In the case of a relation "OneToMany" between two JPA entities, when we make a search on the parent entity, the generated SQL presents duplicated fields. Hele is an example :

                select 
                               school0_.ID as ID1_0_0_,
                               students1_.ID as ID1_1_1_,
                               school0_.NAME as NAME2_0_0_,
                               school0_.ZIP_CODE as ZIP_CODE3_0_0_,
                               students1_.FNAME as FNAME2_1_1_,
                               students1_.LNAME as LNAME3_1_1_,
                               students1_.SCHOOL_ID as SCHOOL_I4_0_0__,
                               students1_.ID as ID1_1_0__ 
                from 
                               SCHOOL school0_ 
                inner join 
                               STUDENT students1_ 
                on 
                               school0_.ID=students1_.SCHOOL_ID 
                where 
                               school0_.ZIP_CODE=? 
                order by 
                               school0_.ID

in more complicated cases, there can be more than two duplicated fields. Although the duplication of fields doesn't distort perfromance at a DB level, this problem is critical for us regarding network performances. We are indeed making hundred millions of queries a year.