You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@openjpa.apache.org by "Kevin Sutter (JIRA)" <ji...@apache.org> on 2012/06/13 20:25:42 UTC

[jira] [Commented] (OPENJPA-2211) Ordering of SQL Statements when using @UniqueConstraints

    [ https://issues.apache.org/jira/browse/OPENJPA-2211?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13294604#comment-13294604 ] 

Kevin Sutter commented on OPENJPA-2211:
---------------------------------------

I reproduced the problem by slightly modifying the org.apache.openjpa.persistence.annotations.TestAdvAnnot JUnit.  I added the following method, expecting the Delete to happen before the Update.  No such luck...

    public void testUniqueConstraintOrder()
    {
        String query = "Select f FROM FlightSchedule f WHERE f.planeName = :pname";
        int upd = 0;

        OpenJPAEntityManager em =(OpenJPAEntityManager) currentEntityManager();
        startTx(em);
        
        FlightSchedule aw2 = (FlightSchedule) em.createQuery(query)
            .setParameter("pname", "AmericaWest2")
            .getSingleResult();
        
        FlightSchedule aw3 = (FlightSchedule) em.createQuery(query)
            .setParameter("pname", "AmericaWest3")
            .getSingleResult();
        
        aw3.setFlightName("AmericaWest2");  // change AW3 to AW2
        em.remove(aw2);                     // remove the original AW2

        endTx(em);  // can we get the order right?
        endEm(em);
    }

It's not exactly the same object model and scenario that John described, but it still shows that the SQL is not being ordered based on the UniqueConstraint annotation.

The discussion that led to this JIRA is documented here:
http://openjpa.208410.n2.nabble.com/Order-of-SQL-Statements-td7580252.html



