Managing Batch Jobs

The Batch Jobs page lets you monitor, inspect, and manage all batch job executions across your accessible projects.

Job List

The Batch Jobs page shows all jobs for your accessible projects:

Column	Description
Name	Auto-generated job identifier
Notebook	Path of the notebook that was executed
Status	Current state: pending, running, succeeded, failed
Project	Owning project
Started	Timestamp when the Kubernetes Job started
Completed	Timestamp when the job finished
Actions	View logs, view output, resubmit, delete

Viewing Logs

Click Logs on any job to stream output from the Papermill execution. Logs show:

• Notebook cell execution progress
• Standard output from print statements
• Error tracebacks if the notebook fails

Executing notebook with kernel: python3
Executing cell 1/24
Executing cell 2/24
...
Notebook saved to /home/jovyan/work/shared/batch_outputs/abc123/output.ipynb

Viewing Output

Successful jobs write the executed notebook to the shared workspace at:

/home/jovyan/work/shared/batch_outputs/<jobID>/output.ipynb

You can access it in two ways:

1. In JupyterLab — open the file browser in any notebook and navigate to shared/batch_outputs/<jobID>/
2. Via Shared Files — click View Output in the job row to browse the output directory

The output notebook contains all cell outputs, including charts, tables, and print statements, captured at execution time.

Failed Jobs

When a job fails, the error is captured and shown in the job detail view. Common causes:

Error	Likely Cause
Notebook not found	The notebook path doesn’t exist in the shared workspace
Kernel error	Python exception in a notebook cell — check logs for traceback
OOMKilled	Notebook exceeded memory limit — increase memory request
DeadlineExceeded	Job ran longer than expected — optimize notebook or increase timeout

To retry a failed job, click Resubmit.

Deleting Jobs

Click Delete to remove a job record. This does not delete the output notebook from the shared workspace.

Note

Kubernetes Job resources are automatically cleaned up 24 hours after completion. The job record in Menatic AI persists until manually deleted.

Filtering by Project

If you have access to multiple projects, use the project filter dropdown to narrow the job list to a specific project.