Checking an execution

When the automation runs in the background — or when it went past the budget of waiting for the result — the call returns an exec_uuid and the run carries on. This route is how you follow it and pick up the result later.

GET https://mwm.datallog.com/api/v4/execution/<exec_uuid>/status
Authorization: Token <YOUR_TOKEN>
X-Api-Key: <YOUR_X_API_KEY>

The response

{
  "exec_uuid": "0199c2f1-7a3b-7e41-9f0a-3c5d2b8e6a11",
  "finished": true,
  "error": false,
  "stopped": false,
  "container_error": false,
  "result_url": "https://…",
  "logs_url": "https://…"
}

Field	What it says
finished	the run is over (with or without an error)
error	the automation raised an error
stopped	someone stopped the run from the dashboard
container_error	the container died before the automation answered
result_url	link to download the result, or null
logs_url	link to download the log, or null

Both links last 1 minute

result_url and logs_url are signed addresses valid for 60 seconds, and they only appear once finished is true — and only when the automation is set to keep the result and the log. Download them as soon as you get them; to fetch again, call the route again.

Polling until it is done

import time, requests

HEADERS = {"Authorization": f"Token {TOKEN}", "X-Api-Key": KEY}
url = f"https://mwm.datallog.com/api/v4/execution/{exec_uuid}/status"

while True:
    state = requests.get(url, headers=HEADERS).json()
    if state["finished"]:
        break
    time.sleep(5)

if state["error"] or state["container_error"]:
    print("it failed; the log is at", state["logs_url"])
else:
    print(requests.get(state["result_url"]).json())

Wait between calls. The run you are following does not go faster for being asked more often, and every check is an API call like any other.

An execution that does not exist, or belongs to another account, answers 404 with {"error": "Execution not found"} — the same answer in both cases, on purpose.