How to debug APScheduler when callback functions don't run

[Braun Brelin]

Hi all,

 

 

I've got a couple of mysterious issues with APScheduler 2.1.0

I've created an application running as a daemon that reads jobs from a text file called 'schedtab'  and loads them into APScheduler

 

The APScheduler no longer calls the callback function that I've been using for the last few weeks.

It fails silently, that is, it simply doesn't execute the callback_method. I know this because I've put in debugging statements

as the first line of code in the callback method. Since there aren't any errors logged , I have no idea how to debug this problem.

i've tried putting in debugging statements into the threadpool.py code where I'm guessing it runs the scheduled jobs as threads.

but nothing outputs.  running the print_jobs method shows that the job(s) are scheduled correctly, but when the time comes to run...

nothing.

 

 

I also have a second issue.

 

The first time that I load the file in and add jobs (via the default jobstore and the cron trigger)

I get a job schedule that looks normal, i.e. it puts the jobs into the schedule correctly.  However, if I reload the

jobs file, the next_run attribute in all of the  jobs changes to None and no jobs will run. Before reloading the file,

I clear all of the jobs out via a method that looks like this:

 

jobs = self.sched.get_jobs()   # sched is the Schedule object

for job in jobs:

    self.sched.unschedule_job(job)

 

I then reload the file with the jobs again via the add_cron_job method.   The reason for this is that it is possible that jobs

may have been deleted out of the  text file so I figured it's simply easiest to wipe out everything and reload the jobs

from scratch. 

 

Any ideas on how I can debug these problems?

 

Thanks,

 

Braun Brelin

 

 

.

 

 

 

[Braun Brelin]

So, i've narrowed down the problem to the fact that running APScheduler inside a daemon doesn't seem to work correctly.

 

Here's some code examples:

 

This code works...

 

#!/usr/bin/env python

import sys

import time

import daemon

from apscheduler.scheduler import Scheduler

 

def main():

    while (True):

         pass

 

class testclass:

    def test_func(self,*args):

         for arg in args:

             print arg

   

    def __init__(self):

        self.s = Scheduler()

        self.s.add_cron_job(self.test_func,year=2013,month=05,day=10,hour=19,minute=53,second=00,args=("Hello World!\n","foo))

          self.s.start()

 

 

t = testclass()

main()

 

 

 

 

this code fails:

 

#!/usr/bin/env python

import sys

import time

import daemon

from apscheduler.scheduler import Scheduler

def main():

     while (True):

          pass

class testclass:

def test_func(self,*args):

for arg in args:

print arg

def __init__(self):

self.s = Scheduler()

self.s.add_cron_job(self.test_func,year=2013,month=05,day=10,hour=19,minute=53,second=00,args=("Hello World!\n","foo))

self.s.start()

t = testclass()

context = Daemon.DaemonContext(

            stdout = sys.stdout,

            stderr = sys.stderr

)

 

with context:

    main()

 

It's not an issue with the daemon closing stdout/stderr, because I can print something from __init__().

Any ideas?

 

Thanks,

 

Braun Brelin

[Alex Grönholm]

Have you tried starting the scheduler thread *after* you've daemonized the process?

 

Thanks,

 

Braun Brelin

 

 

.

 

 

 

--
You received this message because you are subscribed to the Google Groups "APScheduler" group.
To unsubscribe from this group and stop receiving emails from it, send an email to apscheduler...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[Braun Brelin]

Alex,

 

While putting the apscheduler instantiation inside the main function works, this leads me to all sorts of other issues.

For example, if I have code that looks like this:

 

import daemon

class bar():

     s = Scheduler()

     s.start()

    def print_sched(self):

         s.print_jobs()

def sigusr1_handler(signum,frame):

 

    foo.print_sched()

def main():

     foo = bar()

context = daemon.DaemonContext()

context.signal_map = {

     signal.SIGUSR1: sigusr1_handler

}

with context:

     if (__name__="__main__"):

         main()

 

I'll get a NameError when I send a SIGUSR1 to the program as it can't find foo which is defined in the main function.

That's why I was trying to instantiate the class before I created a daemon.  Do you know why the APScheduler

won't work when instantiated before the code becomes a daemon?

 

Thanks,

 

Braun

[Alex Grönholm]

16.05.2013 15:37, Braun Brelin kirjoitti:

Alex,

 

While putting the apscheduler instantiation inside the main function works, this leads me to all sorts of other issues.

For example, if I have code that looks like this:

 

import daemon

class bar():

     s = Scheduler()

     s.start()

    def print_sched(self):

         s.print_jobs()

def sigusr1_handler(signum,frame):

 

    foo.print_sched()

def main():

     foo = bar()

context = daemon.DaemonContext()

context.signal_map = {

     signal.SIGUSR1: sigusr1_handler

}

with context:

     if (__name__="__main__"):

         main()

 

I'll get a NameError when I send a SIGUSR1 to the program as it can't find foo which is defined in the main function.

That's why I was trying to instantiate the class before I created a daemon.  Do you know why the APScheduler

won't work when instantiated before the code becomes a daemon?

You're still starting the scheduler before daemonization. Nothing has changed.

[Braun Brelin]

Perhaps I'm mis-reading the code, but the instantiation of the scheduler happens in the main function, does it not?

I.e. main doesn't get called until the "with context:...main()"?  Does that not mean that the process has already daemonized itself?

 

Braun

[Alex Grönholm]

16.05.2013 16:47, Braun Brelin kirjoitti:

Perhaps I'm mis-reading the code, but the instantiation of the scheduler happens in the main function, does it not?

No, it happens while the class "bar" is being defined, in: s.start()