[转帖]SQL Server JDBC – Set sendStringParametersAsUnicode to false
https://vladmihalcea.com/sql-server-jdbc-sendstringparametersasunicode/
https://learn.microsoft.com/en-us/sql/connect/jdbc/setting-the-connection-properties?view=sql-server-ver16 If the sendStringParametersAsUnicode property is set to "true", String parameters are sent to the server in Unicode format. If the sendStringParametersAsUnicode property is set to "false", String parameters are sent to the server in non-Unicode format such as ASCII/MBCS instead of Unicode. The default value for the sendStringParametersAsUnicode property is "true". Note: The sendStringParametersAsUnicode property is only checked to send a parameter value with CHAR, VARCHAR, or LONGVARCHAR JDBC types. The new JDBC 4.0 national character methods, such as the setNString, setNCharacterStream, and setNClob methods of SQLServerPreparedStatement and SQLServerCallableStatement classes, always send their parameter values to the server in Unicode whatever the setting of this property. For optimal performance with the CHAR, VARCHAR, and LONGVARCHAR JDBC data types, an application should set the sendStringParametersAsUnicode property to "false" and use the setString, setCharacterStream, and setClob non-national character methods of the SQLServerPreparedStatement and SQLServerCallableStatement classes. When the application sets the sendStringParametersAsUnicode property to "false" and uses a non-national character method to access Unicode data types on the server side (such as nchar, nvarchar and ntext), some data might be lost if the database collation doesn't support the characters in the String parameters passed by the non-national character method. An application should use the setNString, setNCharacterStream, and setNClob national character methods of the SQLServerPreparedStatement and SQLServerCallableStatement classes for the NCHAR, NVARCHAR, and LONGNVARCHAR JDBC data types.
Imagine having a tool that can automatically detect JPA and Hibernate performance issues. Wouldn’t that be just awesome?
Well, Hypersistence Optimizer is that tool! And it works with Spring Boot, Spring Framework, Jakarta EE, Java EE, Quarkus, or Play Framework.
So, enjoy spending your time on the things you love rather than fixing performance issues in your production system on a Saturday night!
Introduction
In this article, I’m going to explain why you should always disable the sendStringParametersAsUnicode default JDBC Driver setting when using SQL Server.
Database table
Let’s assume we have the following database table:

The PostID column is the Primary Key, and the Title column is of the VARCHAR type and has a secondary index as well:
|
1
|
CREATE INDEX IDX_Post_Title ON Post (Title) |
The Post table contains the following records:
| PostID | Title ||--------|---------------------------------------------|| 1 | High-Performance Java Persistence, part 1 || 2 | High-Performance Java Persistence, part 2 || 3 | High-Performance Java Persistence, part 3 || 4 | High-Performance Java Persistence, part 4 || .. | .. || 249 | High-Performance Java Persistence, part 249 || 250 | High-Performance Java Persistence, part 250 | |
As you can see, the Title column is highly selective since every record has a different title value.
Unexpected CONVERT_IMPLICIT and Clustered Index Scan
When finding a Post row by its associated Title column value, we expect an Index Seek operation against the IDX_Post_Title index, but this is not what we get when using the default SQL Server JDBC settings.
For instance, if we enable the runtime query statistics to retrieve the associated execution plan of the SQL query that filters by the Title column:
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
|
executeStatement(entityManager, "SET STATISTICS IO, TIME, PROFILE ON");try (PreparedStatement statement = connection.prepareStatement(""" SELECT PostId, Title FROM Post WHERE Title = ? """)) { statement.setString(1, title); if (statement.execute() && statement.getMoreResults()) { LOGGER.info("Execution plan: {}{}", System.lineSeparator(), resultSetToString(statement.getResultSet()) ); }} |
We get the following SQL execution plan:
|
1
2
3
4
5
|
|StmtText ||----------------------------------------------------------------------------------------------------||SELECT PostId, Title FROM Post WHERE Title = @P0 || |--Clustered Index Scan(OBJECT:([high_performance_sql].[dbo].[Post].[PK__Post__AA12603828AEBF55]),|| WHERE:(CONVERT_IMPLICIT(nvarchar(255),[high_performance_sql].[dbo].[Post].[Title],0)=[@P0])) | |
The Clustered Index Scan operation tells us that SQL Server has used the PostId Clustered Index to scan the leaf pages in search of the Title value we provided.
