Deploy a model edition

This page explains how to deploy a new model edition or update an existing one in a data location.

Once the initial model edition is deployed, subsequent model editions can be deployed using the same mechanism. This mechanism allows for:

  • Updating the deployed model edition with the latest changes performed in an open model edition.

  • Deploying a new closed model version to either a production or test environment.

  • Reverting a deployed model edition to a previous edition.

Deploying open model editions is only possible in a development data location.

Deploy a model edition

To deploy a model edition:

  1. In the Management view, expand the Data Locations node.

  2. Right-click the data location node, and select Deploy Model Edition.

  3. Save any unsaved editors if prompted.

  4. In the Deploy Model Edition wizard, select the model edition you want to deploy.

  5. Ensure the Generate Jobs Definition option is selected to generate new integration jobs.

  6. Select the Clear API Enricher Caches option to force the deletion of the enricher caches.

  7. Click Next.
    xDM computes changes required for the data location to support the new model edition. A second page displays the SQL script to execute on the schema for deploying this model edition.

  8. Click Finish to run the script and close the wizard.
    The deployment process initiates jobs before executing the SQL script to create or modify database objects. You can monitor the progress in the Console view at the bottom of Application Builder.

Once the new model edition is deployed, the previous deployment will be visible under the Deployment History node within the data location.

Deploying a model edition does not alter existing data within the data location.

While it is typically advised to update both jobs and schemas simultaneously, there are scenarios where you might prefer to update the data structure first and delay job updates.
For example, if the data currently stored in the data editions (which are using a particular model edition) is not compatible or suitable for the new version of the jobs, you may need to adjust the data to align with the updated data structures before updating the jobs themselves.
You might also choose not to deploy job definitions when the old and new definitions are similar, and you want to preserve existing job logs.
Deploying a model edition in a data location that requires an upgrade is not possible.
When deploying a model edition, Semarchy scans the database objects in the data location schema and identifies any that require creation or modification. If you intend to create database objects like indexes in the data location schema, ensure their names are prefixed with USR_ to prevent them from being removed or overwritten during model deployment.

Deployment history

After several deployments, you might want to manage your deployment history by cleaning it up.

To delete deployment history records:

  1. In the Management view, expand the Deployment History node under the data location.

  2. Select one or multiple deployments (hold down Shift to select multiple options).

  3. Right-click and select Delete.

The selected deployment history records are deleted.

The deployment history can also be purged via a REST API call. For more information, see Manage data hubs with the REST API.