Exporting a project

You can share assets in a project with others by exporting them.

The methods that you can use to export assets in a project depend on the type of project you are working in:

  • For projects with no Git integration, you can:

  • For projects with default Git integration, you can:

    1. Push your changed project files to the Git repository.
    2. Use your Git service to create an archive of your repository. This archive can be used anywhere a project ZIP can be used.

  • For projects with deprecated Git integration, you can:

    • Export the project to desktop

      This project ZIP file can be shared with others and used to create a new project from file.

    • Synchronizing changes with Git

      This functionality is only supported in projects with deprecated Git integration. The assets that are pushed to the Git repository can be shared by creating a new project from Git.

Requirements and restrictions

Required role

You need Admin or Editor role in the project to export a project.

Restrictions
  • You cannot export assets larger than 1024 MB from the projects user interface. To export assets up to 5 GB, use the asset command with the Cloud Pak for Data command line interface (cpd-cli).
  • If your project is marked as sensitive, you can't export data assets, connections or connected data from the project.
  • Be mindful when selecting assets to always also include the dependencies of those assets, for example the data assets or connections for a data flow, a notebook, connected data, metadata imports, data quality definitions and rules, or jobs. There is no check for dependencies. If you don't include the dependencies, subsequent project imports will not work.
  • You can only share assets across projects created in IBM Cloud Pak for Data. You cannot export a project from IBM Cloud Pak for Data and import the project into Cloud Pak for Data as a Service.
  • You cannot export job run metadata. Job run information is not available after you import.
  • Any connections that are created from a custom JDBC connector will not work if you import the project into a different Cloud Pak for Data instance. An administrator must delete the connection and re-create the custom JDBC connector in the new instance.

Exporting a project to desktop

Exporting a project packs the project assets that you select into a single ZIP file that can be shared like any other file.

To export project assets to desktop:

  1. Open the project you want to export assets from.

  2. Check whether the assets that you include in your export, for example notebooks or connections, don't contain credentials or other sensitive information that you don't want to share. You should remove this information before you begin the export. Only private connection credentials are removed. Note as well that any connections that were added from the connections catalog will be local to the project. The connections will no longer be platform-level connections.

  3. Optional. Create a text readme file with a list of the assets that you want to include in the export. Give a brief description of the analytics use case of the added assets and the data analysis methods that are used. Upload the text document as a data asset to the project.

  4. Click the Export to desktop icon from the project toolbar.

  5. Select the assets to add. You can filter by asset type or customize the project export settings by selecting preferences (the settings icon to the right of the window title) which are applied each time you export the project.

    Note that you can currently only select up to 1000 assets and the size of the export file can't exceed 1 GB in size.

  6. Optional: Change the name of the project export file.

  7. You must supply a password if:

    • You want to export connections that have shared credentials. Note that this password must be provided to decrypt these credentials on project import.
    • You want to export connections that have Mask sensitive credentials retrieved through API calls set to On. Without a password, connections that mask sensitive credentials retrieved through API calls are not exported.

  8. Click Export. Do not leave the page while the export is running.

    When you export to desktop, the file is saved to the Downloads folder by default. If a ZIP file with the same name already exists, the existing file isn't overwritten.

    Ensure that your browser settings download the ZIP file to the desktop as a .zip file and not as a folder. Compressing this folder to enable project import leads to an error. Note also that you cannot manually add other assets to an exported project ZIP file on your desktop.

    The status of a project export is tracked on the project's Overview page.

Synchronizing changes with Git

With Git integration enabled, you can synchronize changes between project assets and the Git repository in projects with deprecated Git integration.

Required role
You need Admin permissions on the project to export project assets to a Git repository.

To push project assets to a Git repository:

  1. In your project, click the Sync icon from the project toolbar.

  2. Select Pull and push to push project assets you to the Git repository. You can push incremental changes to the repository.

  3. Select your access token.

  4. Select the assets to add. You can filter by asset type or customize the project export settings by selecting preferences (the settings icon to the right of the window title) which are applied each time you export the project.

    Important:

    Notebook and Python script changes that you committed and pushed to the repository in JupyterLab, are not automatically synchronized. You must first explicitly select these assets to add them to the project synchronization. See Syncing project assets.

    Note that you cannot manually add other assets to the project assets that you exported to the Git repository.

Learn more

Importing a project

Parent topic: Administering projects