Data Lake Formation:Get started with DLF

Prerequisites

Before you begin, ensure that you have:

• Completed DLF setup (activation and authorization, required once).

• (RAM users only) The following permissions:

  • API permissions: The AliyunDLFFullAccess policy, or a policy that includes catalog-related authorization actions. See RAM authorization action reference.

  • Data permissions: The super_administrator or admin system role, or a custom role with catalog-related permissions. See Configure data permissions.

Create a catalog

A catalog is the top-level container for metadata in your data lakehouse. When creating a catalog, choose a storage redundancy type based on your availability requirements: LRS for single-zone storage at lower cost, or ZRS for cross-zone replication at higher availability.

1. Log on to the DLF console.

2. On the Catalogs page, click Create Catalog.

3. Configure the following parameters.

   Parameter	Description
   Catalog Name	A unique name for the catalog.
   Description	A description of the catalog.
   Storage Type	Fixed to Standard Storage.
   Storage Redundancy Type	The redundancy policy for catalog metadata. See the following table.

   Storage redundancy options:

   Option	Behavior	Default
   LRS (Locally Redundant Storage)	Stores metadata in a single zone. If the zone becomes unavailable, the catalog is inaccessible.	Yes
   ZRS (Zone-Redundant Storage)	Replicates metadata across multiple zones within a region for higher availability. Costs more than LRS.	No

   Important: After a catalog is created, you cannot change the redundancy type from ZRS to LRS.

4. Read and select the Terms of Service, then click Create Catalog.

For more information, see Manage catalogs.

Ingest data into your data lakehouse

After creating a catalog, sync raw data into your data lakehouse using tools such as Flink CDC and DataWorks Data Integration.

Analyze data in your data lakehouse

Query and extract insights from your data lakehouse using the following engines:

• EMR Serverless Spark — batch read and write operations

• Realtime Compute for Apache Flink — streaming read and write operations

• EMR Serverless StarRocks — extract insights from data

Next steps

• Manage catalogs — create, modify, and delete catalogs

• Configure data permissions — set up roles and access control

• RAM authorization action reference — review available permission actions