The following is a list of all Connector Interfaces included with the IBM TDI. The Connector Interface is the part of the Connector that implements the actual logic to communicate with the Data Source it is supposed to handle.
We can also make our own Connector Interfaces if needed; the AssemblyLine wraps them so they are available as AssemblyLine Connectors.
Before Connectors can be meaningfully deployed in an AssemblyLine, it needs to be configured. A number of Connectors have different parameter sets, depending on the Mode they are set to; this implies that, for example, a parameter which is significant in Iterator mode, is not necessary and therefore not present in the list of parameters in AddOnly Mode.
All following AssemblyLine Connectors have access to the methods described in the com.ibm.di.server.AssemblyLineComponent in addition to the methods and properties of the Connector Interface. For documentation of the methods, see the JavaDocs (from the CE, choose Help -> Welcome -> JavaDocs.)
For a list of Supported Modes, see Legend for the Supported Mode columns.
For each Connector Interface listed, see the documentation outlined in this chapter.
For a list of Supported Modes, see Legend for the Supported Mode columns. The Script Connector enables you to write our own Connector in JavaScript™.
Generic Connector - custom - You write the Script Connector yourself in JavaScript, and it provides the modes you write into it. See "JavaScript Connector" in IBM TDI V7.1 Users Guide.
In Script-based Connectors, a potential source of problems exists if you made direct Java™ calls into the same libraries as IBM TDI. A new version of IBM TDI might have updated libraries (with different semantics), or you might have upgraded your libraries since the last time we used your Connector.
For a list of Supported Modes, see Legend for the Supported Mode columns.
When a Connector is instantiated, usually it allocates a certain amount of resources to communicate with a particular system (connection objects, session objects, result set, and so forth). When multiple Connectors of the same type are connected to the same system, often it is reasonable to share the underlying resources. This means that a single connection to the given system will be re-used by multiple Connectors.
TDI allows Connector re-use to happen within an AssemblyLine. For a given AssemblyLine the we have the option to re-use an already configured Connector from the same AssemblyLine.
With regards to the TDI Server, when re-using a Connector, a single physical Connector object is instantiated and a number of logical Connectors share it.
With regards to configuration, Connector re-using is a master-slave relation: the re-used ("master") Connector has a full connection and parser configuration and all re-using Connectors have references to the master Connector. All re-using Connectors share the connection and parser settings of the Connector they re-use. Although connection and parser settings are fixed for re-using Connectors, certain other features are configured separately (if any parameter is not configured separately, it is inherited from the master Connector):
Generally, a Connector can be re-used in the same mode (except for Iterator and Server) without any problem. This means that, for example, we can safely re-use a Connector in Lookup mode as many times as you wish.
A problem can potentially arise when a Connector is re-used in different modes. The shared physical Connector object is initialized and terminated only once. So the Connector's initialization and termination procedure must be common for all supported modes.
Following is a list of TDI Connectors which can be re-used in different modes:
Any Connector not in this list can not be re-used in the same AssemblyLine; either because it makes no sense, or because the Connector's internal logic does not allow it.
For configuring a Connector for re-use in an AssemblyLine, refer to IBM TDI V7.1 Users Guide. In the configured AssemblyLine, the re-used Connectors will show up with their name prepended with '@'.