[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
- react+echarts出现“There is a chart instance already initialized on the dom.”
写了一个关于echatrs组件,报错dom重复 配置信息从props拿 let chart; useEffect(() => { if (chart) { updateChartView(); ...
- Java 获取Excel中的表单控件
Excel中可通过[开发工具]菜单栏下插入表单控件,如文本框.单选按钮.复选框.组合框等等,插入后的控件可执行设置控件格式,如大小.是否锁定.位置.可选文字.数据源区域.单元格链接等.当Excel中已 ...
- 教你几个部署多个nginx-ingress的注意事项
本文分享自华为云社区<nginx-ingress工作原理以及多nginx-ingress部署注意事项>,作者: 可以交个朋友. 一.nginx-ingress工作原理 nginx-ingr ...
- Web 全栈开发利器: 强大的在线 Cloud IDE
摘要:近年来,敏捷.DevOps的理念已逐步成为主流.基于云计算的开发环境也正获得越来越多开发者的青睐.不难想象,云端IDE已成未来的趋势. 学了Web全栈开发,就得动手实践,要动手,得先有开发环境. ...
- 华为云IoT设备接入服务全体验
摘要:华为云IoT设备接入服务,海量设备,一键接入,你值得拥有! 本文分享自华为云社区<[云驻共创]Huawei Mate 40产线直击之 华为云IoT设备接入服务全体验>,原文作者:启明 ...
- Open Serverless Benchmark Initiative: 华为云联合上海交大发布ServerlessBench 2.0
Key Takeaways 华为云联合上海交大,首次提出 Open Serverless Benchmark Initiative (OSBI) ,推动Serverless基准测评规范化.标准化: O ...
- 云图说|玩转华为HiLens之端云协同AI开发
阅识风云是华为云信息大咖,擅长将复杂信息多元化呈现,其出品的一张图(云图说).深入浅出的博文(云小课)或短视频(云视厅)总有一款能让您快速上手华为云.更多精彩内容请单击此处. 摘要: 华为HiLens ...
- 华为云企业级Redis揭秘第15期:Redis为什么需要强一致?
摘要:其实开源Redis的弱一致性已经不满足很多应用场景的诉求.怎么,不信? 本文分享自华为云社区<华为云企业级Redis揭秘第15期:Redis为什么需要强一致?>,作者: GaussD ...
- 没想到,学棋五年的我竟然输给了昇腾CANN!
摘要:整整两天,上百场对弈,TA竟然未尝一败,真是让人拍案叫绝. 近日,一位神秘"人物"亮相华为昇腾CANN技术开放日现场,引得众人簇拥,吸粉无数.从现场AI棋艺大战的画面中我们可 ...
- 火山引擎DataLeap如何解决SLA治理难题(二):申报签署流程与复盘详解
申报签署流程详解 火山引擎DataLeap SLA保障的前提是先达成SLA协议.在SLA保障平台中,以申报单签署的形式达成SLA协议.平台核心特点是优化了SLA达成的流程,先通过"系统卡点计 ...