The reason why the IDX_Post_Title index was not used is because of the implicit conversion that was done between the provided NVARCHAR value and the VARCHAR value of the Title column.
Even if we provided the Title bind parameter value as a VARCHAR using the setString method:
|
1
|
statement.setString(1, title); |
The SQL Server JDBC Driver behaved as if we used setNString method instead.
SQL Server JDBC sendStringParametersAsUnicode configuration
By default, SQL Server sends all String parameter values as NVARCHAR since the sendStringParametersAsUnicode configuration is set to true.
So, if we set the sendStringParametersAsUnicode configuration value to false
|
1
|
jdbc:sqlserver://localhost;instance=SQLEXPRESS;databaseName=high_performance_sql;sendStringParametersAsUnicode=false; |
And, rerun the previous SQL query, we will get the following execution plan:
|
1
2
3
4
5
|
|StmtText ||--------------------------------------------------------------------------------||SELECT PostId, Title FROM Post WHERE Title = @P0 || |--Index Seek(OBJECT:([high_performance_sql].[dbo].[Post].[IDX_Post_Title]), || SEEK:([high_performance_sql].[dbo].[Post].[Title]=[@P0]) ORDERED FORWARD)| |
That’s exactly what we were expecting from the start. There’s an Index Seek on the IDX_Post_Title index, and there’s no implicit conversion happening anymore.
Handing Unicode characters
Now, even if you disable the sendStringParametersAsUnicode setting, you can still persist Unicode data in NHAR, NVARCHAR or NLONGVARCHAR column.
So, if the Title column is of the NVARCHAR type:
|
1
2
3
4
5
|
CREATE TABLE Post ( PostID BIGINT NOT NULL, Title NVARCHAR(255), PRIMARY KEY (PostID)) |
We can set the Title column using the setNString PreparedStatement method:
|
1
2
3
4
5
6
7
8
9
10
11
|
try (PreparedStatement statement = connection.prepareStatement(""" INSERT INTO Post (Title, PostID) VALUES (?, ?) """)) { statement.setNString(1, "România"); statement.setLong(2, 1L); assertEquals(1, statement.executeUpdate());} |
And, we can read the Title column using the getNString ResultSet method:
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
|
try (PreparedStatement statement = connection.prepareStatement(""" SELECT Title, PostId FROM Post WHERE Title = ? """)) { statement.setNString(1, "România"); try(ResultSet resultSet = statement.executeQuery()) { if (resultSet.next()) { assertEquals("România", resultSet.getNString(1)); assertEquals(1L, resultSet.getLong(2)); } }} |
If you’re using JPA and Hibernate, the NVARCHAR column needs to be annotated with the @Nationalized Hibernate annotation to instruct Hibernate that the underlying String attribute needs to be handled by the StringNVarcharType, as opposed to the default StringType:
|
1
2
3
4
5
6
7
8
9
10
11
12
|
@Entity(name = "Post")public class Post { @Id @Column(name = "PostID") private Long id; @Column(name = "Title") @Nationalized private String title; } |
Awesome, right?
If you enjoyed this article, I bet you are going to love my Book and Video Courses as well.
2 Comments on “SQL Server JDBC – Set sendStringParametersAsUnicode to false”
-
Thank you for inspiring article, but are you sure about your results? I tried the scenario you described and SQL server have chosen index seek operation even in case of sendStringParametersAsUnicode=true.
May be it is dependent of SQL server version. I used MSSQL 2017.Regards
Pavel Rund-
You’re welcome.
This test provides the proof.
