Duration timer never stops after 1st run

[lzzzam]

Tried a basic pipeline to check how return values are used in the GUI. The first run of the following simple code seems to generate unstable behaviour:

@task
def test():
    return 10

@task
def test1():
    return 20

@task
def test2():
    return 30

register_pipeline(
    id="test",
    description="simple pipeline fails",
    tasks=[test, test1, test2],
)

Here the output log:

INFO:     Will watch for changes in these directories: ['/Users/luca/Documents/JobScraper']
INFO:     Uvicorn running on http://127.0.0.1:8000 (Press CTRL+C to quit)
INFO:     Started reloader process [33667] using StatReload
INFO:     Started server process [33669]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO:     127.0.0.1:60912 - "GET / HTTP/1.1" 304 Not Modified
INFO:     127.0.0.1:60912 - "GET /assets/index-86ef8bb9.js HTTP/1.1" 304 Not Modified
INFO:     127.0.0.1:60912 - "GET /assets/index-4fe84deb.css HTTP/1.1" 304 Not Modified
INFO:     127.0.0.1:60913 - "GET /assets/vendor-3fad8192.js HTTP/1.1" 304 Not Modified
INFO:     127.0.0.1:60913 - "GET /api/auth/whoami HTTP/1.1" 200 OK
INFO:     127.0.0.1:60913 - "GET /api/pipelines/ HTTP/1.1" 200 OK
INFO:     ('127.0.0.1', 60914) - "WebSocket /api/ws/" [accepted]
INFO:     connection open
INFO:     127.0.0.1:60913 - "GET /api/runs/?pipeline_id=&trigger_id= HTTP/1.1" 200 OK
INFO:     127.0.0.1:60912 - "GET /api/pipelines/ HTTP/1.1" 200 OK
INFO:     127.0.0.1:60913 - "GET /api/runs/?pipeline_id=&trigger_id= HTTP/1.1" 200 OK
INFO:     127.0.0.1:60916 - "GET /api/pipelines/ HTTP/1.1" 200 OK
INFO:     127.0.0.1:60915 - "GET /api/runs/?pipeline_id=&trigger_id= HTTP/1.1" 200 OK
INFO:     127.0.0.1:60915 - "GET /api/pipelines/test/input-schema HTTP/1.1" 200 OK
INFO:     127.0.0.1:60915 - "POST /api/runs/ HTTP/1.1" 200 OK
Executing pipeline `test` #1 via trigger `_manual`
Executing task test
Executing task test1
Executing task test2
INFO:     127.0.0.1:60915 - "GET /api/pipelines/test HTTP/1.1" 200 OK
INFO:     ('127.0.0.1', 60917) - "WebSocket /api/ws/" [accepted]
Task exception was never retrieved
future: <Task finished name='Task-33' coro=<WebSocket.send_json() done, defined at /Users/luca/Documents/JobScraper/venv/lib/python3.10/site-packages/starlette/websockets.py:168> exception=ConnectionClosedOK(Close(code=1005, reason=''), Close(code=1005, reason=''), True)>
Traceback (most recent call last):
  File "/Users/luca/Documents/JobScraper/venv/lib/python3.10/site-packages/starlette/websockets.py", line 173, in send_json
    await self.send({"type": "websocket.send", "text": text})
  File "/Users/luca/Documents/JobScraper/venv/lib/python3.10/site-packages/starlette/websockets.py", line 85, in send
    await self._send(message)
  File "/Users/luca/Documents/JobScraper/venv/lib/python3.10/site-packages/starlette/middleware/exceptions.py", line 65, in sender
    await send(message)
  File "/Users/luca/Documents/JobScraper/venv/lib/python3.10/site-packages/uvicorn/protocols/websockets/websockets_impl.py", line 320, in asgi_send
    await self.send(data)  # type: ignore[arg-type]
  File "/Users/luca/Documents/JobScraper/venv/lib/python3.10/site-packages/websockets/legacy/protocol.py", line 635, in send
    await self.ensure_open()
  File "/Users/luca/Documents/JobScraper/venv/lib/python3.10/site-packages/websockets/legacy/protocol.py", line 935, in ensure_open
    raise self.connection_closed_exc()
websockets.exceptions.ConnectionClosedOK: received 1005 (no status code [internal]); then sent 1005 (no status code [internal])
INFO:     connection open
INFO:     connection closed
INFO:     127.0.0.1:60916 - "GET /api/runs/1 HTTP/1.1" 200 OK
INFO:     127.0.0.1:60916 - "GET /api/runs/1/logs HTTP/1.1" 200 OK
INFO:     127.0.0.1:60918 - "GET /api/pipelines/test HTTP/1.1" 200 OK
INFO:     127.0.0.1:60920 - "GET /api/runs/1/logs HTTP/1.1" 200 OK
INFO:     127.0.0.1:60919 - "GET /api/runs/1 HTTP/1.1" 200 OK

[lzzzam]

UPDATE: it seems that adding a little delay into each task seems to fix problem, something like

time.sleep(0.2)

[lucafaggianelli]

hi, thanks for reporting this, I'll check what's going on.

[lucafaggianelli]

After some investigations on different OS and browsers I wasn't able to reproduce this issue, though it seems bound to the websocket and not to the pipeline executor, I'll try to handle those websocket connection issues!