰产品Microsoft SQL Server 2019的General Availability.SQL 2019对其核心数据库引擎进行了各种增强，并提供了与大数据的集成（Apache Spark，Data Lake） ，机器学习，Linux /容器与Kubernetes的兼容性。

Microsoft is very excited about this new release as it takes a big leap in the integration of the database engine with other exciting features and technologies.

微软对此新版本感到非常兴奋，因为它在数据库引擎与其他令人兴奋的功能和技术的集成方面取得了长足的进步。

We can download the General Availability release from the link by clicking on Try now:

通过单击立即尝试，我们可以从链接下载General Availability版本：

SQL 2019版本历史 (SQL 2019 release history)

In the following table, we can get information about SQL 2019 releases, build versions and released dates:

在下表中，我们可以获得有关SQL 2019版本，内部版本和发布日期的信息：

Build Name	Build version	Released date
Servicing Update for SQL Server 2019 RTM	15.0.2070.41	2019-11-04
General Availability – GA	15.0.2000.5	2019-11-04
Release candidate refresh for Big data cluster	15.0.1900.47	2019-08-29
Release candidate -1	15.0.1900.25	2019-08-21
Microsoft SQL Server 2019 Community Technology Preview 3.2	15.0.1800.32	2019-07-24
Microsoft SQL Server 2019 Community Technology Preview 3.1	15.0.1700.37	2019-06-26
Microsoft SQL Server 2019 Community Technology Preview 3.0	15.0.1600.8	2019-05-17
Microsoft SQL Server 2019 Community Technology Preview 2.5	15.0.1500.28	2019-05-17
Microsoft SQL Server 2019 Community Technology Preview 2.4	15.0.1400.75	2019-03-16
Microsoft SQL Server 2019 Community Technology Preview 2.3	15.0.1300.359	2019-02-15
Microsoft SQL Server 2019 Community Technology Preview 2.2	15.0.1200.24	2019-12-05
Microsoft SQL Server 2019 Community Technology Preview 2.1	15.0.1100.94	2018-11-01
Microsoft SQL Server 2019 Community Technology Preview 2.0	15.0.1000.34	2018-09-18

内部名称	内部版本	发布日期
SQL Server 2019 RTM的服务更新	15.0.2070.41	2019-11-04
全面上市– GA	15.0.2000.5	2019-11-04
发布大数据集群的候选刷新	15.0.1900.47	2019-08-29
发布候选-1	15.0.1900.25	2019-08-21
Microsoft SQL Server 2019社区技术预览版3.2	15.0.1800.32	2019-07-24
Microsoft SQL Server 2019社区技术预览版3.1	15.0.1700.37	2019-06-26
Microsoft SQL Server 2019社区技术预览版3.0	15.0.1600.8	2019-05-17
Microsoft SQL Server 2019社区技术预览版2.5	15.0.1500.28	2019-05-17
Microsoft SQL Server 2019社区技术预览版2.4	15.0.1400.75	2019-03-16
Microsoft SQL Server 2019社区技术预览版2.3	15.0.1300.359	2019-02-15
Microsoft SQL Server 2019社区技术预览版2.2	15.0.1200.24	2019-12-05
Microsoft SQL Server 2019社区技术预览版2.1	15.0.1100.94	2018-11-01
Microsoft SQL Server 2019社区技术预览版2.0	15.0.1000.34	2018-09-18

Let us explore in brief new features of SQL Server 2019.

让我们简要探讨SQL Server 2019的新功能。

数据库引擎增强 (Database engine enhancements)

SQL Server 2019 supports various database engine enhancements over previous versions.

SQL Server 2019支持对早期版本的各种数据库引擎增强功能。

兼容性 (Compatibility )

SQL 2019 has compatibility with SQL Server on Azure VMs, Azure SQL Database Azure, SQL Database Edge. Azure SQL Database Edge is a preview version of SQL running on edge devices supporting ARM processors.

