Редактиране

Quickstart: Get data into OneLake

OneLake is the single, unified data lake for Microsoft Fabric. Every Fabric workload reads and writes data through OneLake, so you only need to load data once to use it everywhere. You can bring data into OneLake in several ways:

  • Upload files directly to a lakehouse or warehouse.
  • Ingest data by using pipelines, dataflows, or streaming experiences.
  • Connect to external data by using shortcuts or mirroring.

In this quickstart, you bring data into OneLake two ways: you upload a CSV file to a lakehouse, and you create a OneLake shortcut from a second lakehouse that points back to the same data without copying it. When you finish, you have a queryable Delta table and a shortcut, both available to every Fabric engine through OneLake.

Prerequisites

Create a lakehouse

When you create a Fabric item, such as a lakehouse, warehouse, or eventhouse, that item provisions storage in OneLake on your behalf. In this quickstart, you make a lakehouse, which gives you both a file area (Files) for unstructured or semi-structured data and a Delta table area (Tables) for structured, queryable data. Everything you put in either area is stored in OneLake and immediately accessible to other Fabric workloads.

  1. Sign in to the Fabric portal and select your workspace.

  2. Select New item.

  3. In the New item pane, search for and select Lakehouse.

  4. Enter a name, such as DataLakehouse, then select Create.

    The lakehouse opens to the Explorer view, which shows empty Tables and Files sections. Both sections are already backed by OneLake and are ready for content.

Upload sample data

In this quickstart, you use Dim_Products.csv from a publicly available Fabric sample dataset. It's a small table of product information from a sample coffee retailer.

  1. Open a browser and go to https://fabrictutorialdata.blob.core.windows.net/sampledata/Coffee/Dim_Products.csv.
  2. When prompted, save the file as Dim_Products.csv to a folder on your computer.

In this section, you upload Dim_Products.csv to Files so that you have raw source data sitting in OneLake. The Files area of a lakehouse is a general-purpose storage zone in OneLake. Think of it as the landing zone for raw data in whatever format it arrives. You can drop in CSV, JSON, Parquet, images, logs, or anything else without having to define a schema first.

  1. In the lakehouse Explorer, hover over Files, select the more options (...) menu, and then select Upload > Upload files.

  2. In the Upload files pane, select the folder icon and browse to Dim_Products.csv on your computer.

  3. Select Upload, then close the upload pane.

  4. Select the Files folder to see its contents and confirm that Dim_Products.csv appears.

  5. Select Dim_Products.csv to see its data.

    A screenshot of the Fabric portal that shows unstructured csv data in the Files section of a lakehouse.

The file now lives in OneLake, but as a raw CSV it isn't yet something SQL or Spark can query as a table.

Load the file into a Delta table

Fabric standardizes on Delta Lake as the table format in OneLake. When you load a file into the Tables area, Fabric reads the source file, infers a schema, and writes the data out as a Delta table. From that point on, every Fabric engine can query the same table without you copying or converting the data again.

  1. In the lakehouse Explorer, open the Files folder.

  2. Hover over the Dim_Products.csv file and select the more options (...) menu, then select Load to Tables > New table.

  3. In the Load to table dialog, enter dim_products for the table name, keep the defaults, and select Load.

  4. After the load finishes, expand Tables and select dim_products to preview the rows. The raw CSV in Files is unchanged, and dim_products is a new Delta table built from it.

    A screenshot that shows structured Delta table data in the Tables section of a lakehouse.

  5. Hover over dim_products and select the more options (...) menu, then select Properties.

    The Properties screen shows the various details for the table, including the URL and Azure Blob File System (ABFS) path that you can use to reference this table in other engines.

Reuse the data with a shortcut from a second lakehouse

Uploading and loading is one way to get data into OneLake. The other key pattern is to reference data that already exists somewhere else, without duplicating it. That's what a shortcut is: a pointer in OneLake that references data stored in another lakehouse, in another Fabric workspace, or in supported sources outside of Fabric like Azure Data Lake Storage or Amazon S3. The data isn't copied; it stays in the source location, but you can read it through OneLake as if it were local. Any updates to the source are immediately visible through the shortcut, so you don't have to maintain copies of the data.

In this section, you create a second lakehouse and add a shortcut from it back to the dim_products table in your first lakehouse. This reflects how teams typically work, where one team owns the curated data and other teams or projects consume it through shortcuts in their own workspaces.

  1. In your workspace, select New item.
  2. In the New item pane, search for and select Lakehouse.
  3. Enter a name, such as ShortcutLakehouse, then select Create.
  4. In the new lakehouse's Explorer, hover over Tables, select the more options (...) menu, and then select New shortcut.
  5. On the New shortcut page, under Internal sources, select Microsoft OneLake.
  6. In the data source browser, select the first lakehouse that you made for this quickstart, and then select Next.
  7. Expand Tables, select the dim_products table, and then select Next.
  8. Review the selection and select Create.
  9. Expand Tables in ShortcutLakehouse and confirm that dim_products appears with a shortcut icon (a small link image over the table icon). Select it to preview the rows. The table is the same as in the original lakehouse, but no data was copied.
  10. Hover over the dim_products table, select more options (...), then select Manage Shortcut. On the Manage shortcut pane, you can view the shortcut details, including the shortcut target where the original data is stored.

Clean up resources

If you don't plan to continue to the other OneLake quickstarts, delete the lakehouses to avoid OneLake storage charges against your Fabric capacity.

  1. In your workspace, hover over the lakehouse that you want to delete.
  2. Select the more options (...) menu next to the lakehouse, select Delete, and confirm the deletion.

Deleting the lakehouses also removes the contents within them: the uploaded file, the dim_products Delta table, and the shortcut.

Related content

  • Last updated on