All auto triggered tasks in DataWorks run on a resource group. Configuring the right resource group — and knowing when to allocate compute units (CUs) — keeps tasks running reliably and helps control costs.
Key concepts
Tasks are divided into two types based on whether they consume CUs during execution:
| Task type | Description | CU consumption | Task list |
|---|---|---|---|
| Scheduling task | Triggers and monitors computations in external compute engines, such as MaxCompute and Hologres. The resource group handles only lightweight coordination — no compute resources are consumed, so there is no need to scale compute size for these tasks. | None | Scheduling task configurations |
| Compute-optimized task | Executes task code directly within the DataWorks resource group. | Usage | Compute-optimized task configurations |
This topic covers CU configurations for Serverless resource groups only. If you use exclusive or subscription resource groups, specify the resource group — no CU configuration is needed. This topic also does not cover the resource consumption or billing of third-party engines, such as MaxCompute, called by tasks.
Prerequisites
Before you begin, ensure that you have:
-
An auto triggered task in DataWorks
-
Access to the task code editor and Scheduling Configuration pane
Configure resource properties
All auto triggered tasks require a scheduling resource group. The steps vary slightly depending on the task type and resource group type.
Step 1: Specify the scheduling resource group
This step is required for all auto triggered tasks.
-
In the task code editor, open the Scheduling Configuration pane on the right.
-
In the Resource Properties section, select a resource group from the Scheduling Resource Group drop-down list.
ImportantMake sure the selected resource group can connect to the data sources and compute resources your task accesses. For details, see Network connection configurations.
Step 2: Configure compute resources
Applies to: Compute-optimized tasks on a Serverless resource group only. Skip this step if you are using a scheduling task or a non-Serverless resource group.
After you select a Serverless resource group in Step 1, the Scheduling CUs item appears automatically.
Set the number of CUs required at runtime based on the task's complexity and performance requirements, such as 0.25 CU, 1 CU, or 4 CUs.
Step 3: Configure an image
Applies to: PyODPS and Shell tasks only.
Select an Image to define the runtime environment for the task. For details on managing images, see Custom images.
Set the default scheduling resource group
When you create a task, DataWorks uses the scheduling resource group specified in Scheduling Settings as the default. To change the default, click the 
icon in the lower-left corner, then go to Settings > Scheduling Settings.
Billing
Running auto triggered tasks on a Serverless resource group incurs two types of fees:
| Fee type | When charged | Details |
|---|---|---|
| Task scheduling fees | Based on the number of successfully run instances per day | Task scheduling billing |
| Resource usage fees | Applies only to compute-optimized tasks on a pay-as-you-go Serverless resource group; billed by CUs consumed and runtime duration; excludes third-party engine charges | Serverless resource group billing |
For data integration tasks, resource consumption is determined by the CUs to Occupy setting, not the Scheduling CUs configuration.