Paco Ibañez06/09/2023, 3:45 PM
env variable. Any ideas? Thank you!
Brian06/09/2023, 4:04 PM
It looks like a decorated task has no access to the actual
@task(trigger=always_run) def notify(job_name, **kwargs): tasks = kwargs['upstream_tasks'] if any(s.is_failed() for s in tasks.values()): send_fail_notification(job_name) else: send_success_notification(job_name)
property. I think I could do something similar by crating a custom trigger (which is passed a dictionary of tasks), but I'd like the logic to be in an actual task so that it shows up in the UI and schematics. Our current implementation is to define separate tasks for
The issue with that solution is that the
@task() def notify_success(): # ... @task(trigger=any_failed): def notify_failed(): # ...
task will itself fail if there are not any failed tasks, which means the Flow will fail unless we manually override the reference tasks to include everything except
. That works, but is messy and error prone, since any new tasks have to get added to three different task lists.
Dominic Tarro06/09/2023, 5:03 PM
Peter Nagy06/09/2023, 5:46 PM
method of the context manager is called shortly after the ECSTask starts running - so the task does not actually run. However,
works fine, and the ECSTask execution only starts when the EC2 is up. Could anyone explain me why this is happening and how I could fix it?
with ECSScaler(print_func=<http://logger.info|logger.info>) as scaler: ECSTask(...).run()
Mitch06/09/2023, 7:55 PM
skrawczyk06/09/2023, 8:00 PM
PATCH function, but the docs themselves don't give much information about how to use the command or any example of it being used https://app.prefect.cloud/api/docs#tag/Work-Pools/operation/update_work_pool_api_accounts_[…]t_id__workspaces__workspace_id__work_pools__name__patch
Update Work Pool
Tom A06/09/2023, 8:20 PM
Farhood Etaati06/10/2023, 12:51 PM
name of the flow during its runtime via variables? I don't want to hardcode a storage config in my flow code. Tried using an environment variable + deployment env overriding but apparently it's not working.
Marco Barbero Mota06/11/2023, 4:24 AM
sqlalchemy.exc.OperationalError: (sqlite3.OperationalError) database is locked
. I removed all orphan processes because I suspected the disconnection left something runnign that was accessing the SQLite database from prefect… and nothing. I also tried removing the database. It is neither a question of flow size because I also tried to run a very simple test flow and it gives the same error. Any help is really appreciated, this is extremely annoying. I basically cant run any of the code I had set up with prefect :(
(Background on this error at: <https://sqlalche.me/e/14/e3q8>)
Denys Volokh06/11/2023, 8:38 AM
What could be wrong with workflows? I use github as a storage
Downloading flow code from storage at ''
Ahmed Maged06/11/2023, 4:10 PM
Joshua Greenhalgh06/11/2023, 4:22 PM
Joshua Greenhalgh06/12/2023, 7:37 AM
Giacomo Chiarella06/12/2023, 1:05 PM
Devin06/12/2023, 1:34 PM
Giacomo Chiarella06/12/2023, 2:03 PM
Giacomo Chiarella06/12/2023, 2:24 PM
Jozo06/12/2023, 3:05 PM
GOWRAV TATA06/12/2023, 3:46 PM
Sean Conroy06/12/2023, 3:51 PM
Received response with retryable status code 429. Another attempt will be made in 2.192165185366095s. This is attempt 1/6.
GOWRAV TATA06/12/2023, 3:52 PM
Romain Vincent06/12/2023, 4:16 PM
where all my dependancies are installed for instance) and push to, say, a s3 storage block? Will the agent use this environment? • Should I use the new project related
pull step within my
file? In this case, can I pass a
instead? • Should I make sure that, whatever happens, my agent runs into a python environment that shares the same dependancies as my deployment? • Is there another way I don't see? 🤔
Jehan Abduljabbar06/12/2023, 5:47 PM
James Ashby06/12/2023, 7:38 PM
Gregory Hunt06/12/2023, 8:34 PM
Sean Conroy06/12/2023, 8:51 PM
Jarvis Stubblefield06/12/2023, 11:01 PM