Creating multiple instances of connectors on one server

Creating multiple instances of a connector is in many ways the same as creating a custom connector. You can set your system up to create and run multiple instances of a connector by following the steps below. You must:

Create a new directory

You must create a connector directory for each connector instance. This connector directory should be named:

For Windows:

ProductDir\connectors\connectorInstance

For UNIX:

ProductDir/connectors/connectorInstance

where connectorInstance uniquely identifies the connector instance. If the connector has any connector-specific meta-objects, you must create a meta-object for the connector instance. If you save the meta-object as a file, store the file in the connectorInstance directory.

Create a connector definition

You create a configuration file (connector definition) for the connector instance in Connector Configurator. To do so:

  1. Copy the initial connector's configuration file (connector definition) and rename it.
  2. Make sure each connector instance correctly lists its supported business objects (and any associated meta-objects).
  3. Customize any connector properties as appropriate.

Create a start-up script

To create a startup script:

  1. Copy the initial connector's startup script and name it to include the name of the connector directory:

    dirname

  2. Put this startup script in the connector directory you created in "Create a new directory".
  3. Create a startup script shortcut (Windows only).
  4. Copy the initial connector's shortcut text and change the name of the initial connector (in the command line) to match the name of the new connector instance.

You can now run both instances of the connector on your integration server at the same time.

For more information on creating custom connectors, refer to the Connector Development Guide for C++ or for Java.

Copyright IBM Corp. 1997, 2004