Data and Process Flows

The CMRUN program, which DataMigrator uses to initiate flows or stored procedures, takes an optional parameter of x=filename. This parameter allows you to store the password in a protected file.

In previous releases, the parameter was a DDNAME and only used on IBM zSeries (MVS).

Now it can be used on any other platform, including Windows and UNIX, where it is specified as a file name in the current directory or as the relative or full path to a file name.

When working with a Process Flow that initiates another data flow, you can now prevent the called flow from starting if another instance of the flow is running, by selecting the Do not allow multiple instances check box. This option is available in the Properties pane of a flow object.

In the DMC, when viewing a flow with expanded objects, multiple objects can be selected by pressing the Ctrl key while clicking them.

If there are no objects selected in either a data or process flow, the new Expand All or Collapse All options are available.

In the Data Management Console, if the name of an object is too long to fully display, it is truncated and shown with ellipses. You can see the full name by hovering over the object.

DataMigrator has a new option that allows for faster loading when a source and target are in the same relational database, resulting in an increase in performance. When working with:

• Databases that support the MERGE command, the merge is generated and passed through to the database so that both the extract and load is done in a single SQL statement. This option allows specifying different calculations for insert and update operations. Updates can use values from both the source and target tables. Databases that support a SQL MERGE command include MS SQL Server 2008 and Db2 9.1 (and later).
• Insert only operations, an INSERT INTO ... SELECT command is generated and passed through to the relational database.

In either case, because all operations are done in the database, a much faster throughput is possible.

You can enable this option by opening the flow properties of a data flow and selecting the Optimize Load checkbox under the Execution section.

Note: Since the operation is done as a single SQL command rather than row-at-a-time processing, some capabilities are not available when this option is selected. These include Record Logging and use of VALIDATE to reject individual rows. Also, a count of the number of rows inserted, updated, or deleted is not available.

The user should consider both modes of operation to determine which is best suited for their application.

Target Properties has a new attribute, Display Name, that allows user specification of the name that appears on the flow. The default is the application name, if enabled in flow properties, and the synonym name.

The following new features are available for emails:

• Email File Types for a Process Flow

  Any file type can be attached to a DataMigrator process flow email.

• Email Recipients

  In previous releases, when sending email messages, all recipients appeared in the To line. Now, recipients can also appear on the Cc line, or hidden as Bcc.

• Sending HTML Output on Message Failure

  The Send HTML output as inline message located in the Properties panel of a Process Flow Start object now allows you to send HTML output from a report in an email message on failure, as well as on successful completion.

You can now update, instead of replace, values in a target table using the On Match Expressions tab that can be found in the Transformations dialog box. For example, you can add the value from an incoming row to the value in the target table instead if replacing it completely.