SQL 2019与Azure VM，Azure SQL数据库Azure，SQL数据库边缘上SQL Server兼容。 Azure SQL数据库边缘是在支持ARM处理器的边缘设备上运行SQL的预览版。

智能查询处理 (Intelligent Query Processing)

It extends the SQL Server 2017 Intelligent Query Processing (IQP) feature with the new database compatibility level 150.

它以新的数据库兼容级别150扩展了SQL Server 2017智能查询处理（IQP）功能。

行存储中的批处理模式 (Batch mode on row store)

It provides batch mode execution for data analytics. It does not use columnstore indexes. We can use this feature where the column store index creates overhead for the data workload.

它为数据分析提供批处理模式执行。 它不使用列存储索引。 我们可以在列存储索引为数据工作负载带来开销的情况下使用此功能。

行模式内存可提供反馈 (Row Mode Memory grants feedback)

SQL Server takes feedback from query executions and uses it for granting optimizes memory allocations in subsequent queries. You can refer to the article Row Mode Memory Grant Feedback in SQL Server 2019 for more details.

SQL Server从查询执行中获取反馈，并将其用于授予权限以优化后续查询中的内存分配。 您可以参考SQL Server 2019中的文章行模式内存授予反馈以获得更多详细信息。

表变量推迟编译 (Table variable deferred compilation)

It uses the actual cardinality in a subsequent execution of a table variable as feedback from the first query compilation.

它在随后执行表变量时使用实际基数作为来自第一次查询编译的反馈。

标量UDF内联 (Scalar UDF inlining)

It improves the performance of a scalar user-defined function with transformation in a relational expression. You can refer to Improvements of Scalar User-defined function performance in SQL Server 2019 for it.

通过在关系表达式中进行转换，可以提高标量用户定义函数的性能。 您可以参考SQL Server 2019中的标量用户定义函数性能的改进 。

近似查询处理 (Approximate query processing)

It provides a new function such as Approx_Count_Distinct for providing an approximate number of row counts for improving overall performance. You can refer to article The new SQL Server 2019 function Approx_Count_Distinct.

它提供了一个新函数，例如Approx_Count_Distinct，用于提供大约的行数以提高整体性能。 您可以参考文章新SQL Server 2019函数Approx_Count_Distinct 。

轻量级查询分析 (Lightweight query profiling)

SQL Server 2017 introduced lightweight query profiling for collecting query runtime statistics with a minimal resource overhead. It requires enabling a trace flag 7412. SQL Server 2019 enables lightweight query profiling for all sessions by default. Read more about it in Lightweight performance profiling in SQL Server 2019 article.

SQL Server 2017引入了轻量级查询分析，以最小的资源开销收集查询运行时统计信息。 它需要启用跟踪标志7412.SQL Server 2019默认情况下为所有会话启用轻量级查询分析。 在SQL Server 2019中的轻量级性能分析中阅读有关此内容的更多信息。

静默数据截断 (Silent data Truncation)

In previous versions of SQL Server, it truncates the data if it exceeds the column length. It is a tedious and frustrating task for SQL Server developers because SQL Server does not give information about data that caused this issue. SQL Server 2019 provides useful information for troubleshooting purposes. Refer article SQL truncate enhancement: Silent Data truncation in SQL Server 2019 for this information.

在早期版本SQL Server中，如果数据超过列长度，它将截断数据。 对于SQL Server开发人员来说，这是一个繁琐而令人沮丧的任务，因为SQL Server不提供有关导致此问题的数据的信息。 SQL Server 2019提供有用的信息以进行故障排除。 有关此信息，请参阅文章SQL截断增强：SQL Server 2019中的静默数据截断 。

内存中的改进 (In-memory improvements)

SQL Server provides many enhancements for in-memory features. It provides performance gains for memory-optimized tables, natively compiled stored procedures, and in-memory clustered columnstore indexes.

SQL Server为内存功能提供了许多增强功能。 它为内存优化的表，本机编译的存储过程以及内存中集群的列存储索引提供了性能提升。

内存优化的TempDB (Memory-optimized TempDB)

SQL Server 2019 provides memory-optimized TempDB metadata in removing bottlenecks for TempDB heavy workloads. You can refer to Memory-Optimized TempDB Metadata in SQL Server 2019 article for more details.