Here are the results:
12345678910111213Test with sendStringParametersAsUnicode=true| Rows | Executes | StmtText | StmtId | NodeId | Parent | PhysicalOp | LogicalOp | Argument | DefinedValues | EstimateRows | EstimateIO | EstimateCPU | AvgRowSize | TotalSubtreeCost | OutputList | Warnings | Type | Parallel | EstimateExecutions || ---- | -------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ------ | ------ | ------ | -------------------- | -------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | ------------------------------------------------------------------------------------------------------------------- | ------------ | ------------ | ----------- | ---------- | ---------------- | ------------------------------------------------------------------------------------------------------------------- | -------- | -------- | -------- | ------------------ || 1 | 1 | SELECT PostId, Title FROM Post WHERE Title = @P0 | 1 | 1 | 0 | | | | | 2.0 | | | | 0.0050384817 | | | SELECT | 0 | || 1 | 1 | |--Clustered Index Scan(OBJECT:([high_performance_java_persistence].[dbo].[Post].[PK__Post__AA12603836E8D7BA]), WHERE:(CONVERT_IMPLICIT(nvarchar(255),[high_performance_java_persistence].[dbo].[Post].[Title],0)=[@P0])) | 1 | 2 | 1 | Clustered Index Scan | Clustered Index Scan | OBJECT:([high_performance_java_persistence].[dbo].[Post].[PK__Post__AA12603836E8D7BA]), WHERE:(CONVERT_IMPLICIT(nvarchar(255),[high_performance_java_persistence].[dbo].[Post].[Title],0)=[@P0]) | [high_performance_java_persistence].[dbo].[Post].[PostID], [high_performance_java_persistence].[dbo].[Post].[Title] | 2.0 | 0.0046064816 | 4.32E-4 | 61 | 0.0050384817 | [high_performance_java_persistence].[dbo].[Post].[PostID], [high_performance_java_persistence].[dbo].[Post].[Title] | | PLAN_ROW | 0 | 1.0 |Test with sendStringParametersAsUnicode=false| Rows | Executes | StmtText | StmtId | NodeId | Parent | PhysicalOp | LogicalOp | Argument | DefinedValues | EstimateRows | EstimateIO | EstimateCPU | AvgRowSize | TotalSubtreeCost | OutputList | Warnings | Type | Parallel | EstimateExecutions || ---- | -------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ------ | ------ | ------ | ---------- | ---------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------- | ------------ | ---------- | ----------- | ---------- | ---------------- | ------------------------------------------------------------------------------------------------------------------- | -------- | -------- | -------- | ------------------ || 1 | 1 | SELECT PostId, Title FROM Post WHERE Title = @P0 | 1 | 1 | 0 | | | | | 1.0 | | | | 0.0032831 | | | SELECT | 0 | || 1 | 1 | |--Index Seek(OBJECT:([high_performance_java_persistence].[dbo].[Post].[IDX_Post_Title]), SEEK:([high_performance_java_persistence].[dbo].[Post].[Title]=[@P0]) ORDERED FORWARD) | 1 | 2 | 1 | Index Seek | Index Seek | OBJECT:([high_performance_java_persistence].[dbo].[Post].[IDX_Post_Title]), SEEK:([high_performance_java_persistence].[dbo].[Post].[Title]=[@P0]) ORDERED FORWARD | [high_performance_java_persistence].[dbo].[Post].[PostID], [high_performance_java_persistence].[dbo].[Post].[Title] | 1.0 | 0.003125 | 1.581E-4 | 61 | 0.0032831 | [high_performance_java_persistence].[dbo].[Post].[PostID], [high_performance_java_persistence].[dbo].[Post].[Title] | | PLAN_ROW | 0 | 1.0 |So, you can run the test for yourself and see that it works as explained in the article.
-
[转帖]SQL Server JDBC – Set sendStringParametersAsUnicode to false的更多相关文章
- SQL Server JDBC驱动中sqljdbc和sqljdbc4区别
为了支持向后兼容以及可能的升级方案,JDBC Driver 2.0 在每个安装包中都包括 2 个 JAR 类库:sqljdbc.jar 和 sqljdbc4.jar. qljdbc.jar 类库提供对 ...
- 下载 Microsoft SQL Server JDBC 驱动程序
JDBC 驱动程序中使用 Maven 中心 JDBC 驱动程序可以通过将其添加为依赖项在 POM.xml 文件中使用以下代码添加到 Maven 项目: XML复制 <dependency> ...
- Microsoft SQL Server JDBC 驱动程序支持矩阵
本页包含 Microsoft SQL Server JDBC 驱动程序的支持矩阵和支持生命周期策略. Microsoft JDBC 驱动程序支持生命周期矩阵和策略 Microsoft 支持生命周期 ( ...
- [转帖]SQL Server 索引中include的魅力(具有包含性列的索引)
SQL Server 索引中include的魅力(具有包含性列的索引) http://www.cnblogs.com/gaizai/archive/2010/01/11/1644358.html 上个 ...
