Help Center/ DataArts Studio/ User Guide/ DataArts Factory/ Developing a Notebook Job
Updated on 2026-05-20 GMT+08:00
View PDF
Share

Developing a Notebook Job

Currently, the notebook function is restricted. To use it, contact customer service or technical support.

Currently, this function is available only in the Jakarta region. This function is subject to the ECS resources in the current region.

Currently, the notebook function is in the OBT phase of the first version. It is recommended that you use it for personal development and testing only.

When notebook jobs are interconnected with MRS Spark, individual-based fine-grained data permission control is not supported.

Notes and Constraints

You can configure a maximum of 100,000 jobs.

The project directory (job tree) can store notebook jobs of up to 1 MB (including execution results).

Creating a Notebook job.

  1. Log in to the DataArts Studio console by following the instructions in Accessing the DataArts Studio Instance Console.
  2. On the DataArts Studio console, locate a workspace and click DataArts Factory.
  3. In the left navigation pane of the DataArts Factory console, choose Data Development > Develop Job.
  4. Right-click the job directory tree and select New Notebook Job.
    You can also click Create Now or New Notebook Job shown in the following figure to create a notebook job.
    Figure 1 Creating a notebook job
  5. Enter a job name, select the directory of the job, and enter a job description.

    The job name is case-insensitive.

    The job name can contain only letters, digits, hyphens (-), underscores (_), and periods (.).

  6. Click OK. Access the notebook job development page and perform subsequent operations.

Developing a Notebook Job

On the notebook job development page, enter code, run the cell, and view the execution result.

Figure 2 Developing a notebook job
  • You can create multiple cells and edit the title of each cell.
  • Python is displayed in the upper right corner of each cell. The code type is Python.
  • The cell operations of notebook jobs in the project directory (job directory tree) are the same as the operations on notebook jobs in the notebook environment directory.
  • Click the default Python engine in the upper right corner to view information about the created compute engines. You can choose the compute engine you need. A notebook job supports only one compute engine. After you select a compute engine, all cells in the notebook job use this compute engine.
  • For details about how to bind compute resources, see Binding Compute Resources.

The following table lists the function buttons for notebook jobs.

You can use them to perform operations on notebook jobs.

Figure 3 Common function buttons
Table 1 Common functions

Button

Description

Run All

Stop All

Runs all cells.

Stops all cells.

Save

Saves a notebook job.

You can click the Versions tab on the right to view the saved version. You can roll back, view, and compare versions.

NOTE:

If you select only one version when comparing versions, the selected version will be compared with the content in development state. If you select two versions, they will be compared.

Submit

Submits a notebook job version. The job can be scheduled only after a version is submitted.

When you click Submit, the system displays the message "Not scheduling. This version will be executed when you click Execute."

You can click the Versions tab on the right to view the submitted version. You can roll back, view, and compare versions.

NOTE:

If you select only one version when comparing versions, the selected version will be compared with the content in development state. If you select two versions, they will be compared.

Unlock

Lock

If a notebook job is locked, you can unlock it.

If a notebook job is locked by other users, you can obtain the lock.

Execute

Stop Scheduling

After submitting a notebook job, click Execute. The submitted version will take effect in the scheduling period.

You can click Stop Scheduling to stop the job.

Manual scheduling

Stop Manual scheduling

Manually schedules a notebook job (tests the job).

Stops a notebook job (stops testing the job).

New Cell

Creates a cell under an existing one.

Clear Outputs

Clears the execution results of all cells.

More > Restart Kernel

Restarts the Notebook kernel.

More > Kill Kernel

Stops the Notebook kernel.

More > Monitor

Switches to the O&M monitoring page to view job monitoring information (job instances, job monitoring data, and job attributes).

Click a node name to view the monitoring information and attributes of the node.

Click the Instance DAG tab to view the job dependency graph. Two display modes are supported: DAG view and list view.

If there are too many nodes in the DAG, you can select the list view to quickly filter and check them.

For details about job scheduling, see Monitoring a Batch Job.

More > Export

Exports the current notebook job.

Export Mode: To a local path or To OBS

NOTE:

If you select To OBS, you need to select an OBS path.

Export Scope: Export jobs only or Export jobs and their dependency scripts and resource definitions

Job Status: Developed or Submitted

NOTE:

If you select Submitted and no versions have been submitted for the job to be exported, the job will be skipped.

Python

By default, the Python engine is used, which does not depend on compute resources.

Click the default Python engine in the upper right corner to view information about the created compute engines. You can choose the compute engine you need. A notebook job supports only one compute engine. After you select a compute engine, all cells in the notebook job use this compute engine.

Connection status icon in the upper right corner

Green: Connected, indicating that online debugging is running properly.

Red: Unavailable, indicating that online debugging debugging is disconnected and unavailable. Try reconnecting the kernel to restore the connection.

Configuring the Job

Basic information: For details, see Configuring Basic Job Information.

Job parameters: For details, see Configuring Job Parameters.

You can perform parameter calculation for a notebook job in the next cell of the parameters cell. Do not perform it in the parameters cell. The following figure shows an example.

Figure 4 Job parameter calculation example 1
Figure 5 Job parameter calculation example 2

Job parameters are strings and must be converted before calculated.

For example, if a parameter is set to test and the calculation test + 5 is required, the parameter type needs to be converted. The following is an example.
Figure 6 Job parameter type conversion for calculation

For details about dependency packages, see Exporting Dependency Packages.

Configuring Scheduling

Click the Schedule Config tab on the right to configure the scheduling parameters for the notebook job.

For details, see Setting Up Scheduling for a Job.

When notebook jobs are interconnected with MRS Spark, individual-based fine-grained data permission control is not supported.

Submitting a Version

After saving the job, you need to submit a version before executing it.

For details about submitting a version, see Submitting a Version.

Viewing Job Scheduling

After the notebook job starts, choose Job Monitoring in the navigation pane on the left and click the Batch Jobs tab to view the scheduling status of the job. You can click in front of the job name to view the status of the latest instance or run log.

You can click the job name to view the instances, attributes, and parameters of the job. You can click a node name to view the monitoring information of the node.

For details about job scheduling, see Monitoring a Batch Job.

Parent topic: DataArts Factory