Managing Jobs

A guide for administrators and power users who need to enable, configure, or retire Jobs after they've been installed.

From this point forward we assume the Job code is already present in Nautobot.
The sections below walk through the operational tasks you perform after installation—enabling or hiding a job, overriding its metadata, steering it to Celery or Kubernetes queues, and eventually deleting or archiving the record.
Required permissions for each action are listed at the end of the page.

Job record vs Job class

A Job class is the Python code on disk; a Job record is the database row that stores its metadata and enabled state.

Editing code updates the class immediately, but the record keeps the previous values until you refresh it (e.g., with nautobot-server post_upgrade or a Git repo resync).
Conversely, overriding a field in the record never changes the source code—clear the override flag to revert to whatever the class defines.

Enabling and disabling Jobs

When a new Job record is created for a newly discovered Job class, it defaults to enabled = False, which prevents the Job from being run unintentionally. This provides security and oversight for new installations.

Important

When upgrading from a Nautobot release before 1.3 to Nautobot 1.3.0 or later, Jobs previously run or scheduled will automatically have their records set to enabled = True.

To enable or disable a Job:

1. Navigate to Jobs > Jobs.
2. Select a Job from the list.
3. Click the Edit button.
4. In the Job section, toggle the Enabled checkbox as needed.
5. Click Update.

Overriding Job metadata

Sometimes you need to change how a Job appears or behaves without touching its source code—for example, hiding a vendor‑specific Job from general users or increasing the timeout for a long‑running audit Job. You can accomplish this by overriding metadata attributes directly on the Job record.

Editable attributes

Attribute	Purpose	Typical use‑case
grouping	UI category the Job is listed under.	Organize third-party Jobs into "Compliance" or "Local Automations."
name	Display name in the Jobs list.	Rename technical Job names to user-friendly descriptions.
description	Markdown or HTML description shown on Job detail page.	Add instructions or documentation links.
approval_required	Require an additional user approval before each execution.	Protect Jobs that perform bulk production changes.
dryrun_default	Default state of Dry-Run checkbox.	Audit or reporting Jobs that rarely commit changes.
has_sensitive_variables	Avoid storing sensitive input data.	Jobs that accept credentials or sensitive data.
hidden	Hide Job from the main Job listings.	Helper Jobs executed only by buttons or hooks.
soft_time_limit / time_limit	Override Celery task soft/hard kill timeouts (in seconds).	Long-running imports or external API syncs.
job_queues	Restrict Job execution to specific Job Queues.	Route intensive Jobs to isolated or Kubernetes workers.
is_singleton	Prevent multiple simultaneous executions.	Inventory synchronization Jobs.

How to set or clear an override

1. Navigate to Jobs > Jobs and select the Job.
2. Click Edit.
3. Tick the *_override box beside the field.
4. Set the desired new value, then click Update.
5. A blue pencil icon indicates overridden fields in the Jobs list.

To revert to the original Job class value, clear the override checkbox and save. On next refresh, the original value defined in the Job class will reapply.

Note

Overrides only apply in the database and never update Job source code.

Best‑practice tips

• Combine edits: If enabling a Job, also set metadata overrides (e.g., grouping) simultaneously for cleaner change history.
• Document changes: Clearly document why overrides are set in the description field to help future administrators.
• Use queues: If Jobs consistently exceed default time limits, consider separate queues instead of global limit increases.
• Periodic audits: After major upgrades, review overridden Jobs (blue pencil icons) and consider removing unnecessary overrides.
• Limit overrides: If multiple overrides accumulate, consider adjusting defaults directly in Job code instead.

Job Queues (v2.4+)

As of Nautobot v2.4, the task_queues attribute is deprecated. Use job_queues, which references Job Queues. Queues appear sorted alphabetically.

Assigning Job queues (Celery or Kubernetes)

Jobs in Nautobot can execute on Celery (default) or Kubernetes backends. You manage routing with Job Queues.

Each Job has two queue configuration options:

• Default Job Queue: Used when no queue is specified at runtime.
• Eligible Queues: Limits queue selection options for users when running the Job.

Common scenarios:

• Default Celery worker: Leave Eligible Queues blank to always use default Celery.
• Kubernetes execution: Create a Kubernetes Job Queue (Jobs > Job Queues), add it to Eligible Queues, and optionally set as default.
• Dedicated resource queue: Create a dedicated Celery queue (high-resource-jobs) for CPU-intensive tasks, then assign these Jobs to that queue.

Tip

Refer to Job Queues and Kubernetes Job Support for detailed queue setup instructions.

Viewing installed vs uninstalled Jobs

By default, Nautobot shows only installed Jobs (with corresponding Python code present).

To view uninstalled Jobs (Job records remaining without corresponding Python code):

1. Go to Jobs > Jobs.
2. Click Filter.
3. Set Installed to No, then click Apply.

Uninstalled Jobs appear greyed-out and cannot run or schedule executions, but historical data (JobResult, ScheduledJob) is retained for audit.

Deleting or archiving Jobs

Use this procedure when permanently retiring a Job:

• Delete Job record: Removes Job from Nautobot UI and API. Associated Job Buttons, Hooks, and Scheduled Jobs are automatically disabled. Code files are unaffected.
• Remove code only (keep record): The Job appears as installed = False (see above) but remains in the database.

Impact on JobResults and ScheduledJobs

• Existing JobResults and ScheduledJobs retain historical data but their foreign-key references to the Job become NULL.
• Reinstalling the same Job does not automatically reconnect historical data.

Recommended deletion steps

1. Disable the Job.
2. Verify no active ScheduledJobs reference it (ScheduledJobs > Filter by Job name).
3. Delete the Job (Actions > Delete).
4. Separately remove code files or Git commits.
5. Optionally, remove orphaned JobResults via Admin > Job results.

Permissions checklist

Task	Required permission(s)
Enable/Disable Job	extras.change_job
Override metadata	extras.change_job
Assign Job Queues	extras.change_job + permissions on specific Job Queues
Delete Job	extras.delete_job
Run Job	extras.run_job
Approve scheduled Job	extras.approve_job + (extras.change_scheduledjob or extras.delete_scheduledjob)