- [转帖]SQL Server 10分钟理解游标
SQL Server 10分钟理解游标 https://www.cnblogs.com/VicLiu/p/11671776.html 概述 游标是邪恶的! 在关系数据库中,我们对于查询的思考是面向集合 ...
- [转帖]SQL Server DBCC命令大全
SQL Server DBCC命令大全 原文出处:https://www.cnblogs.com/lyhabc/archive/2013/01/19/2867174.html DBCC DROPC ...
- [转帖]SQL Server 2000~2017补丁包
SQL Server 2000~2017补丁包 https://www.cnblogs.com/VicLiu/p/11510510.html 最新更新 Product Version Latest S ...
- [转帖]sql server版本特性简介、版本介绍简介
sql server版本特性简介.版本介绍简介 https://www.cnblogs.com/gered/p/10986240.html 目录 1.1.sql server的版本信息 1.2.版本重 ...
- JDBC连接SQL Server
下载jdbc驱动包 下载地址,我下载的是exe版本的,其实是格自解压包.下载完毕之后,双击运行,会解压在当前目录下. Microsoft SQL Server JDBC Driver 3.0\sqlj ...
- Java使用JDBC连接SQL Server数据库
Java使用JDBC连接SQL Server数据库 1.下载驱动 1.下载Microsoft SQL Server JDBC 驱动程序 https://docs.microsoft.com/zh-cn ...
随机推荐
- Proxy下的Prepare透传,让GaussDB(for MySQL)更稳固,性能更卓越
本文分享自华为云社区<Proxy下的Prepare透传,让GaussDB(for MySQL)更稳固,性能更卓越>,作者: GaussDB 数据库 . 1.引言 在很多业务场景下,数据库应 ...
- 谁说AI看不懂视频?
摘要:人工智能在视觉领域发展趋于成熟,基于人工智能的视频内容分析能从根本上解决传统内容分析方法性能低下的问题,视频分析开启2.0智能时代. 视频数据量激增,数据处理和内容运营成本居高不下 云计算.大数 ...
- 【有奖征文】WEB前端大作战,走在技术最前端!
摘要:投稿分享你在前端领域的积累,秀出你的技术"肌肉",为自己,也为技术发声. 近几年大家对于WEB前端的关注度很高, 比如整体势头发展良好,各种技术百花齐放,人才稀缺, 随着互联 ...
- vue3溢出文本tooltip或title展示解决方案—如何获取文本宽度
vue3溢出文本tooltip或title展示解决方案-如何获取文本宽度 Author:zhoulujun Date:2023-03-06 Hits:5 解决文本溢出,鼠标悬浮展示tooltips,要 ...
- 基于AIO架构smarthttp开发的完整MVC框架
基于AIO架构smarthttp开发的完整MVC框架 写了篇<基于jdk自带httpserver开发的最小完整MVC框架>,就再写篇AIO的MVC DEMO启动时间:0.1s(应该算少的吧 ...
- Jenkins Pipeline 流水线 - 完整构建 Pipeline Script 脚本
Docker Jenkins 安装配置 Windows 2016 安装 Jenkins 前置条件可参考 Jenkins Pipeline 流水线 - 拉代码(SVN) + Maven 编译打包 Jen ...
- 多线程 ThreadPoolTaskExecutor 应用
1.如何判断线程池所有任务是否执行完毕 package com.vipsoft.web; import org.junit.jupiter.api.Test; import org.slf4j.Log ...
- Python中字符前添加r,b,u,f前缀的含义
1.在python字符串前添加r,意思为消除转义字符 2.在python字符串前添加f,意思为支持大括号内的python 表达式. 3.在python字符串前添加b,意思为字符串类型为byte类型,在 ...
- 国内申请微软新必应(New Bing)
国内申请微软新必应(New Bing) 本文解决了两个问题: 1 需国外网络环境 2 点击加入候补名单无限返回错误 注册outlook邮箱 https://outlook.live.com/ 一步一步 ...
- PS 新建作业DUMP DBSQL_DUPLICATE_KEY_ERROR
1.CJ20N新建作业 在CJ20N中新建作业后,保存DUMP,报以下错误 2.相关NOTE 605584 - CN22: update termination when creating activ ...