Hello,
I'm looking for advice on how to use the schedule in the dashboard. I think that when it is working, it should be the case that:
- Experiments remain in the schedule while they are running
- When one experiment is running and another is submitted, the second one is added to the queue in the schedule
- Experiments can be halted by right clicking on them in the schedule and requesting termination
Currently, experiments that we submit in the dashboard briefly appear in the schedule but vanish before they stop running. We can't halt the experiments using the "terminate instances" button in the gui, or by right clicking and requesting termination in the schedule. This is the case whether we run the experiment script or we use the scheduler method to submit the experiment (see code below). We do not use the prepare() or analyze() methods in our experiments, only run().
When we set a due date for an experiment, it does appear in the schedule until the due date arrives, and before the due date we can delete/terminate them.
Here is our current scheduler script:
from artiq.experiment import *
class Scheduler(EnvExperiment):
def build(self):
self.setattr_device("core")
self.setattr_device("scheduler")
def run(self):
expid={
"repo_rev": None,
"file": "TEST_zot_test.py",
"class_name": "zot_test",
"log_level": 30,
"arguments": None
}
self.scheduler.submit("main", expid)
How can we make it possible to pause/end experiments before they finish running? Is it possible to make experiments stay in the schedule until they are completed?