Managing CloudTrail Lake federation resources with Amazon Lake Formation - Amazon CloudTrail
Services or capabilities described in Amazon Web Services documentation might vary by Region. To see the differences applicable to the China Regions, see Getting Started with Amazon Web Services in China (PDF).

Managing CloudTrail Lake federation resources with Amazon Lake Formation

When you federate an event data store, CloudTrail registers the federation role ARN and event data store in Amazon Lake Formation, the service responsible for allowing fine-grained access control of the federated resources in the Amazon Glue Data Catalog. This section describes how you can use Lake Formation to manage the CloudTrail Lake federation resources.

When you enable federation, CloudTrail creates the following resources in the Amazon Glue Data Catalog.

  • Managed database – CloudTrail creates 1 database with the name aws:cloudtrail per account. CloudTrail manages the database. You can't delete or modify the database in Amazon Glue.

  • Managed federated table – CloudTrail creates 1 table for each federated event data store and uses the event data store ID for the table name. CloudTrail manages the tables. You can't delete or modify the tables in Amazon Glue. To delete a table, you must disable federation on the event data store.

Controlling access to federated resources

You can use one of two permissions methods to control access to the managed database and tables.

  • IAM only access control – With IAM only access control, all users in the account with the required IAM permissions are given access to all Data Catalog resources. For information about how Amazon Glue works with IAM, see How Amazon Glue works with IAM.

    On the Lake Formation console, this method appears as Use only IAM access control.

    Note

    If you want to create data filters and use other Lake Formation features, you must use Lake Formation access control.

  • Lake Formation access control – This methods provides the following advantages.

    • You can implement column-level, row-level, and cell-level security by creating data filters.

    • Database and tables are only visible to Lake Formation administrators and creators of the database and resources. If another user needs access to these resources, you must explicitly grant access by using Lake Formation permissions.

For more information about access control, see Methods for fine-grained access control.

Determining the permissions method for a federated resource

When you enable federation for the first time, CloudTrail creates a managed database and managed federated table using your Lake Formation data lake settings.

After CloudTrail enables federation, you can verify which permissions method you are using for the managed database and managed federated table by checking the permissions for those resources. If the ALL (Super) to IAM_ALLOWED_PRINCIPALS setting is present for the resource, the resource is managed exclusively by IAM permissions. If the setting is missing, the resource is managed by Lake Formation permissions. For more information about Lake Formation permissions, see Lake Formation permissions reference.

The permissions method for the managed database and managed federated table can differ. For example, if you check the values for the database and table, you could see the following:

  • For the database, the value that assigns ALL (Super) to IAM_ALLOWED_PRINCIPALS is present in the permissions indicating that the you're using IAM only access control for the database.

  • For the table, the value that assigns ALL (Super) to IAM_ALLOWED_PRINCIPALS not present, which indicates access control by Lake Formation permissions.

You can switch between access methods at any time by adding or removing ALL (Super) to IAM_ALLOWED_PRINCIPALS permission on any federated resource in Lake Formation.

Cross-account sharing using Lake Formation

This section describes how to share a managed database and managed federated table across accounts by using Lake Formation.

You can share a managed database across accounts by taking these steps:

  1. Update the cross-account data sharing version to version 4.

  2. Remove Super to IAM_ALLOWED_PRINCIPALS permissions from the database if present to switch to Lake Formation access control.

  3. Grant Describe permissions to the external account on the database.

  4. If a Data Catalog resource is shared with your Amazon Web Services account and your account is not in the same Amazon organization as the sharing account, accept the resource share invitation from Amazon Resource Access Manager (Amazon RAM). For more information, see Accepting a resource share invitation from Amazon RAM.

After completing these steps, the database should be visible to the external account. By default, sharing the database does not give access to any tables in the database.

You can share all or individual managed federated tables with an external account by taking these steps:

  1. Update the cross-account data sharing version to version 4.

  2. Remove Super to IAM_ALLOWED_PRINCIPALS permissions from the table if present to switch to Lake Formation access control.

  3. (Optional) Specify any data filters to restrict columns or rows.

  4. Grant Select permissions to the external account on the table.

  5. If a Data Catalog resource is shared with your Amazon Web Services account and your account is not in the same Amazon organization as the sharing account, accept the resource share invitation from Amazon Resource Access Manager (Amazon RAM). For an organization, you can auto accept using RAM settings. For more information, see Accepting a resource share invitation from Amazon RAM.

  6. The table should now be visible. To enable Amazon Athena queries on this table, create a resource link in this account with the shared table.

The owning account can revoke sharing at any point by removing permissions for the external account from Lake Formation, or by disabling federation in CloudTrail.