Integrating Data in an ANSI SQL-92 Compliant Technology

Generic SQL 4-9 Using a Multi-connection IKM A multi-connection IKM allows updating a target where the staging area and sources are on different data servers. Oracle Data Integrator provides the following multi-connection IKMs for ANSI SQL-92 compliant technologies: IKM SQL to SQL Incremental Update and IKM SQL to SQL Control Append. See Table 4–5 for more information on when to use a multi-connection IKM. To use a multi-connection IKM in an ETL-style interface: 1. Create an integration interface with an ANSI SQL-92 compliant staging area and target using the standard procedure as described in Creating an Interface in the Oracle Fusion Middleware Developer’s Guide for Oracle Data Integrator. This section describes only the ETL-style specific steps.

2. In the Definition tab of the Interface Editor, select Staging Area different from

Target and select the logical schema of the source tables or a third schema. 3. In the Flow tab, select one of the Source Sets, by clicking its title. The Property Inspector opens for this object. 4. Select an LKM from the LKM Selector list to load from the sources to the staging area. See Table 4–5 to determine the LKM you can use. 5. Optionally, modify the KM options. 6. In the Flow tab, select the Target by clicking its title. The Property Inspector opens for this object. In the Property Inspector, select an ETL multi-connection IKM from the IKM Selector list to load the data from the staging area to the target. See Table 4–5 to determine the IKM you can use. Note the following when setting the KM options: ■ For IKM SQL to SQL Incremental Update ■ If you do not want to create any tables on the target system, set FLOW_ CONTROL=false and FLOW_TABLE_LOCATION=STAGING. Please note that this will lead to row-by-row processing and therefore significantly lower performance. ■ If you set the options FLOW_CONTROL or STATIC_CONTROL to true, select a CKM in the Controls tab. Note that if FLOW_CONTROL is set to true, the flow table is created on the target, regardless of the value of FLOW_ TABLE_LOCATION. ■ The FLOW_TABLE_LOCATION option can take the following values: Value Description Comment TARGET Objects are created on the target. Default value. STAGING Objects are created only on the staging area, not on the target. Cannot be used with flow control. Leads to to row-by-row processing and therefore loss of performance. NONE No objects are created on staging area nor target. Cannot be used with flow control. Leads to to row-by-row processing and therefore loss of performance. Requires to read source data twice in case of journalized data sources