SQL Server 2019提供了内存优化的TempDB元数据，以消除TempDB繁重工作负载的瓶颈。 您可以参考SQL Server 2019中的内存优化的TempDB元数据一文以获得更多详细信息。

混合缓冲池 (Hybrid buffer pool)

SQL Server 2019 supports a hybrid buffer pool to access data pages stored on persistent memory (PMEM) devices. It eliminates the process of copying the data page in the buffer pool from the disk. SQL Server supports PMEM devices both on Windows and on Linux OS.

SQL Server 2019支持混合缓冲区池，以访问存储在持久性内存（PMEM）设备上的数据页。 它消除了从磁盘复制缓冲池中的数据页的过程。 SQL Server在Windows和Linux OS上都支持PMEM设备。

加速数据库恢复 (Accelerated Database Recovery)

Before SQL Server, database recovery or rollback takes a long time for a significant transaction or SQL Service restart. It is a painful situation for DBAs for massive production databases. SQL Server provides instant database recovery in such cases using new feature Accelerated database recovery. Refer to article, Accelerated Database Recovery; Instant Rollback and Database Recovery for more details.

在使用SQL Server之前，数据库恢复或回滚需要很长时间才能完成重要的事务或重新启动SQL Service。 对于大规模生产数据库的DBA来说，这是一个痛苦的情况。 在这种情况下，SQL Server使用新功能“加速数据库恢复”提供即时数据库恢复。 请参阅文章加速数据库恢复； 即时回滚和数据库恢复以获取更多详细信息。

列存储索引增强 (Columnstore Index Enhancements)

SQL Server provides columnstore index enhancements for:

SQL Server为以下内容提供了列存储索引增强功能：

• Columnstore Index Enhancements – data compression, estimates and savings 存储索引增强-数据压缩，估计和节省”
• Columnstore Index Enhancements – Index stats update in clone databases for more details 存储索引增强功能–克隆数据库中的索引统计信息更新。
• Columnstore Index Enhancements – online and offline (re)builds for more details Columnstore Index Enhancements –在线和离线（重新）构建

高可用性解决方案 (High Availability Solutions)

SQL Server 2019 provides the following enhancements for improving disaster recovery and high availability solutions.

SQL Server 2019提供了以下增强功能，以改善灾难恢复和高可用性解决方案。

五个同步副本对 (Five synchronous replica pairs)

SQL 2019 supports one primary and four secondary replicas in synchronous mode. It also allows automatic failover between these replicas.

SQL 2019在同步模式下支持一个主副本和四个辅助副本。 它还允许在这些副本之间进行自动故障转移。

辅助副本到主副本连接重定向 (Secondary to primary replica connection redirection)

SQL 2019 supports connection redirection from secondary to a primary replica without SQL listener as well.

SQL 2019还支持从辅助副本到主副本的连接重定向，而无需SQL侦听器。

针对AG组的增强的数据库级运行状况检测 (Enhanced database-level health detection for AG groups)

SQL 2019 improves database health detection for the Availability group databases. It initiates a failover in case of errors in health detection.

SQL 2019改进了可用性组数据库的数据库运行状况检测。 如果运行状况检测错误，它将启动故障转移。

在Kubernetes上具有远程存储的高可用性解决方案 (High availability solution with remote storage on Kubernetes)

SQL Server 2019 allows configuring AG groups using Kubernetes as an orchestration layer. It also provides database health detection for errors and failover if needed.

SQL Server 2019允许使用Kubernetes作为业务流程层配置AG组。 它还根据需要提供数据库运行状况检测以进行错误和故障转移。

大数据集群 (Big Data Clusters)

SQL 2019 supports the configuration of scalable clusters, Spark and HDFS containers. We can use T-SQL to process big data. It enables us to analyze a high volume of big data. It also supports SQL Server on Linux, Hadoop, and Kubernetes.

