def get_learner(
url: str, log_fname: str, job_id: str, job_name: str
) -> tuple[BaseLearner, str | list[str]]:
"""Get a learner from the database running at `url` and this learner's
process will be logged in `log_fname` and running under `job_id`.
Parameters
----------
url : str
The url of the database manager running via
(`adaptive_scheduler.server_support.manage_database`).
log_fname : str
The filename of the log-file. Should be passed in the job-script.
job_id : str
The job_id of the process the job. Should be passed in the job-script.
job_name : str
The name of the job. Should be passed in the job-script.
Returns
-------
learner : `adaptive.BaseLearner`
Learner that is chosen.
fname : str
The filename of the learner that was chosen.
"""
_add_log_file_handler(log_fname)
log.info(
"trying to get learner", job_id=job_id, log_fname=log_fname, job_name=job_name
)
with ctx.socket(zmq.REQ) as socket:
socket.setsockopt(zmq.LINGER, 0)
socket.setsockopt(zmq.SNDTIMEO, 300_000) # timeout after 300s
socket.connect(url)
socket.send_serialized(("start", job_id, log_fname, job_name), _serialize)
log.info("sent start signal, going to wait 60s for a reply.")
socket.setsockopt(zmq.RCVTIMEO, 300_000) # timeout after 300s
reply = socket.recv_serialized(_deserialize)
log.info("got reply", reply=str(reply))
if reply is None:
msg = "No learners to be run."
exception = RuntimeError(msg)
log_exception(log, msg, exception)
raise exception
elif isinstance(reply, Exception):
log_exception(log, "got an exception", exception=reply)
raise reply
else:
learner, fname = reply
log.info("got fname and learner")
log.info("picked a learner")
return learner, maybe_lst(fname)
def tell_done(url: str, fname: str) -> None:
"""Tell the database that the learner has reached it's goal.
Parameters
----------
url : str
The url of the database manager running via
(`adaptive_scheduler.server_support.manage_database`).
fname : str
The filename of the learner that is done.
"""
log.info("goal reached! 🎉🎊🥳")
with ctx.socket(zmq.REQ) as socket:
socket.connect(url)
socket.send_serialized(("stop", fname), _serialize)
socket.setsockopt(zmq.RCVTIMEO, 10_000) # timeout after 10s
log.info("sent stop signal, going to wait 10s for a reply", fname=fname)
socket.recv_serialized(_deserialize) # Needed because of socket type