Hi @<1523701260895653888:profile|QuaintJellyfish58>
You mean some "daemon service" aborting Tasks that do not end after X hours? or is it based on CPU/GPU utilization?

i see thanks for the answer, i will read that reference.

my case more like there is a task/process that running but somehow its takes too long to completed. it can be because connection issue forgot to put connection timeout, a problem connection database, etc that makes status still running, but its traped in a situation like that.

so i want to force shutdown a task to failed if that happen

it should be fairly easy to write such a daemon

from clearml.backend_api.session.client import APIClient
client = APIClient()
timestamp = time() - 60 * 60 * 2  # last 2 hours
tasks = client.tasks.get_all(
            status=["in_progress"],
            only_fields=["id"],
            order_by=["-last_update"],
            page_size=100,
            page=0,
            created =[">{}".format(datetime.utcfromtimestamp(timestamp))],
        )
...

references:
None
None