Assembling the Publishing Process |
Previous
|
|
Next
|
Loading Data Into the Hub |
|
Running the Publishing Process |
Assembling the Publishing Process
Data is published in the hub using a transaction named an
external load. We will implement the external load in Convergence for Data Integration as a
Process. This process will create the external load, run the mappings designed in the previous chapter and then submit this external load to the MDM hub. The process will also manage issues and automatically cancel the external load in the event of an error.
Creating the Publishing Process
- Select the
GettingStarted > Process folder in your project.
- Click the
Create New Process button in the
Project Explorer toolbar. The
New Process Diagram wizard opens.
- Make sure that the
GettingStarted/Process folder is selected as the parent folder.
- In
File Name enter
PublishDatatoHub and then click
Finish. The
PublishDatatoHub process editor opens.
Initializing the External Load
The first step of the external load process is the Initialization. This step requests a Load ID from the Semarchy Repository using a PL/SQL function call and stores it in the
SEM_LOAD_ID
variable used in the mappings.
- In the
Project Explorer, expand the
global > template.semarchy > Semarchy Convergence for MDM Integration node.
- Drag and drop the
PROCESS Get LoadID process from the project explorer to the process editor.
This process must be configured with the connection information to the Semarchy Repository and with information about the data location.
- Expand the
GettingStarted > Metadata > OracleServer node.
- From the
Project Explorer, drag and drop the
OracleServer data server in the
Semarchy Get LoadID process reference in the process editor. This metadata link to the data server appears in the process reference as shown below.
- Select the metadata link named
OracleServer, then click on it again to make it editable.
- Rename it to
SOURCE
and then press
ENTER.

The
Semarchy Get LoadID process assumes that the repository data server is provided as a metadata link named
SOURCE
. By adding this metadata link to the data server containing the repository and renaming it accordingly, you allow this process to use the data server configuration previously defined.
- Select the
Semarchy Get LoadID process reference in the process diagram.
- In the
Properties view, select the
Standard tab.
- In the
Name, enter
Initialize External Load. The
Semarchy Get LoadID step in the diagram is renamed to
Initialize External Load.
- Click the
Data Location Name link to unlock the field and then enter
CustomerAndFinancialMDM
. This parameter of the process is the name of the data location hosting the data edition into which the publishing process takes place.
- Click the
User Name link to unlock the field and then enter
Semarchy DI Publisher
. This name is used to initialize the load and manage it afterwards.
- Click the
Sem Repo Schema link to unlock the field and then enter
SEMARCHY_DEMO_REPOSITORY
. This value identifies the repository schema in the data server referred to as SOURCE.
- Click the
edit parameters icon to expand the content of the
Initialize External Load step and view all its parameters.

Loading Data
The second step of the external load process is the data load. For this step, the mappings created previously are directly re-used.
- In the
PublishDatatoHub process editor, in the
Palette that appears on the right hand, select the
Process tool in the
Component accordion.
- Click in the
PublishDatatoHub process editor background. A new sub-process appears, and you can edit its name.
- Enter
Load Data and then press
ENTER.
- A new blank editor named
Load Data opens to edit this sub-process.
- In the
Project Explorer Expand the
GettingStarted > Mapping node.
- Press the
SHIFT key and select all six mappings in this folder. Drag and drop the selection into the
Load Data sub-process editor.
The mappings are added to the process.
- Close the
Load Data sub-process editor to return to the
PublishDatatoHub process editor. When prompted for saving the
Load Data sub-process, select
Yes.
Submitting the External Load
After loading the publisher data into the landing tables using the mappings, the external load must be submitted to Semarchy Convergence for MDM. This step submits the Load using a PL/SQL function call to the Semarchy Repository.
- From the
Project Explorer, drag and drop the
global > template.semarchy > Semarchy Convergence for MDM Integration > PROCESS Submit Load process into the process editor.
- From the
Project Explorer, drag and drop the
OracleServer data server in the
Semarchy Submit Load process reference in the process editor. This metadata link to the data server appears in the process reference.
- Select the metadata link named
OracleServer, then click on it again to make it editable.
- Rename it to
SOURCE
and then press
ENTER.
- Select the
Semarchy Submit Load process reference in the process diagram.
- In the
Properties view, select the
Standard tab.
- In the
Name, enter
Submit External Load. The step in the diagram is renamed to
Submit External Load.
- Click the
Integration Job Name link to unlock the field and then enter
INTEGRATE_DATA
. This parameter of the process is the name of the Semarchy Convergence for MDM integration job that must be started when the external load is submitted.
- Click the
User Name link to unlock the field and then enter
Semarchy DI Publisher
.
- Click the
Sem Repo Schema link to unlock the field and then enter
SEMARCHY_DEMO_REPOSITORY
.
- Click the
edit parameters icon to expand the content of the
Submit External Load step and view all its parameters.
Canceling the External Load
In the event of an error while loading the data, the external load should be cancelled.
Canceling the load is also performed using a PL/SQL function call to the Semarchy Repository.
- From the
Project Explorer, drag and drop the
global > template.semarchy > Semarchy Convergence for MDM Integration > PROCESS Cancel Load process into the process editor.
- From the
Project Explorer, drag and drop the
OracleServer data server in the
Semarchy Cancel Load process reference in the process editor. This metadata link to the data server appears in the process reference.
- Select the metadata link named
OracleServer, then click on it again to make it editable.
- Rename it to
SOURCE
and then press
ENTER.
- Select the
Semarchy Cancel Load process reference in the process diagram.
- In the
Properties view, select the
Standard tab.
- In the
Name, enter
Cancel External Load. The step in the diagram is renamed to
Cancel External Load.
- Click the
Cleanup Data link to unlock the field and then select this option. This parameter forces this process to cleanup the data loaded in the Load Data step.
- Click the
User Name link to unlock the field and then enter
Semarchy DI Publisher
.
- Click the
Sem Repo Schema link to unlock the field and then enter
SEMARCHY_DEMO_REPOSITORY
.
- Click the
MDM Schema link to unlock the field and then enter
SEMARCHY_DEMO_MDM
. This value identifies the schema in the data server containing the MDM Hub.
- Click the
edit parameters icon to expand the content of the
Cancel External Load step and view all its parameters.
Designing the Publishing Process Flow
All the steps of the process are now created. It is time to link them together.
Creating the Publishing Process
- In the
PublishDatatoHub process editor, in the Palette that appears on the right hand, select the
Successful Link tool in the
Link accordion.
- Click and drag from
Initialize External Load to
Load Data. A green arrow appears between these two steps.
- Repeat the two previous steps to create a link from
Load Data to
Submit External Load.
- In the Palette, select the
Unsuccessful Link tool.
- Create a link from
Load Data to
Cancel External Load. This link will be followed if the
Load Data step fails, and appears in red.
- Press
CTRL+S to save the
PublishDatatoHub process.
Editing the Load Data Sub-Process
- Double-click the
Load Data sub-process to open it.
- Use the
Successful Link tool in the palette to link the steps of this sub-process. Note that these mappings target tables with no referential constraints implemented in the database. They can be executed in any order in the sub-process. The sub-process appears as shown below:
- Close the
Load Data sub-process editor to return to the
PublishDatatoHub process editor. When prompted for saving the
Load Data sub-process, select
Yes.
- Press
CTRL+S to save the
PublishDatatoHub process.