How to create and update deployments¶

This guide focuses on deployment lifecycle operations in IronFlow's self-hosted API:

create deployments
update deployments
trigger deployment runs
enable interval or cron schedules

It assumes the API is already running (see How to run the server and UI).

1. List existing deployments¶

curl -s http://127.0.0.1:8000/api/deployments | python -m json.tool

2. Create a deployment¶

Use POST /api/deployments to register a deployment for a known flow.

curl -s -X POST http://127.0.0.1:8000/api/deployments \
  -H 'Content-Type: application/json' \
  -d '{
    "name": "my-simple",
    "flow_name": "simple_flow",
    "default_parameters": {"n": 4},
    "paused": false
  }' | python -m json.tool

Schedule fields are also accepted on create:

schedule_enabled (bool)
schedule_interval_seconds (positive integer)
schedule_cron (cron expression)
schedule_next_run_at (RFC3339 timestamp)

3. Update a deployment¶

Use PATCH /api/deployments/{id} to change deployment settings, including schedules.

curl -s -X PATCH http://127.0.0.1:8000/api/deployments/DEPLOYMENT_ID \
  -H 'Content-Type: application/json' \
  -d '{
    "default_parameters": {"n": 8},
    "paused": false
  }' | python -m json.tool

Schedule update example (cron)¶

curl -s -X PATCH http://127.0.0.1:8000/api/deployments/DEPLOYMENT_ID \
  -H 'Content-Type: application/json' \
  -d '{
    "schedule_enabled": true,
    "schedule_cron": "*/10 * * * *"
  }' | python -m json.tool

Schedule update example (interval)¶

curl -s -X PATCH http://127.0.0.1:8000/api/deployments/DEPLOYMENT_ID \
  -H 'Content-Type: application/json' \
  -d '{
    "schedule_enabled": true,
    "schedule_interval_seconds": 300
  }' | python -m json.tool

Interval and cron are mutually exclusive in deployment state. Setting one schedule type clears the other.

4. Trigger a deployment run manually¶

curl -s -X POST http://127.0.0.1:8000/api/deployments/DEPLOYMENT_ID/run \
  -H 'Content-Type: application/json' \
  -d '{"parameters": {"n": 2}}' | python -m json.tool

5. Observe run progress¶

curl -s http://127.0.0.1:8000/api/deployment-runs | python -m json.tool
curl -s http://127.0.0.1:8000/api/flow-runs | python -m json.tool

With the default embedded worker enabled, deployment runs progress from SCHEDULED to execution in the same process.

Notes¶

Cron scheduling is Rust-first when the native engine is available with DB binding.
Without that Rust path, cron schedules may require schedule_next_run_at to be provided explicitly.
IronFlow currently provides a local subset of deployment/worker behavior, not full Prefect Cloud parity.