All Products
Search

This Product

    Data Lake Formation:Lifecycle management

    Document Center

    Data Lake Formation:Lifecycle management

    Last Updated:Aug 24, 2026

    Lifecycle management supports various rule types that allow you to manage the data lifecycle within your data lake, to help you save on storage costs. This topic describes how to manage lifecycle rules.

    How it works

    You can use lifecycle management to configure data management rules for databases and tables in your data lake. Based on four rule types—last access time of data, partition value by time, partition/table creation time, and last partition/table update time—you can periodically transition data to different storage classes to reduce storage costs. Data lakes often contain large amounts of historical data that are no longer actively used. You can transition this data to more cost-effective storage classes such as Infrequent Access, Archive, or Cold Archive. For example:

    • For an order table partitioned by date (for example, pt=20220101), you may only need to analyze data from the last three years. You can configure a rule based on the partition value to automatically transition older partitions to Cold Archive, which reduces storage costs.

    • If a database belongs to a discontinued business line, you can configure a rule to transition the entire database to Cold Archive for long-term storage.

    Important

    Configure rules carefully based on your actual business needs.

    • Data transitioned to Archive or Cold Archive is inaccessible to compute engines. To use the data again, you must manually restore it, which incurs fees. For more information, see OSS storage classes and Convert storage classes.

    • When data is transitioned to Infrequent Access, access performance is degraded for compute engines.

    Limitations

    • Lifecycle management requires Data Lake Formation (DLF 1.0), and the data must be stored in Object Storage Service (OSS).

    • Lifecycle management does not support unstructured data. To manage unstructured data, use the lifecycle management feature in OSS.

    Billing

    Lifecycle management involves two types of fees:

    • The lifecycle management feature is currently in public preview and is free of charge.

    • For information about OSS lifecycle fees, see Fees related to lifecycle rules.

    Procedure

    Prerequisites

    1. You must have Object Storage Service (OSS) activated. To do so, go to the OSS console.

    2. Access to databases and tables for lifecycle management is controlled by DLF data permissions. You can configure lifecycle rules only for databases and tables that you have permission to access.

    Create a lifecycle rule

    Follow these steps to create a lifecycle rule.

    1. Log on to the Data Lake Formation console.

    2. In the navigation pane on the left, choose Lake Management > Lifecycle Management.

    3. Click Create Rule to configure the rule.

      1. Specify the basic information: Name, Description, Catalog List, and Resource Type.

        For Resource Type, you can select Database or Table to configure lifecycle rules for databases or tables, respectively.

      2. Select a rule type. DLF currently supports the following four types:

        • Last access time of data: Defines the lifecycle based on the last access time of data. For a partitioned table, the last access time of the finest-grain partition is used. For non-partitioned tables, the last access time of the table is used.

        • Partition value by time: Defines the lifecycle based on the partition value. This rule applies only to tables whose level-one partition contains a time format.

        • Last partition/table update time: Defines the lifecycle based on the last update time of a partition or table. For a partitioned table, the last update time of the finest-grain partition is used. For non-partitioned tables, the last update time of the table is used.

        • Partition/table creation time: Defines the lifecycle based on the creation time of a partition or table. For a partitioned table, the creation time of the finest-grain partition is used. For non-partitioned tables, the creation time of the table is used.

      3. Select the time interval for transitioning data to Infrequent Access, Archive, or Cold Archive.

      4. Configure the rule execution mechanism.

        Enable Execution Scheduling to run the rule automatically every day. If scheduling is disabled, you can run the rule manually after creating it. Scheduled executions are completed daily before 08:00.

    4. Click Next to select the databases or tables to which the rule applies.

    5. Click Add Database Resources, select the resources to bind (search and cross-page selection are supported), and then click Add.

    6. After you add the resources, click OK to view the binding result.

      If the binding succeeds, a message displays the number of successfully bound resources. If it fails, a message shows the reason for the failure.

      Note

      • If the resource type is Database, you can bind database resources. If the resource type is Table, you can bind table resources.

      • Table rules have a higher priority than database rules. If a table is already bound by a database rule, this operation overwrites the existing rule for that table.

      • Each database or table can be bound to only one rule at a time.

      • Each rule can be bound to a maximum of 1,000 resources.

      • You can create a rule without binding resources and bind them later. To do so, click Save.

    Edit a lifecycle rule

    To edit a lifecycle rule, find it in the list and click Edit on the right.

    Important

    1. After you modify a rule, if Execution Scheduling is enabled, the changes take effect during the next day's execution.

    2. When a modified rule is executed again, it affects all its bound resources as follows:

      • Data already transitioned to Infrequent Access, Archive, or Cold Archive remains in its current state.

      • Data not yet transitioned is processed by the new rule.

    View rule details

    1. In the navigation pane on the left, choose Lake Management > Lifecycle Management.

    2. Select a rule and click its ID to view its details.

      • Basic Information: Displays rule details and the execution mechanism.

      • Resource Information: Lists the databases or tables bound to the rule.

      • Execution History: Lists manual and scheduled rule executions.

    Delete a lifecycle rule

    1. In the navigation pane on the left, choose Lake Management > Lifecycle Management.

    2. Find the lifecycle rule that you want to delete, click Delete on the right, and then click OK in the confirmation dialog box.

    Note

    • After a lifecycle rule is deleted, it can no longer be run manually or by the scheduler.

    • Data previously affected by the rule remains in its current state after the rule is deleted.

    Manually run a rule

    1. In the navigation pane on the left, choose Lake Management > Lifecycle Management.

    2. Find the lifecycle rule that you want to run manually, click Manual Execution on the right, carefully read the prompt, and then click OK to start the task.

    Important

    A manual execution runs immediately and affects all currently bound resources. This may impact your business applications. Assess the risks carefully before you proceed.

    View execution records

    1. In the navigation pane on the left, choose Lake Management > Lifecycle Management.

    2. Click the Execution History tab to view all historical lifecycle tasks and their execution logs.

    3. Click a task name to view its execution information and logs.

    Restore a table

    1. In the navigation pane on the left, choose Metadata > Metadata.

    2. Click the Data Table tab, and then click the name of the table.

    3. Click the Storage Rule tab. In DLF, you can click Restore Table to transition the storage class of the table to Standard Storage.

    4. For advanced restoration options, see the following documents: