Workflow Jobs

Monitor AWX workflow job executions via the API

Overview

Workflow jobs are instances of workflow job templates that have been executed. They contain multiple workflow job nodes, each representing a job, project update, or inventory update.

Endpoints

Method	Endpoint	Description
GET	/api/v2/workflow_jobs/	List workflow jobs
GET	/api/v2/workflow_jobs/{id}/	Retrieve workflow job
POST	/api/v2/workflow_jobs/{id}/cancel/	Cancel workflow
POST	/api/v2/workflow_jobs/{id}/relaunch/	Relaunch workflow

List Workflow Jobs

curl -X GET \
  https://awx.example.com/api/v2/workflow_jobs/ \
  -H "Authorization: Bearer YOUR_TOKEN"

Retrieve Workflow Job

curl -X GET \
  https://awx.example.com/api/v2/workflow_jobs/50/ \
  -H "Authorization: Bearer YOUR_TOKEN"

id · integer

Workflow job ID

name · string

Workflow job name

status · string

Status: new, pending, waiting, running, successful, failed, error, canceled

started · string

Start timestamp

finished · string

Completion timestamp

elapsed · number

Elapsed time in seconds

workflow_job_template · integer

Source workflow template ID

launch_type · string

How workflow was launched: manual, relaunch, scheduled, dependency, webhook

extra_vars · string

Extra variables used

related · object

Links to related resources:

• workflow_job_template - Source template
• workflow_nodes - Workflow execution nodes
• labels - Job labels
• activity_stream - Activity log
• notifications - Notifications sent
• cancel - Cancel endpoint
• relaunch - Relaunch endpoint

Workflow Nodes

Get the execution graph:

curl -X GET \
  https://awx.example.com/api/v2/workflow_jobs/50/workflow_nodes/ \
  -H "Authorization: Bearer YOUR_TOKEN"

Each node includes:

• id - Node ID
• job - Associated job/update ID
• unified_job_template - Template that was executed
• identifier - Node identifier
• do_not_run - Whether node was skipped
• success - Whether node succeeded
• failed - Whether node failed
• related.success_nodes - Nodes executed on success
• related.failure_nodes - Nodes executed on failure
• related.always_nodes - Nodes always executed

Cancel Workflow

curl -X POST \
  https://awx.example.com/api/v2/workflow_jobs/50/cancel/ \
  -H "Authorization: Bearer YOUR_TOKEN"

Cancels all running jobs in the workflow.

Relaunch Workflow

curl -X POST \
  https://awx.example.com/api/v2/workflow_jobs/50/relaunch/ \
  -H "Authorization: Bearer YOUR_TOKEN"

Labels

curl -X GET \
  https://awx.example.com/api/v2/workflow_jobs/50/labels/ \
  -H "Authorization: Bearer YOUR_TOKEN"

Notifications

curl -X GET \
  https://awx.example.com/api/v2/workflow_jobs/50/notifications/ \
  -H "Authorization: Bearer YOUR_TOKEN"

Activity Stream

curl -X GET \
  https://awx.example.com/api/v2/workflow_jobs/50/activity_stream/ \
  -H "Authorization: Bearer YOUR_TOKEN"

Filtering

# By status
?status=successful

# By workflow template
?workflow_job_template=5

# By date range
?created__gte=2024-01-01

# By launch type
?launch_type=scheduled

Monitor Workflow Execution

import requests
import time

base_url = "https://awx.example.com/api/v2"
token = "YOUR_TOKEN"
headers = {"Authorization": f"Bearer {token}"}
workflow_job_id = 50

def get_workflow_status():
    response = requests.get(
        f"{base_url}/workflow_jobs/{workflow_job_id}/",
        headers=headers
    )
    return response.json()

def get_workflow_nodes():
    response = requests.get(
        f"{base_url}/workflow_jobs/{workflow_job_id}/workflow_nodes/",
        headers=headers
    )
    return response.json()['results']

# Monitor workflow
while True:
    workflow = get_workflow_status()
    status = workflow['status']
    
    print(f"Workflow status: {status}")
    
    # Get node statuses
    nodes = get_workflow_nodes()
    for node in nodes:
        node_status = "pending"
        if node.get('job'):
            node_status = "completed"
        if node.get('do_not_run'):
            node_status = "skipped"
        
        print(f"  Node {node['identifier']}: {node_status}")
    
    if status in ['successful', 'failed', 'error', 'canceled']:
        print(f"\nWorkflow finished: {status}")
        break
    
    time.sleep(5)