Saving an existing document leads to insert when the identifier field is named `_id` without using `@Id`

Hi! I’m using mongodb to handle some data. But something is wrong when using XX.save(Object object) to update an existed document. My entity is like

@Document(value = "test")
public class TestDO
{
    @Id
    @Field(value = "_id", targetType = FieldType.OBJECT_ID)
    private String _id;

    @Field(value = "pa")
    private Integer pa;

    @Field(value = "pb")
    private String pb;
}

And following is the DAO interface

public interface TestDAO extends MongoRepository<TestDO, String>
{
}

But When I use the following code and try to update the document, error appears

TestDO testDO = testDAO.findBy_idEquals(_id);
        testDO.setPc(new Date());
        testDO.setPa(98);
        testDO.setPb("66666");
        testDAO.save(testDO);

The error is E11000 duplicate key error collection: hospital.test index: _id_ dup key: { _id: ObjectId('6130d0e926df0b133148b57a') }; nested exception is com.mongodb.MongoWriteException: E11000 duplicate key error collection: hospital.test index: _id_ dup key: { _id: ObjectId('6130d0e926df0b133148b57a') }

seems save() can’t update the data. But I google that save() can do this. Can someone fix this? Would be appreciated.

[mp911de]

Thanks for the reproducer. It’s quite a mess and it took a while to clean up the project to make it runnable. There are two issues that should be considered:

1. The used @Id annotation comes from the JPA project (javax.persistence.Id). Since MongoDB isn’t related to JPA at all, the annotation is silently ignored. Please use org.springframework.data.annotation.Id.
2. The identifier field is named _id. When renaming that one to id, then Spring Data MongoDB will identify that field as primary key and is able to correctly detect whether the entity is a new one or whether it should exist in the database.

[ppooii12009]

Thanks for the reproducer. It’s quite a mess and it took a while to clean up the project to make it runnable. There are two issues that should be considered:

1. The used @Id annotation comes from the JPA project (javax.persistence.Id). Since MongoDB isn’t related to JPA at all, the annotation is silently ignored. Please use org.springframework.data.annotation.Id.
2. The identifier field is named _id. When renaming that one to id, then Spring Data MongoDB will identify that field as primary key and is able to correctly detect whether the entity is a new one or whether it should exist in the database.

Thanks! I have resolved this problem. Like what you said, I use the wrong @Id annotation from javax.persistence.Id Very appreciated for your patience!