                
> Ordering of SQL Statements when using @UniqueConstraints
> --------------------------------------------------------
>
>                 Key: OPENJPA-2211
>                 URL: https://issues.apache.org/jira/browse/OPENJPA-2211
>             Project: OpenJPA
>          Issue Type: Bug
>    Affects Versions: 2.2.0
>            Reporter: John Boblitz
>
> Scenario:
> In my Model I have an Entity "Equipment" which contains a bi-directional @OneToMany relationship with "Axle".  All entities have a primary key (uniqueId) which is contained in a MappedSuperClass "BaseEntity".
> Further, all Entities have a secondary unique key based on the business requirements.  In the case of Axles, this is the equipmentId and the axleNumber.  This is annotated in Axle through @UniqueConstraint.
> Axle is also defined as @ElementDependant in Equipment so a deletion of an Equipment or the delinking of an Axle from the Equipment will result in the deletion of the Axle.  This works as intended.
> However, it is also necessary that any Axles defined after the deleted Axle (i.e. with a higher axelNumber) should "move up" the chain.
> Here is an illustration:
> Start with 
> Equipment #413908
> Axle 1 #413966
> Axle 2 #498801
> Axle 3 #413965
> Axle 4 #504805
> User removes Axle 1 from the Collection (on screen)
> What "should" happen:
> Since the reference bewteen Axle 1 and equipment is set to null - Axel 1(413966) should be Removed from the Database.
> Axle 2 (498801) should become Axle 1
> Axle 3 (431965) should become Axle 2
> Axle 4 (504805) should become Axle 3
> However, the removal Axle 1 is occuring AFTER the statements which reorganize the other Axles.  This causes a duplicate key error as Axle 2 tries to become Axle 1 before Axle 1 is removed from the database.
> according to the Manual, this should not occur:
> 13.2. Unique Constraints
> ...
> "The unique constraints you define are used during table creation to generate the proper database constraints, and may also be used at runtime to order INSERT, UPDATE , and DELETE statements. For example, suppose there is a unique constraint on the columns of field F. In the same transaction, you remove an object A and persist a new object B, both with the same F value. The JPA runtime must ensure that the SQL deleting A is sent to the database before the SQL inserting B to avoid a unique constraint violation."
> The following are the relevant portions of the code:
> ENTITIES
> @Entity
> @Table(schema = "galaxy11", name = "Equipment") public class Equipment extends BaseEntity {
> @JsonManagedReference
> @OneToMany(cascade = CascadeType.ALL, mappedBy = "equipmentId", fetch = FetchType.EAGER) @ElementDependent 
> private Set<Axle> axles = new HashSet<Axle>();
> @Entity
> @Table(schema = "galaxy11", name = "Axle", uniqueConstraints= @UniqueConstraint(name="uq_Axle_equipmentId_axleNumber", columnNames = { "equipmentId", "axleNumber" })
> )
> public class Axle extends BaseEntity {
> @Basic
> @Column(columnDefinition = "int2")
> private short axleNumber;
> @JsonBackReference
> @ManyToOne(fetch = FetchType.EAGER)
> @JoinColumn(name = "equipmentId", columnDefinition = "int8", nullable = false) 
> private Equipment equipmentId;
> THE DELETION CODE
> TreeSet<AxleDTO> axleDtos = new TreeSet<AxleDTO>(new Comparator<AxleDTO>() {
> @Override
> public int compare(AxleDTO pO1, AxleDTO pO2) { 
> if (pO1.getAxleNumber() < pO2.getAxleNumber()) { 
> return -1; 
> } else if (pO1.getAxleNumber() == pO2.getAxleNumber()) { 
> return 0; 
> } else { 
> return 1; 
> } } });
> axleDtos.addAll(pData.getAxles());
> if (axleDtos != null && !axleDtos.isEmpty()) { 
> short i = 1; 
> Set<Axle> axles = new HashSet<Axle>(); 
> AxleMgr axleMgr = new AxleMgr(mPm); 
> for (AxleDTO axleDto : axleDtos) { 
> Axle axle = new Axle(); 
> axleMgr.toEntity(axleDto, axle); 
> if (!axle.isActive()) { 
> axle.setEquipmentId(null); 
> } else { 
> axle.setAxleNumber(i++); 
> axle.setEquipmentId(pEntity); 
> axles.add(axle); 
> } 
> } 
> pEntity.setAxles(axles); 
> }
> ** pData is the Data passed from the GUI
> TRACE OF STATMENTS
> UPDATE galaxy11.Equipment
> SET modified = ?, versionid = ?, length = ?, height = ?, wheelbase = ?, tareweight = ?, grossweight = ? 
> WHERE uniqueid = ? AND versionid = ? 
> [params=(Timestamp) 2012-06-11 13:47:18.21, (int) 33, (BigDecimal) 6.2, (BigDecimal) 3.1, (BigDecimal) 3.8, (BigDecimal) 6.7, (BigDecimal) 18, (long) 413908, (int) 32]
> - <t 656806424, conn 972792495> [27 ms] spent
> - <t 656806424, conn 972792495> executing prepstmnt 684003742 UPDATE galaxy11.Axle SET modified = ?, versionid = ?, equipmentid = ?, axlenumber = ? 
> WHERE uniqueid = ? AND versionid = ? 
> [params=(Timestamp) 2012-06-11 13:47:18.23, (int) 15, (long) 413908, (short) 2, (long) 413965, (int) 14]
> - <t 656806424, conn 972792495> [23 ms] spent
> - <t 656806424, conn 972792495> executing prepstmnt 10578812 UPDATE galaxy11.Axle SET modified = ?, versionid = ?, equipmentid = ?, axlenumber = ? 
> WHERE uniqueid = ? AND versionid = ? 
> [params=(Timestamp) 2012-06-11 13:47:18.24, (int) 2, (long) 413908, (short) 3, (long) 504805, (int) 1]
> - <t 656806424, conn 972792495> [21 ms] spent
> - <t 656806424, conn 972792495> executing prepstmnt 1704434230 UPDATE galaxy11.Axle SET modified = ?, versionid = ?, equipmentid = ?, axlenumber = ? 
> WHERE uniqueid = ? AND versionid = ? 
> [params=(Timestamp) 2012-06-11 13:47:18.24, (int) 13, (long) 413908, (short) 1, (long) 498801, (int) 12]
> - <t 656806424, conn 972792495> [22 ms] spent
> - <t 656806424, conn 972792495> executing prepstmnt 1141736277 UPDATE galaxy11.Motor SET modified = ?, versionid = ? 
> WHERE uniqueid = ? AND versionid = ? 
> [params=(Timestamp) 2012-06-11 13:47:18.24, (int) 28, (long) 414058, (int) 27]
> - <t 656806424, conn 972792495> [21 ms] spent
> - <t 656806424, conn 972792495> executing prepstmnt 1395221594 DELETE FROM galaxy11.Axle WHERE uniqueid = ? AND versionid = ? 
> [params=(long) 413966, (int) 14]

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira