Some jobs occasionally cannot be executed

[yuqiuwen]

Things to check first

• I have checked that my issue does not already have a solution in the FAQ

• I have searched the existing issues and didn't find my bug already reported there

• I have checked that my bug is still present in the latest release

Version

3.10.4

What happened?

I defined a timed job function, and set the job_id as aps_job:pub_bp:{bp_id}. However, some job_id occasionally cannot be executed, which can be queryed in the apscheduler_jobs table. There are no errors printed.

def publish_bp(bp_id: int):
    ...

scheduler.add_job(
    publish_bp,
    id=job_id,
    name=job_id,
    args=[bp_id],
    run_date=DT.ts2time(ptime),
    misfire_grace_time=600,
)

How can we reproduce the bug?

this is apscheduler config:

SCHEDULER_JOBSTORES = {"default": SQLAlchemyJobStore(url=SQLALCHEMY_BINDS["main"])}
SCHEDULER_EXECUTORS = {
    "default": ThreadPoolExecutor(20),
    "processpool": ProcessPoolExecutor(10),
}
SCHEDULER_JOB_DEFAULTS = {
    "coalesce": False,
    "replace_existing": True,
    "max_instances": 20,
}

[agronholm]

If you were me, could you reproduce this issue with the information you've provided?

[yuqiuwen]

If you were me, could you reproduce this issue with the information you've provided?

Actually, the scheduler was not alive using celery periodic task to test it, and there are no errors printed.

@shared_task()
def monitor_apscheduler():
    if not scheduler.running:
        logger.critical("scheduler was crashed, try restarting...")
        scheduler.start()

[agronholm]

So what do you expect me to do? If the scheduler isn't running, it's obviously not going to run your jobs. And when it restarted, isn't it likely that the misfire grace time had passed so the job was not run because of that?

[yuqiuwen]

My question is that I don't know why the scheduler crashed, and why there is no log when it crashed.

[agronholm]

I don't have your computer, so how would I know that? If a Python interpreter crashes, something is always written to the log. If that's not happening, then I have nothing to work with.

[agronholm]

By the way, have you read this yet? https://apscheduler.readthedocs.io/en/3.x/userguide.html#troubleshooting