Deleting a Dataset from All Projects

When you delete a dataset using the dataset catalog, Tamr Core removes it from all projects. To remove a dataset from a project without deleting it from the system, see Removing a Dataset from a Project.

Before You Begin:

Before you delete a dataset, be sure to:

• Back up your Tamr Core instance.
• Identify all “downstream” projects that depend on the dataset by calling the /v1/datasets/{datasetId}/usage API endpoint. That is, before you delete the input dataset from one project, you must identify all other projects that use any dataset generated by Tamr Core for that project as an input dataset. See Report usages of a dataset by project steps and downstream datasets.
• Remove the dataset from all projects in which it is an input dataset.
• Update the unified datasets of all projects in which it was an input dataset.
• Delete the sample dataset that resulted from profiling the dataset.
• Remove the dataset from any projects or datasets that depend on it.
  Tip: the Tamr Toolbox includes a utility that gets a list of datasets and projects downstream from a specified dataset.
• Run jobs to update all downstream projects.

Tip: If the dataset you want to delete is a derived dataset for a project that has been deleted, a system administrator can use the Dataset Cleanup utility to find and convert these to source datasets for deletion.

Note: Deleting a unified dataset submits a job that can be resource-intensive to run. To monitor the progress of the job see Monitoring Job Status.

To delete a dataset from all projects:

1. At the top right of your browser, select Dataset Catalog. A list of all datasets appears.
2. Locate the dataset and select the checkbox that appears as you move your cursor over the left end of its row.
3. Select Delete. A message indicates whether it is possible to delete the dataset or if any preliminary action is required. To delete the dataset, select Delete.