> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.umnai.com/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.umnai.com/_mcp/server.

# Ingestion jobs

An ingestion job imports files from source storage and creates a datasource. Prepare a [storage connector](/ui-guides/setup/storage-connectors), [schema](/ui-guides/data/schemas), [ingestion configuration](/ui-guides/data/ingestion-configuration) and processing compute configuration before creating the job.

## Before you begin

Confirm that the connector can read the source files and that the ingestion configuration references the correct schema. Prepare a [processing compute configuration](/ui-guides/setup/compute-configurations). Decide whether this run creates a new datasource or adds an increment to an existing one.

## Create an initial ingestion job

An initial ingestion job creates a new datasource. Open **Datasources > Ingestion Jobs**, then select **Create**. Select the configuration and source location for the data you want to import, using the fields below.

| **Field**               | **What to enter or select**                                                       |
| ----------------------- | --------------------------------------------------------------------------------- |
| Job name                | A descriptive name for this import job                                            |
| Job type                | Initial for the first import that creates this datasource                         |
| Datasource name         | The name for the output datasource, for example adult-income-datasource           |
| Ingestion configuration | The configuration associated with the intended data schema                        |
| Storage connector       | The connector for the bucket holding your source files                            |
| Data path               | The bucket-relative path of the data to import, consistent with the chosen schema |
| Compute configuration   | A processing compute configuration for the ingestion job                          |
| Maximum runtime seconds | The runtime limit in seconds; the example uses 3600 (60 minutes)                  |

1. Enter the job name and output datasource name, and select **Initial** for the job type.

2. Select the ingestion configuration and storage connector, then enter the bucket-relative data path. Check that these refer to the intended schema and source files.

3. Select a processing **Compute configuration** and set **Maximum runtime (seconds)**. Enter **3600** to allow 60 minutes.

4. Select **Create** to submit the job for asynchronous processing.

### Review the selected inputs

Complete the required inputs before creating the job.

![Initial ingestion job form before completion](/_fern-files/_img/f6f4f601c87188f42933c93db931050df0c223495da89ebafc94e98550be70b8.webp)

![Compute configuration choices and runtime in seconds](/_fern-files/_img/1ecb9e1cfaab11cce3c9f4642b87efbab236900874e320c4bc80f260efd8ce61.webp)

## Monitor the job and check the datasource

Ingestion jobs run asynchronously. Open the job's **Overview** and inspect **Status**, **Progress** and **Stage** under **Execution status**. The example below shows **COMPLETED**, **100%** and **Persistence**. Use **Job details** to verify the source path, ingestion configuration, compute configuration, target datasource and maximum runtime. If the job fails, inspect its error details before submitting another run.

Return to the **Datasources** tab and locate the output by its datasource name. Check that its status is **AVAILABLE** before using it to create a dataset. The list also displays **Increment**, **Created at**, and **Updated at**.

![Existing completed ingestion job with configuration references](/_fern-files/_img/7947b125529aa37b24709b6c5dac5c3f32e0a523f9d44e146e1d6753f08e6a75.webp)

![Existing datasources with AVAILABLE status](/_fern-files/_img/a82c99a3f557351f49ab06aa408649b9b78c0f0f7b33d14ef4e0c172cbd782bc.webp)

## Add an increment to a datasource

1. Open **Ingestion Jobs > Create** and enter a job name.
2. Set **Job type** to **Increment** and select the existing datasource. Use the existing datasource's schema when preparing the additional files.
3. Supply an **Increment label** if you need a descriptive label for this update.
4. Select the storage connector, enter the path to the additional data and choose the processing compute configuration.
5. Set the runtime in seconds, review the existing datasource selection and select **Create**.
6. Monitor the new job, then inspect the datasource and its increment information after completion.

Changing job type clears selections specific to the previous type. Review the form again after switching between Initial and Increment. An increment targets an existing datasource; it does not create a separately named initial datasource.

## Next steps

Review the [datasource](/ui-guides/data/datasources), then prepare its [onboarding configuration](/ui-guides/data/onboarding-configuration).