管理MaxCompute作业

更新时间:
复制 MD 格式

The Jobs page in the MaxCompute console lets you monitor and manage job execution across your projects. Use it to track running and historical jobs, investigate failures, identify resource bottlenecks, and stop jobs consuming excessive resources.

Data developers can find a specific job, check its status, view execution details in LogView, or stop a runaway job.

Administrators can analyze resource load across quota groups at a point in time, identify which jobs are competing for compute resources, and adjust allocations accordingly.

What you can do

Action

How

Filter jobs by time range, project, quota, and job type

Use the global filters above the job list

Narrow results within the job list

Use the Instance ID, Job Owner, ExtNodeId, ExtNodeName, or Signature search fields above the list

Filter by status, priority, or diagnostic tag

Click the filter icon next to a column header

Sort by a single column

Click the sort button in a column header

Sort by multiple columns

Click Advanced Sorting, add columns, and set the order for each, then click OK

View execution details

Click LogView in the Actions column

Analyze resource consumption and quota allocation

Click Insights in the Actions column

Stop a running job

Click Terminate or Batch Terminate for jobs with Latest Status Running

By default, jobs are sorted by end time in descending order, with running jobs listed first.

Job status reference

Status

Description

Running

The job is currently executing and has not finished

Successful

The job completed without errors

Failed

The job encountered an error and did not complete

Waiting

The job is queued, waiting for compute resources

Canceled

The job was stopped before completion

Common O&M scenarios

Find and inspect a specific job

Use this approach when you need to audit a specific MaxCompute job or investigate a job triggered by an orchestration tool such as DataWorks.

  1. Log in to the MaxCompute console and select a region.

  2. In the left navigation pane, choose Observation O&M > Jobs.

  3. Set a Time Range and click Search.

  4. Above the job list, select ExtNodeId or Instance ID and enter the corresponding value.

  5. Click the filter icon to apply the filter.

  6. In the results, click LogView in the Actions column to view detailed execution information.

Stop a job consuming excessive resources

Use this approach to terminate a runaway job that is monopolizing compute resources.

  1. Log in to the MaxCompute console and select a region.

  2. In the left navigation pane, choose Observation O&M > Jobs.

  3. Set a Time Range and click Search.

  4. In the job list, click the filter icon next to Latest Status and select Running.

  5. Identify the target job, then click Terminate in the Actions column. To stop multiple jobs at once, select them and click Batch Terminate.

View accelerated query job details

Use this approach to check the status and execution details of SQLRT (Query Acceleration) jobs.

  1. Log in to the MaxCompute console and select a region.

  2. In the left navigation pane, choose Observation O&M > Jobs.

  3. Set Time Range to 1d and Job Type to SQLRT (Query Acceleration), then click Search.

  4. In the results, click LogView in the Actions column to view execution details.

For SQLRT jobs, multiple SQL commands may run within the same session, and one session corresponds to one instance ID. If some SQL commands have finished but others are still running, Latest Status shows Running. If the session expired or the interface was closed, Latest Status shows Cancelled.

Related topics