您需要指定 Type（type =uuid-char），另请参阅在JPA / hibernate中映射UUID的问题。另外，您可以使用字符串字段作为id在Java中仍然保留SQL Server中的 uniqueidentifier 。

I am trying to map a UUID column in POJO to SQL Server table column using Hibernate.

The annotations are applied as follows:

@Id
@GeneratedValue
@Column(name = "Id", columnDefinition = "uniqueidentifier") 
public UUID getId(){ ... }

However, it seems that there is some endianness problem between the Java Hibernate mapping and SQL server.

For example, in my Java app, I have ids represented as:

4375CF8E-DEF5-43F6-92F3-074D34A4CE35
ADE3DAF8-A62B-4CE2-9D8C-B4E4A54E3DA1

whereas in SQL Server, these are represented as:

8ECF7543-F5DE-F643-92F3-074D34A4CE35
F8DAE3AD-2BA6-E24C-9D8C-B4E4A54E3DA1

Is there any way to have same representation at both sides?

Please note that uniqueidentifier is used only to have a uniqueidentifier typed id in SQL server instead of type binary; the same problem exists when uniqueidentifier is removed from annotation (the problem can be observed by converting binary is to uniqueidentifier).

You need to specify the @Type(type = "uuid-char"), see also Problems mapping UUID in JPA/hibernate .

Alternatively you can use a String field for the id in Java and still keep uniqueidentifier in SQL Server.