3 Replies Latest reply: May 24, 2013 6:45 AM by David Last-Oracle RSS

    Modeling Partitioned Tables on SODDM

    nelsonalberti
      Hello,

      Is it possible modeling partitioned tables on SODDM 3.3.0.747 ?

      Regards
      Nelson
        • 1. Re: Modeling Partitioned Tables on SODDM
          Philip Stoyanov-Oracle
          Hi Nelson,
          Is it possible modeling partitioned tables on SODDM 3.3.0.747 ?
          yes it's possible in physical models for Oracle, DB2 and MS SQL Server 2005 - since DM 2.0

          Philip
          • 2. Re: Modeling Partitioned Tables on SODDM
            nelsonalberti
            Great, but could I have some little orientation in how to use partitioning in the OSSDM ?
            The documentation provided is very weak, neither talks about that.

            Regards
            Nelson
            • 3. Re: Modeling Partitioned Tables on SODDM
              David Last-Oracle
              Hi Nelson,

              To define partitioning for an Oracle Table, you first need to set the "Partitioned" option to YES on the General tab in the Physical Model dialog for the Table.
              You can then select the Partitioning tab, which provides a further set of tabs for partitioning details.

              On the Partitioning/General tab, you can select the Partition type. This can be RANGE, HASH, LIST, REFERENCE or SYSTEM for Oracle 11g.
              You also need to select which Columns are to be used for partitioning (unless you are using REFERENCE or SYSTEM partitioning).
              For REFERENCE partitioning you need to select which foreign key is to be used in the "Reference Constraint" field. (This should be a foreign key referring to another partitioned Table.)
              For SYSTEM partitioning you should enter the number of partitions in the "System Partitions" field.

              HASH partitioning is defined as either HASH BY QUANTITY or HASH BY LIST OF PARTITIONS.
              If you specify HASH BY QUANTITY, you should specify the number of Hash Partitions using the "Hash Part Quantity" field (on the "Hash Part Tablespaces" tab).

              INTERVAL Partitioning can be defined by first setting the Partition Type to RANGE, and then entering an Interval Expression on the "Range Interval Clause" tab.

              For most Partitioning types, you will need to define a set of partition instances.
              First close the Table Properties dialog, then go to the "Partitions[]" node for the Table in the Browser tree and select "New" from its right-click context menu.
              For RANGE partitioning, set its "Value List" field to the upper limit for that partition instance,
              for LIST partitioning, set its "Value List" field to the value or list of values for that partition instance.
              (For HASH BY LIST OF PARTITIONS, just create the relevant partition instances.)

              If you wish to use sub-partitioning, you can specify the Subpartition Type and whether you wish to define a set of Subpartition Templates (on the Table's Partitioning/General tab).
              You should also select the Columns that are to be used for the sub-partitioning (on the "Subpart Columns" tab).

              If you set the "Use Subpartition Template" option to YES, you can create a set of Subpartition Templates using the "Subpartition Templates []" node for the Table in the Browser tree.
              Alternatively you can expand the Browser node for a specific partition instance and define a set of sub-partition instances for it.

              Regards,
              David