SQL 2019支持可伸缩群集，Spark和HDFS容器的配置。 我们可以使用T-SQL处理大数据。 它使我们能够分析大量的大数据。 它还支持Linux，Hadoop和Kubernetes上SQL Server。

增强型PolyBase (Enhanced PolyBase)

SQL Server 2019 provides feature enhancements to PolyBase and integrates with various data sources such as Oracle, MongoDB, HDFS, ODBC based data sources. You can refer the following articles for more details:

SQL Server 2019对PolyBase进行了功能增强，并与各种数据源集成，例如Oracle，MongoDB，HDFS，基于ODBC的数据源。 您可以参考以下文章以了解更多详细信息：

• Enhanced PolyBase SQL 2019 – Installation and basic overview 增强的PolyBase SQL 2019-安装和基本概述
• Enhanced PolyBase SQL 2019 – External tables for Oracle DB 增强的PolyBase SQL 2019-Oracle DB的外部表
• Enhanced PolyBase SQL 2019 – MongoDB and external table 增强的PolyBase SQL 2019 – MongoDB和外部表
• Enhanced PolyBase SQL 2019 – External tables using t-SQL 增强的PolyBase SQL 2019-使用t-SQL的外部表
• Enhanced PolyBase SQL 2019 – External tables SQL Server, Catalog view and Pushdown 增强的PolyBase SQL 2019-外部表SQL Server，目录视图和下推式

SQL Server安全性和合规性 (SQL Server Security and compliance)

SQL Server has the lowest number of security vulnerabilities for 2010-2019 as per the National Institute of Standards and Technology (NIST).

根据美国国家标准技术研究院（NIST）的数据，SQL Server在2010-2019年的安全漏洞数量最少。

SQL 2019 provides the following enhancements for security and compliance.

SQL 2019提供了以下增强的安全性和合规性。

证书管理 (Certificate Management)

SQL Server 2019 is integrated with the certificate management and provides benefits such as:

SQL Server 2019与证书管理集成在一起，提供了以下好处：

• View and validate SSL certificates 查看和验证SSL证书
• Check certificate expiration dates 检查证书的到期日期
• Deploy certificate for Always on Availability groups 部署始终可用性组的证书
• Deploy certificate for failover cluster instance 部署故障转移群集实例的证书

Read more about it using Certificate Management in SQL Server 2019.

在SQL Server 2019中使用证书管理了解更多信息。

SQL漏洞评估 (SQL Vulnerability Assessment)

We can use this tool for assessment, track and remediate database vulnerability issues in SQL Server. It is available in SSMS 17.4 or later versions. Read more about What’s new in SSMS 17.5; Data Discovery and Classification and more.

我们可以使用此工具评估，跟踪和修复SQL Server中的数据库漏洞问题。 它在SSMS 17.4或更高版本中可用。 阅读有关SSMS 17.5的新功能的更多信息。 数据发现和分类等 。

SQL数据发现和分类 (SQL Data Discovery and Classification)

We can classify database columns and provide labels such as public, general, and confidential. It is helpful for regulatory requirements such as GDPR. Read more about SQL data classification – Add sensitivity classification in SQL Server 2019

我们可以对数据库列进行分类，并提供诸如公共，常规和机密之类的标签。 这对于法规要求（例如GDPR）很有帮助。 了解有关SQL数据分类的更多信息-在SQL Server 2019中添加敏感度分类

始终使用安全区域加密 (Always Encrypted with Secure Enclaves)

SQL Server 2019 introduces Secure Enclaves in Always Encrypted technology. It protects the sensitivity of data while initial data encryption using Always Encrypted feature of SQL Server. We can also do data computation or mathematical operations, pattern matching, range comparisons as well on secure data.

SQL Server 2019在始终加密技术中引入了安全区域。 使用SQL Server的始终加密功能进行初始数据加密时，它可以保护数据的敏感性。 我们还可以对安全数据进行数据计算或数学运算，模式匹配，范围比较。

Linux上SQL Server (SQL Server on Linux)

SQL Server 2017 and 2019 works on cross-platform operation systems such as Windows and Linux. It provides the following enhancements in SQL 2019 Linux version:

SQL Server 2017和2019可在Windows和Linux等跨平台操作系统上运行。 它在SQL 2019 Linux版本中提供了以下增强功能：

• SQL Server on Linux can acts in the transactional, merge, and snapshot replication. We can use it publisher, distributor or subscriber role Linux上SQL Server可以执行事务复制，合并和快照复制。 我们可以使用它的发布者，发行者或订阅者角色
• SQL Server 2019 Linux supports Change Data Capture (CDC) feature as well for insert, update and delete activity recording SQL Server 2019 Linux支持更改数据捕获（CDC）功能以及插入，更新和删除活动记录
• It also supports distributed transactions using Microsoft distributed transaction coordinator (MSDTC) 它还使用Microsoft分布式事务处理协调器（MSDTC）支持分布式事务处理
• SQL 2019 Linux supports machine learning languages such as Python, R Scripts SQL 2019 Linux支持机器学习语言，例如Python，R脚本
• It also supports PolyBase for configuration of external data sources for Oracle, MongoDB, Teradata and ODBC based data source 它还支持PolyBase，用于为基于Oracle，MongoDB，Teradata和ODBC的数据源配置外部数据源
• SQL 2019 also supports TempDB configuration in Linux instance similar to a Windows-based instance SQL 2019还支持Linux实例中的TempDB配置，类似于基于Windows的实例

You can follow these articles for more details about SQL 2019 on Linux:

您可以按照以下文章获取有关Linux上SQL 2019的更多详细信息：

• SQL Server 2019 on Linux with Ubuntu and Azure Data Studio 具有Ubuntu和Azure Data Studio的Linux上SQL Server 2019
• SQL Server 2019 on Linux with a Docker container on Ubuntu Linux上SQL Server 2019和Ubuntu上的Docker容器
• SQL Server 2019 on Linux with Ubuntu Linux与Ubuntu上SQL Server 2019
• SQL Server 2019 installation on Ubuntu without a Docker Container 在没有Docker容器的Ubuntu上安装SQL Server 2019

图形数据库增强 (Graph database enhancements)

SQL Server supports graph databases for a complicated relationship, and hierarchical data.SQL 2019 enhances its capabilities using edge constraints, MATCH predicates, SHORTEST_PATH and derived tables support in the graph match query. Read more about in using Graph Database features in SQL Server 2019.

SQL Server支持用于复杂关系和层次结构数据的图形数据库.SQL 2019在图形匹配查询中使用边缘约束，MATCH谓词，SHORTEST_PATH和派生表支持来增强其功能。 阅读有关在SQL Server 2019中使用图数据库功能的更多信息。

增强的许可权益 (Enhanced License benefit)

SQL Server 2019 customer with software assurance program gets the following benefits:

具有软件保障计划SQL Server 2019客户获得以下好处：

• It provides one free passive SQL Server license for high availability or DR 它为高可用性或灾难恢复提供了一个免费的被动SQL Server许可证。
• It also provides a free passive asynchronous SQL Server license for DR purpose on a separate OS as well 它还在单独的OS上还提供了用于灾难恢复目的的免费的被动式异步SQL Server许可证。
• We can also use free asynchronous replica in a disaster recovery environment 我们还可以在灾难恢复环境中使用免费的异步副本

Refer to article New high availability and disaster recovery benefits for SQL Server for more details.

有关更多详细信息，请参阅文章SQL Server的新高可用性和灾难恢复优势 。

安装变更 (Installation changes)

We see changes in SQL Server installation with every release. Microsoft’s focus is to make the installation more straightforward, easy to use with maximum user-friendly configurations. In previous versions, we did TempDB configurations during the installation of SQL Server.

我们看到每个版本SQL Server安装都发生了变化。 Microsoft的重点是使安装更直接，更易于使用，并且具有最大程度的用户友好配置。 在以前的版本中，我们在安装SQL Server时进行了TempDB配置。

In the article, SQL Server 2019 overview and installation, we installed SQL Server vNext CTP 2.0. SQL 2019 is in General Availability phase now, and we should look at the installation changes. You can follow the article for installation; however, I will highlight the differences in the setup.

在SQL Server 2019概述和安装文章中 ，我们安装了SQL Server vNext CTP 2.0。 SQL 2019现在处于常规可用性阶段，我们应该查看安装更改。 您可以按照本文进行安装； 但是，我将重点介绍设置上的差异。

功能选择 (Feature Selection)

You can see below SQL Server vNext CTP 2.0 feature selection screen:

您可以在下面看到SQL Server vNext CTP 2.0功能选择屏幕：

In the General Availability release, we can see a new language option Java in machine learning services:

在General Availability版本中，我们可以在机器学习服务中看到一个新的语言选项Java：

If we select Java language, we get JRE 11.0.3 installation option in the installation process:

如果选择Java语言，则会在安装过程中获得JRE 11.0.3安装选项：

数据库引擎配置 (Database Engine Configurations)

In the previous setup, we get the following options in the database engine configurations:

在先前的设置中，我们在数据库引擎配置中获得以下选项：

• Server configuration 服务器配置
• Data directories 资料目录
• TempDB 临时数据库
• FILESTREAM 文件流

In the General Availability release setup, we get two new options in the database engine configuration.

在General Availability版本设置中，我们在数据库引擎配置中获得了两个新选项。

• MaxDOP 最大DOP
• Memory 记忆

Click on MaxDOP, and it gives you the configuration for the max degree of parallelism. It detects the logical CPU on OS and displays the recommendation for the MAXDOP configuration. You can modify it as per your requirement, but you should be careful while changing the recommended value.

单击MaxDOP ，它为您提供了最大并行度的配置。 它检测OS上的逻辑CPU，并显示有关MAXDOP配置的建议。 您可以根据需要对其进行修改，但是在更改建议值时应格外小心。

You can also click on the hyperlink specified in the page for reference purpose:

您也可以单击页面中指定的超链接以供参考：

Next, click on the Memory tab. On this page, you can configure the minimum and maximum memory configurations for SQL Server. Usually, DBA sets the min and max memory after the installation. Sometimes if you forget to set the memory configuration, you face high memory consumptions by SQL Server processes, and it might cause issues for you in terms of high memory alerts. SQL Server 2019 provides flexibility to do memory configurations during installation. It is also beneficial for a junior DBA’s as well because you can use the recommended value provided by the SQL installer:

接下来，单击“ 内存”选项卡。 在此页面上，您可以配置SQL Server的最小和最大内存配置。 通常，DBA在安装后设置最小和最大内存。 有时，如果您忘记设置内存配置，则SQL Server进程将面临大量的内存消耗，这可能会给您带来高内存警报方面的问题。 SQL Server 2019提供了在安装过程中进行内存配置的灵活性。 对于初级DBA来说，这也是有益的，因为您可以使用SQL安装程序提供的推荐值：

You can choose either recommended or default memory configuration. If you use the recommended configuration, you need to accept the recommended memory configuration using the checkbox.

您可以选择推荐或默认的内存配置。 如果使用建议的配置，则需要使用复选框接受建议的内存配置。

结论 (Conclusion)

SQL Server 2019 provides many exciting features. It integrates SQL Server relational database with big data, data virtualization, analytical capabilities with enhanced performance and troubleshooting support. You should explore it in the test environment and be familiar with the solutions provided. You can also follow SQLShack for being updated with SQL Server 2019 features.

SQL Server 2019提供了许多令人兴奋的功能。 它将SQL Server关系数据库与大数据，数据虚拟化，分析功能集成在一起，具有增强的性能和故障排除支持。 您应该在测试环境中进行探索，并熟悉提供的解决方案。 您还可以按照SQLShack进行SQL Server 2019功能更新。

翻译自: https://www.sqlshack.com/overview-of-sql-server-2019-general-availability-and-installation/

sql server高可用