- Releases Notes and Announcements
- Product Introduction
- Purchase Guide
- Getting Started
- Operation Guide
- SQL Developer Guide
- Python Developer Guide
- ETL Developer Guide
- FAQ
- Contact Us
- Releases Notes and Announcements
- Product Introduction
- Purchase Guide
- Getting Started
- Operation Guide
- SQL Developer Guide
- Python Developer Guide
- ETL Developer Guide
- FAQ
- Contact Us
You can perform the following four operations on a job in the Stream Compute Service console: Publish, Run, Stop, and Create copy. Some of these operations can be performed in batches. This document describes these operations in detail.
Publish
Description
After a job draft is developed, click Save and Publish draft to publish it as an online version. After the job is published successfully, depending on the job status, you can run a published version or stop an online running version of it.
Directions
After a job draft is developed, click Save and Publish draft to publish it as an online version. The version number of a new version is automatically generated by the system, but you need to enter a version description.
If no job version is running online, directly run the published version.
If a job version is running, stop it first and select Create a checkpoint when stopping the job, if necessary, in the pop-up window. After the job is completely stopped, run the new version.
Run
Description
Run corresponds to Stop. It means starting a new job running instance. To run a running job, stop the online running version first, and then perform Run.
Directions
You can run a Stopped job using one of the following methods.
1. In the Stream Compute Service console, click Jobs > Operations > More, and select Run in the drop-down list. The job status will become Operating. After a moment, the job status will become Running, indicating that the job is successfully started.
2. Publish and run a new version of the job. The job status will become Operating. After a moment, the job status will become Running, indicating that the job is successfully started.
Resuming a job from a checkpoint
When running a job, you can select Use a checkpoint.
This option is unavailable to a job without checkpoints.
Note
A cluster of an old version does not support selecting a checkpoint. To use this feature, submit a ticket to upgrade your cluster first.
Notes
If any exception occurs when you are operating a job, the job will go back to its actual status. For example, if the operation of stopping a job is aborted, the job status will be Stopped; if the operation fails and the job is still running, the job status will be Running. If this is the case, a triangle with an exclamation point will be displayed on the right of the status field, which will show the error message when you move the pointer over it.
Operation suggestions will be provided in common error messages.
If an error message is not easy for you to understand, please contact us. We will continuously improve error messages.
Note
When a job is running, modify the configurations of upstream and downstream products with caution, including but not limited to deleting or adding topics of CKafka data sources and sinks, and locking tables, modifying table structure, adding constraints, and stopping operations on MySQL data sources and sinks. Otherwise, the running job may be affected, resulting in incomplete data or job exception.
Stop
Description
Stop means stopping the execution of a job. When performing this operation, you need to decide whether to save all runtime status data.
Directions
You can stop a Running job using one of the following methods.
1. In the Stream Compute Service console, click Jobs > Operations > More, and select Stop in the drop-down list. The job status will become Operating. After a moment, the job status will become Stopped, indicating that the job is completely stopped.
2. Publish a new version, and stop the current running version. The job status will become Operating. After a moment, the job status will become Stopped, indicating that the job is completely stopped.
Note
If you want to retain the current state data of the job and make the job resume when it is started next time, select Create a checkpoint when stopping the job.
Create copy
Description
Job copies can be created quickly to facilitate job migration or development of similar jobs.
Directions
Creating a job copy
1. Go to the job details page, and select Create copy in the drop-down list of Job operations in the top right corner.
2. Click Create copy, and then select a target directory and target cluster and enter a copy name in the pop-up window.
3. Click Confirm. You will see the result shown in the pop-up window.
4. Click Job name to go to the details page of the created job.
Creating job copies in batches
1. Click Batch on the top of the job list, and select Create copy.
2. Select the jobs for which you want to create a copy.
3. Click Create copy, and then select a target cluster in the pop-up window. In this operation, you cannot select a target directory. The directory of a copy will be the same as that of the source job.
4. Click *Confirm**. You will see the result (success or failure) in the pop-up window.
Note
If the target cluster does not support the Flink version of the source job, the Flink version of the job copy will be changed to the default Flink version of the target cluster.
If the source job uses fine-grained resources, but the target cluster does not support such resources, both the JobManager and TaskManager specs of the job will be set to 1 CU.
Jobs can be copied only within the workspace where they reside, and a target cluster must be one associated with the workspace.
Checkpoints of source jobs will be synced (by sharing the same checkpoint path) to new jobs created by copying, but only checkpoints manually triggered will be synced.
Yes
No
Was this page helpful?